BUS1003H: Introduction to

Financial Risk
Course Reader

Compiled by the Actuarial Science Section

University of Cape Town
February 26, 2020
Contents

0 Course Structure 1

1 Risk and Value 5

1 Financial Risk 7
1.1 What do Actuaries and Quants do? . . . . . . . . . . . . . . . . . . . 7
1.2 Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.2.1 Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.2.2 Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2.3 Defining Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3 Actuarial Control Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2 Value 17
2.1 Cashflow patterns of assets . . . . . . . . . . . . . . . . . . . . . . . 17
2.1.1 Fixed-interest securities . . . . . . . . . . . . . . . . . . . . . 18
2.1.2 Equities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2 Interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.2.1 Compound vs. Simple interest . . . . . . . . . . . . . . . . . 23
2.2.2 Nominal and effective interest rates . . . . . . . . . . . . . . 24
2.3 Single cashflows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.3.1 Future values . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.3.2 Present values . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.3.3 Relationship between interest, term and present and future
values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.3.4 Comparing single cashflows . . . . . . . . . . . . . . . . . . 29
2.4 Series of regular cashflows . . . . . . . . . . . . . . . . . . . . . . . 31
2.4.1 Present values of level annuities certain . . . . . . . . . . . 33
2.4.2 Accumulations of level annuities certain . . . . . . . . . . . 35
2.4.3 Pricing bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.4.4 Perpetuities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.4.5 Increasing annuities certain . . . . . . . . . . . . . . . . . . 38
2.4.6 Pricing equities . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.4.7 Deferred series of cashflows . . . . . . . . . . . . . . . . . . 45
2.5 Equations of value . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

iii
iv CONTENTS

2.5.1 Solving for the yield (i ) . . . . . . . . . . . . . . . . . . . . . 47

2.5.2 Solving for the number of payments or the term . . . . . . . 49
2.6 Risk discount rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.7 Practical applications . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.7.1 Comparing assets to each other using present values . . . . 55
2.7.2 Accounting for risk . . . . . . . . . . . . . . . . . . . . . . . . 56
2.7.3 Solving for interest rates . . . . . . . . . . . . . . . . . . . . . 57
2.7.4 Identifying cashflow patterns . . . . . . . . . . . . . . . . . . 57
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

2 Uncertainty and Risk Management 67

3 Uncertainty 69
3.1 How do we measure uncertainty? . . . . . . . . . . . . . . . . . . . 70
3.2 Elementary Probability . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.3 Intrinsic vs. estimated probability . . . . . . . . . . . . . . . . . . . 74
3.4 Expected value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.5 Expected values of cashflows at different times . . . . . . . . . . . 78
3.6 Case studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.6.1 Case Study 1: Death . . . . . . . . . . . . . . . . . . . . . . . 81
3.6.2 Case Study 2: Property damage . . . . . . . . . . . . . . . . 82

4 Managing Risks 85
4.1 Types of risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.2 Pure vs. speculative risks . . . . . . . . . . . . . . . . . . . . . . . . 86
4.3 Dynamic vs. static risks . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.4 Fundamental vs. particular risks . . . . . . . . . . . . . . . . . . . . 88
4.5 Risk management . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.5.1 Acceptance . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.5.2 Avoidance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.5.3 Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.5.4 Self-funding . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.5.5 Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.5.6 Risk sharing (or pooling) . . . . . . . . . . . . . . . . . . . . . 93
4.5.7 Risk transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.5.8 Risk pooling in detail . . . . . . . . . . . . . . . . . . . . . . . 94

5 Insurance Principles 99
5.1 What is Insurance? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.2 Why Use Insurance? . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.2.1 Risk Aversion . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.2.2 Risk Pooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.2.3 Economies of Scale . . . . . . . . . . . . . . . . . . . . . . . 102
5.2.4 Protection from unacceptable risks . . . . . . . . . . . . . . 103
CONTENTS v

5.2.5 Better use for capital . . . . . . . . . . . . . . . . . . . . . . . 104

5.2.6 Smoothing of cashflows . . . . . . . . . . . . . . . . . . . . . 105
5.2.7 Social perspective . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.3 Insurable Risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
5.3.1 It has to be a risk . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.3.2 Ideally, the risk should be a pure risk and not a speculative
risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.3.3 Ideally, the risk should be a static risk and not a dynamic risk 108
5.3.4 Ideally, the risk should be a particular risk not a fundamental
risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
5.3.5 Financial, quantifiable and limited . . . . . . . . . . . . . . 109
5.3.6 Small probability . . . . . . . . . . . . . . . . . . . . . . . . . 109
5.3.7 Large pool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5.3.8 Limited opportunity for anti-selection and moral hazard . 110
5.3.9 Availability of past data . . . . . . . . . . . . . . . . . . . . . 112

6 Insurance Provision 113

6.1 Types of Insurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
6.1.1 Life insurance . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
6.1.2 Disability insurance . . . . . . . . . . . . . . . . . . . . . . . 114
6.1.3 General insurance . . . . . . . . . . . . . . . . . . . . . . . . 114
6.1.4 Pension funds . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
6.1.5 Medical aid schemes . . . . . . . . . . . . . . . . . . . . . . . 115
6.1.6 Reinsurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
6.2 Insurance Providers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
6.2.1 Insurance companies . . . . . . . . . . . . . . . . . . . . . . 116
6.2.2 The state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
6.3 The role of actuaries and quants . . . . . . . . . . . . . . . . . . . . 119
6.3.1 Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
6.3.2 Underwriting . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
6.3.3 Contract design . . . . . . . . . . . . . . . . . . . . . . . . . . 122
6.3.4 Reinsurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
6.3.5 Prudent reserving and capital management . . . . . . . . . 122
6.3.6 Asset-liability management . . . . . . . . . . . . . . . . . . . 123
6.3.7 Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
6.3.8 Expense budgeting . . . . . . . . . . . . . . . . . . . . . . . . 124
6.3.9 Asset Management . . . . . . . . . . . . . . . . . . . . . . . . 124

7 Pricing, Reserving and Emergence of Profits 125

7.1 Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
7.1.1 Pricing Assumptions . . . . . . . . . . . . . . . . . . . . . . . 126
7.1.2 Claims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
7.1.3 Expenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
7.1.4 Premiums . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
7.1.5 Risk margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
vi CONTENTS

7.1.6 Profit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

7.1.7 Calculating the risk premium . . . . . . . . . . . . . . . . . . 133
7.1.8 Calculating the office premium . . . . . . . . . . . . . . . . . 134
7.2 Reserving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
7.2.1 Reserving assumptions . . . . . . . . . . . . . . . . . . . . . 136
7.2.2 Claims, Expenses and Premiums . . . . . . . . . . . . . . . . 137
7.2.3 Profits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
7.2.4 Risk Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
7.2.5 Calculating the reserve . . . . . . . . . . . . . . . . . . . . . . 138
7.3 Emergence of profit . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
7.3.1 Actual claims experience . . . . . . . . . . . . . . . . . . . . 141
7.3.2 Real investment returns . . . . . . . . . . . . . . . . . . . . . 142
7.3.3 Real expense experience . . . . . . . . . . . . . . . . . . . . . 143
7.3.4 Real premium experience . . . . . . . . . . . . . . . . . . . . 143
7.3.5 Building up actual assets to time t . . . . . . . . . . . . . . . 143
7.3.6 Calculating the emerging profit . . . . . . . . . . . . . . . . . 145
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

3 Institutions 151

8 General Insurance 153

8.1 Who needs general insurance, and what types of products are there? 154
8.1.1 Property damage insurance . . . . . . . . . . . . . . . . . . 154
8.1.2 Financial loss . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
8.1.3 Fixed benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
8.1.4 Liability insurance . . . . . . . . . . . . . . . . . . . . . . . . 156
8.1.5 Policies as combinations of cover . . . . . . . . . . . . . . . 157
8.2 Accessing general insurance products . . . . . . . . . . . . . . . . . 157
8.3 Underwriting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
8.4 Cashflows of a general insurance policy . . . . . . . . . . . . . . . 158
8.5 Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
8.5.1 Risk and rating factors . . . . . . . . . . . . . . . . . . . . . . 160
8.5.2 Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
8.5.3 Pricing general insurance policies . . . . . . . . . . . . . . . 164
8.6 Claims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
8.6.1 Excess . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
8.6.2 No claims bonus . . . . . . . . . . . . . . . . . . . . . . . . . 168
8.7 Reserving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
8.8 Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

9 Life Insurance 171

9.1 What are people’s needs and what types of products do life insurers
offer? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
9.1.1 Products related to death . . . . . . . . . . . . . . . . . . . . 172
CONTENTS vii

9.1.2 Products related to survival . . . . . . . . . . . . . . . . . . . 173

9.1.3 Products which pay out on both death and survival . . . . . 175
9.1.4 Products associated with disability . . . . . . . . . . . . . . 176
9.2 Accessing life insurance products . . . . . . . . . . . . . . . . . . . 176
9.2.1 Distribution channels . . . . . . . . . . . . . . . . . . . . . . 176
9.2.2 Individual vs. group business . . . . . . . . . . . . . . . . . . 178
9.3 Underwriting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
9.4 Cashflows of life insurance products . . . . . . . . . . . . . . . . . 179
9.5 Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
9.6 Reserving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
9.7 Claims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
9.8 Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186

10 Reinsurance 189
10.1 Why do insurers need reinsurance? . . . . . . . . . . . . . . . . . . 189
10.1.1 Smoothing of claims experience . . . . . . . . . . . . . . . . 189
10.1.2 Limiting large losses . . . . . . . . . . . . . . . . . . . . . . . 190
10.1.3 Accessing the reinsurer’s expertise . . . . . . . . . . . . . . . 190
10.2 Ways of writing reinsurance business . . . . . . . . . . . . . . . . . 191
10.2.1 Treaty reinsurance . . . . . . . . . . . . . . . . . . . . . . . . 191
10.2.2 Facultative reinsurance . . . . . . . . . . . . . . . . . . . . . 192
10.3 Types of reinsurance . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
10.3.1 Proportional reinsurance . . . . . . . . . . . . . . . . . . . . 193
10.3.2 Non-proportional reinsurance . . . . . . . . . . . . . . . . 200
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206

11 Life Contingencies 207

11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
11.2 Life tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
11.2.1 l x and d x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
11.2.2 p x and q x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
11.2.3 k p x and k q x . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
11.3 Curtate future lifetime . . . . . . . . . . . . . . . . . . . . . . . . . . 213
11.3.1 Defining K x . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
11.3.2 K x as a discrete random variable . . . . . . . . . . . . . . . . 213
11.4 Expected present values of lump sum payments . . . . . . . . . . 215
11.4.1 Pure endowments . . . . . . . . . . . . . . . . . . . . . . . . 215
11.4.2 Whole-life policies . . . . . . . . . . . . . . . . . . . . . . . . 216
11.4.3 Endowment assurance policies . . . . . . . . . . . . . . . . . 217
11.4.4 Term assurance . . . . . . . . . . . . . . . . . . . . . . . . . . 219
11.5 Expected present values of series of payments . . . . . . . . . . . . 222
11.5.1 Whole-life annuities . . . . . . . . . . . . . . . . . . . . . . . 223
11.5.2 Temporary annuities . . . . . . . . . . . . . . . . . . . . . . . 226
11.6 Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
11.6.1 Net Premiums . . . . . . . . . . . . . . . . . . . . . . . . . . 229
viii CONTENTS

11.6.2 Gross premiums . . . . . . . . . . . . . . . . . . . . . . . . . 231

11.7 Reserving: Gross premium policy values . . . . . . . . . . . . . . . 235
11.8 Some more examples . . . . . . . . . . . . . . . . . . . . . . . . . . 238
11.9 A1967–70 (ultimate) . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
11.10English Life Table No. 12 (Males) . . . . . . . . . . . . . . . . . . . . 253

12 Pensions and Related Benefits 257

12.1 What are people’s needs with respect to retirement, and how can
those needs be met? . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
12.1.1 Defined contribution funds: . . . . . . . . . . . . . . . . . . 262
12.1.2 Defined Benefit Funds . . . . . . . . . . . . . . . . . . . . . . 265
12.2 Accessing retirement funds . . . . . . . . . . . . . . . . . . . . . . . 267
12.2.1 State retirement provision . . . . . . . . . . . . . . . . . . . . 268
12.2.2 Occupational retirement funds . . . . . . . . . . . . . . . . . 268
12.2.3 Individual provision . . . . . . . . . . . . . . . . . . . . . . . 268
12.3 Underwriting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
12.4 Setting contribution rates . . . . . . . . . . . . . . . . . . . . . . . . 269
12.5 Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
12.6 Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274

13 Ethics and Professionalism 277

13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
13.2 Professional societies and codes of conduct . . . . . . . . . . . . . 279
13.2.1 The Actuaries’ Code . . . . . . . . . . . . . . . . . . . . . . . 281
13.2.2 The CFA Institute . . . . . . . . . . . . . . . . . . . . . . . . . 282
13.3 What does it mean to be “professional”? . . . . . . . . . . . . . . . 284
13.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287

4 Assets 293

14 Overview of Assets 295

14.1 Why study assets? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
14.2 Matching assets to liabilities . . . . . . . . . . . . . . . . . . . . . . 295
14.3 The investment market . . . . . . . . . . . . . . . . . . . . . . . . . 297
14.4 Characteristics of investment markets . . . . . . . . . . . . . . . . . 298
14.4.1 Domestic vs. international investment . . . . . . . . . . . . 298
14.4.2 Direct vs. indirect investments . . . . . . . . . . . . . . . . . 298
14.4.3 Primary vs. secondary markets . . . . . . . . . . . . . . . . . 299
14.5 Income vs. capital gains . . . . . . . . . . . . . . . . . . . . . . . . . 299
14.6 The SYSTEM T framework . . . . . . . . . . . . . . . . . . . . . . . 300

15 Money Market Instruments 301

15.1 Cash and bank deposits . . . . . . . . . . . . . . . . . . . . . . . . . 302
15.2 Money market securities . . . . . . . . . . . . . . . . . . . . . . . . 305
CONTENTS ix

15.2.1 Treasury bills . . . . . . . . . . . . . . . . . . . . . . . . . . . 306

15.2.2 Commercial Paper . . . . . . . . . . . . . . . . . . . . . . . . 307
15.2.3 Banker’s acceptances and bearer deposit notes . . . . . . . 308
15.2.4 Money market strips . . . . . . . . . . . . . . . . . . . . . . . 308
15.2.5 Negotiable certificate of deposit . . . . . . . . . . . . . . . . 308
15.2.6 Repurchase rate agreements . . . . . . . . . . . . . . . . . . 309
15.3 Money market summary . . . . . . . . . . . . . . . . . . . . . . . . . 310
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312

16 Bonds 313
16.1 Bond Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
16.1.1 Domestic government bonds . . . . . . . . . . . . . . . . . . 314
16.1.2 Corporate bonds . . . . . . . . . . . . . . . . . . . . . . . . . 314
16.1.3 Overseas bonds . . . . . . . . . . . . . . . . . . . . . . . . . . 314
16.2 Types of bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
16.2.1 Zero-coupon bonds . . . . . . . . . . . . . . . . . . . . . . . 315
16.2.2 Fixed-interest coupon paying bonds . . . . . . . . . . . . . . 317
16.2.3 Inflation-linked bonds . . . . . . . . . . . . . . . . . . . . . 324
16.3 Credit ratings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331

17 Equities 333
17.1 Characteristics of equity . . . . . . . . . . . . . . . . . . . . . . . . . 334
17.2 Types of equity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
17.2.1 Ordinary shares . . . . . . . . . . . . . . . . . . . . . . . . . . 335
17.2.2 Preference shares . . . . . . . . . . . . . . . . . . . . . . . . . 336
17.3 Classification of equities . . . . . . . . . . . . . . . . . . . . . . . . . 337
17.3.1 Industry grouping . . . . . . . . . . . . . . . . . . . . . . . . 337
17.3.2 Size of the company . . . . . . . . . . . . . . . . . . . . . . . 338
17.3.3 Exposure to the economic cycle . . . . . . . . . . . . . . . . 340
17.4 Equity valuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
17.5 Evaluating equities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
17.5.1 Earnings per share . . . . . . . . . . . . . . . . . . . . . . . . 343
17.5.2 Price/Earnings ratio . . . . . . . . . . . . . . . . . . . . . . . 343
17.5.3 Dividend yield . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
17.6 Equity indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
17.7 Historic asset class returns in South Africa . . . . . . . . . . . . . . 346
17.8 Summary of equity characteristics . . . . . . . . . . . . . . . . . . . 347

18 Property 349
18.1 Characteristics of property . . . . . . . . . . . . . . . . . . . . . . . 349
18.1.1 Nature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
18.1.2 Marketability . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
18.1.3 Yield . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
18.2 Property markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
x CONTENTS

18.3 Property types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351

18.4 Residential vs. non-residential property markets . . . . . . . . . . 352
18.5 Direct vs. indirect property investments . . . . . . . . . . . . . . . . 352
18.5.1 Direct property investments . . . . . . . . . . . . . . . . . . 352
18.5.2 Indirect property investment . . . . . . . . . . . . . . . . . . 353
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355

19 Derivatives 357
19.1 Derivative Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
19.1.1 Over-the-counter Markets . . . . . . . . . . . . . . . . . . . . 358
19.1.2 Exchange Traded Markets . . . . . . . . . . . . . . . . . . . . 358
19.2 Forwards and futures . . . . . . . . . . . . . . . . . . . . . . . . . . 359
19.3 Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
19.3.1 European call options . . . . . . . . . . . . . . . . . . . . . . 363
19.4 European put options . . . . . . . . . . . . . . . . . . . . . . . . . . 366
19.5 Uses of derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
19.6 Derivative Traders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
19.6.1 Hedgers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
19.6.2 Speculators . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
19.6.3 Arbitrageurs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372

20 Investors and Investment Strategy 373

20.1 Investment strategy considerations . . . . . . . . . . . . . . . . . . 374
20.1.1 Liabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
20.1.2 Risk appetite . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
20.1.3 Diversification . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
20.1.4 Expected returns and volatility of returns of asset classes . 377
20.1.5 Level of free assets . . . . . . . . . . . . . . . . . . . . . . . . 377
20.1.6 Investment restrictions . . . . . . . . . . . . . . . . . . . . . 379
20.1.7 Tax and transaction costs . . . . . . . . . . . . . . . . . . . . 379
20.2 Institutional investors . . . . . . . . . . . . . . . . . . . . . . . . . . 379
20.2.1 Life insurers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
20.2.2 General (short-term) insurers . . . . . . . . . . . . . . . . . 382
20.2.3 Retirement funds . . . . . . . . . . . . . . . . . . . . . . . . . 382
20.2.4 Medical schemes . . . . . . . . . . . . . . . . . . . . . . . . . 383
20.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384

21 Asset and Portfolio Management 385

21.1 Asset allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
21.1.1 Strategic asset allocation . . . . . . . . . . . . . . . . . . . . . 385
21.1.2 Tactical asset allocation . . . . . . . . . . . . . . . . . . . . . 386
21.2 Portfolio construction approaches . . . . . . . . . . . . . . . . . . . 388
21.2.1 Top-down approach . . . . . . . . . . . . . . . . . . . . . . . 388
21.2.2 Bottom-up approach . . . . . . . . . . . . . . . . . . . . . . . 390
CONTENTS xi

21.3 Active and passive portfolio management . . . . . . . . . . . . . . 391

21.3.1 Active portfolio management . . . . . . . . . . . . . . . . . . 391
21.3.2 Passive portfolio management . . . . . . . . . . . . . . . . . 393
21.4 Choosing an investment style and implementation . . . . . . . . . 394
21.5 Monitoring and revision of the portfolio . . . . . . . . . . . . . . . 394
21.5.1 Monitoring the market . . . . . . . . . . . . . . . . . . . . . . 394
21.5.2 Monitoring the investor’s objectives . . . . . . . . . . . . . . 395
21.5.3 Monitoring the asset mix . . . . . . . . . . . . . . . . . . . . 395
21.5.4 Monitoring diversification . . . . . . . . . . . . . . . . . . . . 396
21.5.5 Monitoring the asset manager . . . . . . . . . . . . . . . . . 396
Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
Chapter 0

Course Structure

The objective of this course is to introduce you to the key concepts of actuarial
science and quantitative finance, giving you an overview of the program and
also of your career as future quant or actuary. The course is structured as four
modules:

Module 1: Risk and Value

This module introduces the concept of risk, which is at the core of the actuarial
and quant professions. Risk is related to uncertain future events that have a
financial value. This module is mainly concerned with the concept of value, and
we will focus on how the value of money changes over time due to interest that
can be earned. We consider cashflows and series of cashflows at different points
in time, and how these cashflows can be compared to each other and valued.
We also consider different financial products and the types of cashflows they
represent. At the end of this module, you should be able to:

• Explain what an actuary and quantitative analyst do.

• Explain the concept of risk.

• Understand the concept of time value of money, and be able to do calcula-

tions of the value of cashflows at different points in time.

• Be able to adjust the interest rate to reflect the uncertainty of payment.

• Be able to compare the value of different cashflows, to set a price for differ-
ent cashflows, and to calculate any missing component of the equation of
value.

• Be able to understand the different ways that interest rates can be ex-
pressed.

1
2 Chapter 0 Course Structure

Module 2: Uncertainty and Risk Management

This module delves deeper into the concept of risk, by considering how uncer-
tainty can be estimated and allowed for in the evaluation of risk. The South
African environment is described to demonstrate the types of risk faced by in-
dividuals in our country, and how the government and private institutions help
people manage those risks. Different ways of managing risk are introduced, with
a particular focus on risk transfer and insurance. We discuss the principles of
insurance using a general insurance company as an example. At the end of this
module, you should be able to:

• Estimate probabilities of future events from past data.

• Describe the South African environment with respect to risk and protection
from risk.

• Explain how risk can be managed.

• Explain the principles of risk pooling, risk transfer and insurance.

• Describe how general insurance works.

• Calculate the premiums that a general insurer would charge for various
insurance products.

Module 3: Institutions
This module explores the different institutions which are particularly exposed
to risk because their business is linked to uncertain events. These institutions
include life insurance companies, pension funds, general insurance companies
and reinsurers. Life insurance is based on the study of mortality rates, and we
will discuss how mortality rates and tables are calculated (demography), and how
they can be used to price life insurance policies. This area of study is referred to
as “life contingencies”. Contingencies is the study of valuing cashflows which
are attached to a risk – for example, cashflows which are only paid out when a
policyholder dies, or survives to a certain time. We also consider the ethical and
professional obligations of actuaries working for different institutions. At the end
of this module, you should be able to:

• Describe the operation of a life insurance company, and the risks life insur-
ance business is exposed to.

• Describe the operation of a pension fund, different types of pension funds,

and the risks associated with saving for retirement.

• Describe the operation of a general insurance company, and the risks these
companies are exposed to.
 3

• Describe the operation of a reinsurer, and how a reinsurer interacts with

other insurance providers.

• Be able to use mortality tables to value cashflows which are related to

the death or survival of policyholders, including whole and temporary life
insurance, annuity and endowment policies.

• Assess situations where a professional needs to consider the impact of their

actions on various stakeholders and evaluate which is the most ethical way
to proceed.

Module 4: Assets
This module describes different types of assets that can be used for investment
purposes, including money market instruments, bonds, equities, property, deriva-
tives and other investment vehicles. We then consider different types of investors
and their needs, in terms of the liabilities they need to meet, their attitude to risk,
and other factors. This allows us to evaluate the different types of assets in terms
of their suitability as investments for institutions with certain types of liabilities –
this is called matching the assets to the liabilities. At the end of the module, you
should be able to:

• Describe the characteristics of the major asset classes.

• Be able to value an asset in each of the major asset classes.

• Recommend an investment strategy for different types of investors, or

comment on their existing strategies.

• Describe and analyse the investment needs of different institutional in-

vestors.
 Module 1

Risk and Value

5
Chapter 1

Financial Risk

1.1 What do Actuaries and Quants do?

Actuaries and Quantitative Analysts (we’ll call them quants from here on) are
financial professionals who deal with the financial impact of risk, typically with
a long-term outlook.

Example 1.1
Financial risks are everywhere:

• Someone who owes you money may not pay you back
• The interest rate on your student loan may go up, making your repayments
more expensive
• You may fail a year and have to pay for an additional year of study
• You could crash your car and need to pay for repairs
• You could get a girlfriend/boyfriend - they are expensive!

Think of the financial risks that someone who runs a corner shop may face. What
about the financial risks of a building contractor?

The slogan of the actuarial profession is “Making financial sense of the future”.
This could very well apply to quants, too. How do we “make financial sense of the
future”?

• We analyse the past;

• Model the future (i.e. forecast future financial outcomes);

• Assess and mitigate the risks involved; and

• Communicate what the results mean in financial terms.

7
8 Chapter 1 Financial Risk

Example 1.2
How could you use this process to prepare yourself for the financial risks you are
facing?
Say you are concerned about the financial implications of failing a year:

• Analyse the past: what information could you gather about your chances of
failing? Are there statistics on pass rates you could look up? Ideally it would
be nice to see how students with similar grade 12 marks as yourself have
done in the past. What percentage of them end up failing a year? And what
is the financial impact of failing - how much does it cost to repeat a year?
• Model the future: once you have some information, you can project it for-
ward - has anything changed? Are there new lecturers, a new syllabus? Will it
be easier or harder this year? How is the cost of studying likely to change in
the future?
• Assess and mitigate the risks: What are the risks that affect your perfor-
mance? Is it how much you study, how many lectures you attend, who you
make friends with, how much you use the hot seat? Is there something you
can do to reduce these risks? Attending lectures and actually studying for
What are “Assets” and “Liabili- tests may have some effect on performance!
ties”? • Communicate the results: In this case, you will be communicating to your-
Assets are the possessions of a self, but it may be useful to be able to quantify the risk: say the cost of failing
company. They can include prop- is R50 000, and my self-assessed probability of failing is 10%. If I can reduce
erty, equipment, money in the this risk to 2% by doing certain things, that is a significant reduction and may
bank, investments, and much be worth the extra 4 hours a week spent studying!
more. If you consider your own
situation, your assets are things This is a very simple case. See if you can do a similar risk assessment on the other
like your clothes, computer, and risks listed in the previous example - you may find the risk of dating is too high to
maybe a car. bear!
Liabilities are what a company
owes to other parties, i.e. promises Working with risk is a complex and important function in the financial sector.
it has made to pay or give some-
thing to someone in the future.
Actuaries/quants have to work hard to maintain their reputations as financial
This could be loans which it needs problem-solvers. It is expected that we are:
to repay, or products that have
been paid for but not shipped. An • experts in the analysis and modelling of situations involving financial risk
insurance company, for example, and contingent events (i.e. those which are dependent on other events);
has made a promise to pay out
insurance claims when something • concerned with both the asset and liability (see the box on the left) sides of
bad happens - those promises are the balance sheet (quants are more concerned with assets than liabilities,
its liabilities. In your case, your
but do need a good understanding of both);
liabilities could be a student loan
(or bursary which requires you
• able to provide realistic solutions to complex problems with a long-term
to work later), money you owe
a friend, or the round of drinks forward view;
you promised to buy your friends
when you qualify! • practical, innovative and numerate; and

• able to communicate our findings clearly to different parties.

In order to live up to these expectations, actuaries and quants must have compe-
Table 1.1 Assets and liabilities. tence in a wide range of areas. Some of the key skills required are:
1.1 What do Actuaries and Quants do? 9

• financial and mathematical modelling;

• a solid understanding of the broader economic and business environ-

ment;

• analysis and management of uncertainty;

• evaluation of financial consequences;

• analysis of risk, and risk management;

• scientific pricing and reserving techniques (these are relevant to actuar-

ies only, and refer to managing the products and liabilities of insurance
companies);

• asset-liability management; and

• overall financial management.

We will explore these skills during the course of this year, in order to give you an
idea what kind of work you could be doing in the future. The following quote
from the Actuarial Society of South Africa summarises the role of actuaries well:

“An actuary applies analytical, statistical and mathematical skills

to financial and business problems, especially those which involve
uncertain future events. This helps individuals and businesses to
safeguard their future, confidently and at a fair price, in an ever-
changing world.”

What are the differences between actuaries and quants?

Both professions focus on the analysis of financial risks. The main difference is
that actuaries generally consider both risks related to liabilities and risks related to
assets, where quants are mainly concerned with assets.
Regular companies, like manufacturers and retailers for example, have fairly cer-
tain liabilities - monthly bills which become due, or loan repayment. There is not
much need for sophisticated financial modelling, so you won’t find many actuaries
working there. Actuaries are attracted to companies which have complex, uncer-
tain, often long-term, liabilities, where interesting models are needed to predict
what will happen. That is why, traditionally, actuaries have been very involved with
life insurance and pensions business. In both cases, the liability can have a time
horizon of several decades, and the timing and/or the quantity of the required
payment are unknown - perfect work for actuaries. In recent years, the areas for
actuarial work have expanded to include many entities which traditionally did
not employ actuaries - for example, general insurance companies, banks, and
the government. This has happened as these entities have realised that they do
in fact face long term risks and even in the short term, their risks could be better
evaluated by actuaries.
10 Chapter 1 Financial Risk

Quants don’t analyse liabilities in detail (in fact, in some cases quants work with
actuaries and rely on actuarial analysis of the liabilities in their work with assets).
Instead, they focus on assets at a level not usually attempted by actuaries, mak-
ing use of mathematical and numerical techniques to price financial assets or
facilitate investment decision-making. Quants typically work in the fields of asset
management, investment banking or investment advice.

It is worth noting that the problems actuaries and quants help to solve are
usually financial in nature. The one aspect which requires some discussion is
that of “uncertain future events”. The combination of uncertainty and financial
implications may be summed up in one word: risk.

1.2 Risk
What do we mean when we use the term “risk”?

Risk = Uncertainty and Value.

The two main characteristics of something that is risky are that it is uncertain,
and that the outcome has a financial consequence, or value, attached to it. There
are non-financial risks , but for the purposes of this course, a financial value must
T Can you think of any risks be attached to the outcome in order for us to consider the event “risky”.
which do not have a finan-
cial consequence?
1.2.1 Uncertainty
It is clear what we mean by the outcomes being uncertain. If we are guaranteed
to be paid R100 for a particular job, then by definition there is no risk attached
to the payment: certainty is the opposite of risk. But if there is a chance that our
employer will refuse to make the payment, then we can legitimately speak of the
payment being risky.
Uncertainty can relate to the event happening or not (we do/don’t win the
game; you do/don’t pass this course), but it can also relate to the timing of an
event (when do the oil reserves run out?) and also the value can be uncertain
(what amount of salary are you going to be offered? What will be the cost of
repairs for your car after the accident?)
In order to assess risk, you will have to learn to assess all these types of
uncertainty. This relates to calculating the probability that something happens,
and the probability that it happens at a certain time and the probability that
the loss it causes will have a certain value. These calculations are at the core
of actuarial work, and they are the reason why statistics are a core part of your
curriculum.

Example 1.3
1.2 Risk 11

Think of some examples of uncertain events: weather, sports, politics, economy,

relationships, job prospects, health. . . it’s actually much harder to think of anything
that is certain! What types of uncertainty (occurrence, timing, amount) can you
think of for each of these uncertain events?
What is value?
There is a saying: “Nothing is certain except death and taxes” - but can you think
of how both of these are actually uncertain in some of the above ways, too? Value seems like a very straight-
forward concept – we mostly know
how much something is worth.
A loaf of bread has a sticker on
it with a price; a house for sale
1.2.2 Value has a listed price you can find
out from an estate agent; same
The other important aspect of risk as it is commonly used for actuarial appli- goes for a car. And money is worth
cations is the financial consequence, or value, of the event. In the scenario in what it says on the note – R50 is
section 1.2.1, we expect to receive R100 - that is the value associated with the risk. worth fifty rand, right? Where this
Much of the work you will be doing as a quant/actuary will have to do with gets much more interesting, and
complicated, is when we start
calculating the value of something, so that you can properly assess the risk of it. thinking about the time value of
The greater the value at stake, the greater the risk: the risk of losing R5 is much money. Is R50 in ten year’s time
smaller than the risk of losing R500. worth the same as R50 now? In the
next chapter, you will learn that
money changes value over time,
Example 1.4 and R50 today can be worth R150
in ten year’s time. This is linked
Go back to those examples of uncertain events in Example 1.3: which of them have to the idea of interest. For now
a financial consequence attached to them? keep in mind that since actuaries
and quants are often consider-
Some events have a direct financial consequence: getting a certain job will mean a ing money that’s due in the far
certain paycheck, being ill means expenses on doctors, an increase in the interest future, we don’t take amounts at
rate may mean higher loan repayments. face value – so calculations are
needed to work out what this fu-
Other events may be indirectly related to financial outcomes: a drought could
ture money is worth today.
affect farmers and drive up food prices, making your expenses go up. A sporting
event could have a financial consequence if you bet on it.
Note that the financial consequences of an event are completely different for
Table 1.2 Value.
different people: a drought in South Africa will have huge financial consequences
for South African farmers, lower consequences for South African consumers, and
likely virtually no consequences on consumers in Japan.
! A murky world You may
be used to a clear world of
definitions and concepts
So actuaries and quants are mostly concerned with those parts of running a from school. Now that you
are at university, you will
business that have risky outcomes, i.e. those that are uncertain and have finan-
discover that a concept may
cial consequences. And the greatest risks are those where there is the highest
be defined differently by dif-
probability of making the greatest loss. ferent people - because they
are in different disciplines,
1.2.3 Defining Risk or because they deal with
a different aspect of the
The description of risk as a combination of uncertainty and financial consequence concept. None of these def-
always holds at a high level; but at a detailed level, risk can mean many different initions are “wrong”, they
things to different people. Over the coming years you will encounter many dif- are just best suited to their
ferent definitions and different measures of risk. One of the defining features of circumstances. Be careful
actuaries specifically, however, is that they are concerned with both assets and when encountering similar
terms in different classes or
reference works.
12 Chapter 1 Financial Risk

liabilities, and quite often what we mean by actuarial risk is the possibility of
the assets being insufficient to meet the liabilities at some future point (i.e. not
having enough money to pay what you promised, essentially).
This type of definition generally applies to companies: the risk of not having
enough assets to meet liabilities could for example be the risk of not having
enough money to pay salaries, or an insurer not having enough funds to meet
claims. But even in your personal life, not having enough money to pay for repairs
to someone’s car after you drove into it is an example of not having sufficient
assets to meet liabilities.
A lot of actuarial work is concerned with working out how much money you
need to have on hand that you won’t ever be caught in a situation where you can’t
pay what you owe.
“Risk” is arguably the defining concept of actuarial science and quantitative
finance. It is our ability to understand, model and manage financial risk that sets
our professions apart from others. This process is not a once-off, static exercise;
the dynamic nature of the business and financial environment means that risk
management is an ongoing process. This idea is captured in the very important
concept of the Actuarial Control Cycle (with which you should familiarise your-
selves, as it will be a common theme throughout your studies, both actuarial and
quant).

1.3 Actuarial Control Cycle

The Actuarial Control Cycle is a representation of the dynamic nature of actuarial
work, and a useful framework for actuarial problem-solving. It consists of three
stages:

1. Specifying the problem: solving any actuarial problem means having a

clear specification upfront of what that problem entails. For example, in
pricing (i.e. setting the premium for) a life insurance contract we would
need to know, among other things, what the expected future mortality rates
(the uncertainty part of the calculation - the chance of death) will be, what
future investment returns we expect to earn and how much capital needs
to be held to protect against volatile experience.

2. Developing the solution: we might now employ research, past experience

and actuarial judgement to provide proposed answers to those questions.
In our example, this gives us the tools necessary to price the contract.

3. Monitoring the experience: it is then critical that we observe to see how

things turn out to ensure that the solution we have developed is as close
to correct as possible. For example, we might find that we have seriously
under- or overstated the mortality experience, and need to return to our
models to update our expectations based on the emerging experience. The
importance of this leg of the cycle cannot be overstated: it is central to
1.3 Actuarial Control Cycle 13

actuarial practice. Note that it then feeds back both into the specification
of the problem and the development of the solution.

A graphical representation of the Actuarial Control Cycle is shown below. As

the figure suggests, this process is influenced primarily by two forces: the general
commercial and economic environment and the professionalism expected of an
actuary.
We will discuss professionalism in Module 3, when you have had enough ex-
posure to the types of work that you might be doing to be able to work through the
ethical problems that you may encounter. For now you should simply understand
this term to encompass the behaviours that the public expects of a professional.
The general commercial and economic environment may influence actu-
arial practice in many ways, such as through regulation, tax, the interest rate
environment and technology, to name but a few. We will look at the general
commercial and economic environment in South Africa in Module 2.

Figure 1.1 Actuarial Control Cycle.

Example 1.5 Let’s see if we can apply the Actuarial Control Cycle to our
personal problems (this is why people think actuaries are strange).
Let’s say that the problem is organising a party.

(a) Specify the problem: What is required for a truly splendid party? We
need to know how many people are likely to turn up, what they like
to drink, how many loudspeakers will be required, how many trolleys
full of potato chips, and how much capital will be required to finance
all these fine things. This may be a good time to consider the source
of capital (Dad?), and what the limitations are (R200. No way!).
14 Chapter 1 Financial Risk

(b) Develop the solution: after specifying the problem, you will probably
realise that budget is a big part of the issue, with a secondary problem
being. . . getting the cool kids to come, of course. How do you solve this
problem? Budget problems could be solved by getting more funds, or
cutting costs. More funds – could you charge admission (. . . only if the
cool kids are coming)? Maybe get your friends to pitch in? Are there
charitable organisations that help finance parties? (Tip: No.) What
about cost cutting? One beer is surely like any other? Do we really
need alco-pops? (Yes. . . if you want the cool kids to come). And so on –
until you develop a solution (Promise the cool kids free drinks, charge
everyone else admission?)
(c) Monitor the experience: This is a really important part (and probably
requires you to stay sober!). Monitoring the experience may make
you realise that you did not fully understand the problem: turns out
that the main issue is the supervisory regime (your parents, who are
not leaving the house as promised) which is causing the stakeholders
(that’s your friends) to turn to alternative options (the neighbourhood
bar). This can take you back to re-specifying the problem (How do I
make my parents leave the house?), developing a new solution (reser-
vation at local restaurant), and go back to monitoring the experience.
This monitoring step is crucial – this is where you test your planning
against reality and fine-tune the solution until the party is really rock-
ing. Or whatever you kids are calling it these days.
Exercises 15

Exercises

Exercises for 1 Financial Risk

Ex.1.1 List the financial risks that the following entities may face:

1. A hospital
2. Your tutor
3. A Hollywood movie star
4. A plumber
5. A company that prints business cards

Ex.1.2 Describe the steps you could follow to analyse the past, model the
future, assess and mitigate risks and communicate the solution to the
following problems:

1. Deciding whether to have LASIK eye surgery (a procedure which

fixes eye problems so you don’t need glasses – but it’s quite expen-
sive and some people experience side effects).
2. Deciding whether to buy or rent a house.
3. Deciding between bartending and working in a library as a holiday
job.

Ex.1.3 What type of assets may the following have:

1. Your parents
2. A private hospital
3. A carpenter
4. A university

Ex.1.4 What kind of liabilities may the same parties as in 3 have?

Ex.1.5 What kind of events may put each party in danger of not having enough
assets to meet liabilities?

Ex.1.6 Consider the following situation, and list the uncertainties associated
with each:

1. A storm
2. A mugging
3. A job interview
4. A wedding
5. An exam
16 Chapter 1 Financial Risk

6. An overseas holiday

Ex.1.7 What could the financial consequences of each of the above situations
be? How would they be different for different people?

Ex.1.8 Use the actuarial control cycle to solve the following problems:

1. Getting from Joburg to Cape Town as cheaply as possible

2. Deciding how much to sell your car for
3. Finding a holiday job

Don’t forget to describe the “general economic environment” and the

“professionalism/ethics” for each situation!
Chapter 2

Value

In Chapter 1 we established that Risk is composed of uncertainty and value.

Actuaries and quants rarely consider value as simply the amount of a transaction
or an event. Instead, we are very concerned with the timing of the cashflow, and
the effect that time has on the value of money. This concept is referred to as the
time value of money.

Example 2.1
If you are offered R50 now or R50 in one year’s time, which would you prefer and
why?

Solution: A rational person would prefer to receive the money now as they can use
the money and get access to its value immediately. A harder question to answer
would be, would you prefer R100 now or R110 in one year’s time? Now it is a bit
more difficult to answer the question. One way to compare the two amounts would
be to figure out how much the R100 would grow to in one year if you invested it
now. If you can earn 12% interest, you could grow it into R112. In that case, the
R100 now is worth more than R110 in a year. But if you could only earn 5% interest,
it would be better to wait for one year and take the R110. This example illustrates
that money has different values at different time points.

This chapter is concerned with how time value of money and related concepts
are calculated. The methods for these calculations are called mathematics of fi-
nance. In your second year, there is a full course associated with maths of finance,
where you will refine and build on the concepts introduced in this chapter.
A good example of the application of mathematics of finance is to evaluate
different types of assets.
U Money you receive or pay
out is often referred to as
2.1 Cashflow patterns of assets cashflows. In general, a
positive cashflow will be
Most of the activity in the world of finance is to do with buying and selling assets, money that you receive and
which can be seen as streams of cashflows. You have heard of investment traders a negative cashflow will be
money that you pay out.
17
18 Chapter 2 Value

buying and selling equities, bonds and property. Each of those assets can be seen
as a series of cashflows. For example, if you buy a property, the series of cashflows
that you are buying is the monthly rent you will get from the property, plus the
final price you will get for the property (if and) when you decide to sell it. The
following figure illustrates the cashflow pattern from a property investment.
T This type of diagram is
called a cashflow diagram.
Drawing cashflow diagrams
(or cashflow timelines,
which you will see later in
the chapter) can be very
useful in visualising when
money is expected to be
paid!

Figure 2.1 Cashflow pattern of a property purchase, rental, and sale.

The term asset is very broad. You can think of assets as resources which an
individual, company or government owns and/or controls, expecting that they
will provide future benefits. There are many different types of assets, and many
more are being invented every year. We are going to look at assets in a lot of detail
in Module 4.
For now, we only briefly introduce two very common types of assets: fixed-
interest securities and equities.

2.1.1 Fixed-interest securities

Fixed-interest securities are also called “bonds”. Bonds are issued (this means
created) by companies or governments in order to borrow money to fund their
operations. So if you buy a bond, you are effectively giving a loan to the issuer.
In return, they pay you regular interest payments (called coupons) and at the
end of the period (known as the term), they will make a final payment (called the
redemption value). This final payment can be equal to the nominal value of the
bond (in which case the bond is said to be redeemed at par), but it is sometimes
set as higher or lower than the nominal value (in which case the bond is said to
be redeemed above or below par respectively.
The nominal value of the bond is the amount that all of the other payments
are calculated with respect to. So the coupon rate is applied to the nominal value
to calculate the regular coupon payment; the redemption amount is often 100%
of the nominal value but could be a different percentage as well.
2.1 Cashflow patterns of assets 19

It is important to understand that all of these cashflows are known at the

outset, so it is possible to predict the cashflows emerging from the bond with
close to complete accuracy.

Example 2.2

Say the South African government needs to raise some money to build some
hospitals, or prisons. Or prison hospitals! They decide to borrow money from the
public by issuing a bond. They decide (and the term and coupon are completely
up to them!) to issue R100 million (that is the nominal value) of a 20 year bond
with a coupon rate of 5%, which is to be redeemed at par. That means that whoever
buys the bond (let’s assume one person with far too much money on their hands
buys the whole R100 million issue) will get a coupon of R5 million (R100 million ×
5%) every year for 20 years, and at the end of 20 years, they will get R100 million.
The following cashflow diagram illustrates the payments that this person would
make and receive.

Figure 2.2 Cashflow pattern of a bond investment.

If the bond were to be redeemed at 110% of par, the final payment would be R110
million, plus a final R5 million coupon as before - R115 million altogether.

Note that in reality, coupon payments happen half-yearly. In this section, we

often assume that they occur annually. In the next example we consider a real
government bond issued by the South African government.

Example 2.3

A real example of a South African bond is the R2023 bond which was issued in 2012.
R500 000 000 nominal was issued on the 22nd of June 2012 and will be redeemed
at par on the 28th of February 2023. The bond has semi-annual coupons and a
coupon rate of 7.75% p.a.
U The details of this bond is-
The coupons are paid on the 31st of August and 28th of February each year until sue can be found on the
redemption. The issue price was R99.81 per R100 nominal. The cashflow diagram South African Treasury web-
site ([Link]).
20 Chapter 2 Value

below illustrates the cashflows you could expect if you invested in this bond on
22nd of June 2012 and bought R100 nominal of the bond. The coupon amounts
for R100 nominal are calculated as (7.75% × R100)/2 = R3.88 where we divide by
two as coupons are paid twice a year (semi-annually).

Figure 2.3 Cashflow pattern from R100 nominal investment in the

R2023 government bond.

Note also that not all bonds pay coupons. The simplest kind of bond is a zero
coupon bond. As the name suggests, the coupon is zero – no coupon payments
are made to the owner of the bond each year. The only payment is the repayment
T Note that zero coupon of the bond par value at the end of the term. The next example illustrates the
bonds are always redeemed payments from a zero coupon bond.
at par.

Example 2.4
Suppose that you invest in a zero coupon bond with a ten year term. You will
pay the price now and then in ten years’ time you will receive the par value. The
cashflow diagram below illustrates the payments you will make and receive.
2.1 Cashflow patterns of assets 21

Redemption
Amount

0 1 2 3 4 5 6 7 8 9 10

Price

Figure 2.4 Cashflow pattern from a 10 year zero coupon bond.

2.1.2 Equities
You have seen that companies can raise money by issuing bonds, i.e. basically U Equities are also called
getting a loan from investors. There is another way to get money: companies can shares or stock.
sell shares in the company to investors. This means that they are giving part of Bonds and equities are both
the ownership to the investors, and those investors now share in the profits of the referred to as securities.
company.

The difference between bonds and equities

Let’s say your uncle Elmo has a corner shop. Things are going well, and he thinks
he can expand the shop. But he will need money to rent some more space, put in
fittings, and buy more stock. He offers you a chance to invest in his business, in
one of two ways:

(a) Loan him R200 000. He will pay you R1 000 a month, and repay the R200 000
in full after 2 years.
(b) Buy into the business: his shop is valued at R1 000 000 now. He will sell you
20% of it in exchange for R200 000. You will become a part owner of the shop,
and every year, 20% of the profits will be yours - forever!

While uncle Elmo is just a small timer, and he is not issuing Bonds or Equities in the
strictest sense, his two options are very similar to issuing a bond (a) or an equity
(b). Can you see that each option will be very different for you as an investor? Can
you list of some of the differences you will experience?
22 Chapter 2 Value

Investors who buy shares own a portion of the company and are entitled to a
share in the net profits of the company.
T The dividend yield is the The payments made to the shareholders are called dividends. Dividends are
dividend per share divided related, but not equal, to the net profits the company makes; sometimes compa-
by the price per share. It is
nies may choose to keep part of their profits and use them in their operations.
often expressed as a per-
centage.
If a company does pay dividends, the amount of the dividend will depend on
the profits the company makes. This is unpredictable from year to year, and so
the dividend amount is uncertain.
The pattern of cashflows from a share generally looks something like this:

Figure 2.5 Example of a cashflow pattern from a dividend paying

share.
Now that you are familiar with some basic investments and the timing of their
cashflows, we can figure out how to put a value on these investments.

2.2 Interest
In this chapter we will learn to calculate the value of an amount of money at
different times by applying interest.
T Interest rates are sometimes Interest can be defined as the price paid by the borrower to the lender for the
stated in percentage form, use of an asset, usually called the capital or principal.
e.g. ‘the rate of interest per
To see how interest works, suppose that you lend R100 to a friend for a year
annum is 6%’, and some-
times in decimal form,
and they pay interest of 10% p.a. At the end of the year, you will receive R110
e.g. ‘the rate of interest is (R100+R10).
i = 0.06’. The phrase ‘per an-
num’, which is Latin for ‘per Exercises for 2.2 Interest
year’, is usually abbreviated
to p.a. Ex.2.1 How much would the repayment be if the loan was for 3 years?
2.2 Interest 23

2.2.1 Compound vs. Simple interest

The answer to Exercise 2.1 depends on whether you charged your friend simple
or compound interest.
Under simple interest, only the principal (i.e. the original amount) earns
interest, i.e. interest already earned does not itself earn interest. In our example,
the amount owed for a loan of R100 after 3 years at a simple interest rate of 10%
p.a. will simply be R100 × (1 + 0.10 × 3) = R130.00. In general, the accumulation of
RC invested for t years at simple rate i p.a. is:

C (1 + t i ).

Under compound interest, not just the principal, but the interest itself earns
interest. So after 1 year, the R100 will have grown to R110 as before. But after the
second year, the R110 will have grown to R110 × 1.10 = R121.00. And after 3 years,
that amount will have grown to R121.00 × 1.10 = R133.10.
In general, RC invested at compound interest of i p.a. accumulates as follows:
after one year it has accumulated to C + iC = C (1 + i ),
after two years to
C (1 + i ) + i ×C (1 + i ) = C (1 + i )2 ,
after three years to

C (1 + i )2 + i ×C (1 + i )2 = C (1 + i )3 ,

and after t years to

C (1 + i )t . (How would you show this more formally?)

The expression C (1 + i )t actually holds even if t is not an integer, although we

have not demonstrated that.
From the equations above the difference between simple and compound
interest should be clear.
U Think of simple interest like
taking the interest you are
Example 2.5 earning out and spending it
– it does not have a chance
To illustrate the difference between compound and simple interest rates consider
to stay in your account and
the following example. Suppose that you can invest R100 now for 20 years at either
carry on making more inter-
a compound interest rate of 6% p.a. or a simple interest rate of 6% p.a. What would
est for you!
the accumulated investment amounts in 2, 5, 10, and 20 years’ time be?
Figure 2.6 depicts the accumulated amounts over the 20 year term for each of these
investment options. We can see that the investment which offers a compound
interest rate of 6% p.a. has a greater value than that which offers a simple interest
rate of 6% p.a.

Simple interest rates are not common in practice and you can assume, unless
stated otherwise, interest is compound interest.
24 Chapter 2 Value

compound interest
simple interest

300
250
Amount (R)
200
150
100

0 5 10 15 20
Term

Figure 2.6 Accumulation of R100 at an interest of 6% p.a.

2.2.2 Nominal and effective interest rates

Another important distinction is between effective and nominal interest rates.
T Effective interest rates The effective rate is the compound annual interest that would have been
can also be calculated earned to get the investment to the position in which it is.
over other time periods -
monthly effective rates, for
example. Example 2.6
Earlier, we calculated that R100 increased with simple interest of 10% for 3 years
will amount to R130. What is the effective interest rate (i.e. the annual compound
rate) the R100 would have had to earn to grow to R130 after 3 years? Before you
make any calculations, do you think the effective rate will be lower or higher than
10% ?

Solution: We are looking for an annual effective rate i , such that

R100 × (1 + i )3 = R130

i = 1.31/3 − 1
i = 0.09139
So if R100 was invested at an annual compound interest of 9.139% for 3 years, this
would be equivalent to investing R100 at 10% simple for the same time period.

Effective rates are intuitively easy to understand and to compare to each

other. However, interest rates are in reality often stated as ‘nominal rates per
annum, convertible monthly’ (or convertible quarterly, or half-yearly, or p times
per annum, as the case may be).
2.2 Interest 25

This is purely a vocabulary (and sometimes a marketing!) issue – but in order

to be able to understand financial products out in the market, you need to be able
to interpret such statements.
U Can you see how nominal
rates may be good for mar-
Example 2.7 keting loans? Have a look
at these:
So for example, a nominal rate of interest of 18% per annum, convertible monthly,
18% p.a. convertible quar-
actually means 1.5% (= 18%/12) per month (effective). What effective rate p.a.
terly = (1 + 0.18/4)4 − 1 =
does this monthly effective rate imply?
19.25% annual effective;
18% p.a. convertible
Solution: The effective rate p.a. can be found by considering what an amount of monthly = (1 + 0.18/12)12 −
R1 would accumulate to over 12 months at 1.5% per month. Then we can figure 1 = 19.56% annual effective.
out the actual, “effective” rate that is being charged. In this case the p.a. effective You can see that any con-
rate is 19.56%, because 1 × 1.01512 − 1 = 0.1956. vertible rate is in actual fact
higher than the nominal
rate which is quoted.
More generally, a nominal rate of interest of j p.a., convertible p times p.a.,
just means an effective rate of j /p per 1/p of a year. What effective rate p.a. does
this imply? (1 + j /p)p − 1.
The conventional symbol for a nominal rate of interest p.a., convertible p
times p.a. (or ‘compounded’ p times p.a., or ‘payable’ p times p.a.), is i (p) . From
the above we see that the link between i (p) and i , the corresponding effective rate
p.a., is:
¶p
i (p)
µ
1+i = 1+ .
p
Equivalently:
i (p) = p (1 + i )1/p − 1 .
¡ ¢

The easiest way to handle nominal rates is to convert them to effective rates
and then use the effective rates to solve the problem.

Exercises for 2.2 Interest

Ex.2.2 Consider an investment account that earns a nominal rate of 12% p.a.
convertible monthly. Given an initial investment of R100, find the
accumulated value in the account after three months, and after three
years. Compare this to the same amount invested at 12% p.a. effective,
which one is higher?

Ex.2.3 Consider a bank that is marketing a loan with a nominal interest rate
of 6% p.a. convertible half-yearly.

1. Find the annual effective rate of interest that is being charged on

this loan.
2. Consider a loan of R1 000 that has a term of ten years and a single
repayment due in ten years’ time. Find the repayment value.
26 Chapter 2 Value

Ex.2.4 Given an interest rate of 10% p.a., find the accumulated value of a R100
investment after 54 months for the following types of interest rate:

1. a simple interest rate;

2. an annual effective rate;
3. and a nominal rate, convertible monthly.

In the coming sections we will see how we can answer these types of compari-
son questions for different types of cashflows.
T The term ‘to accumulate’ is
used when we are finding
the value of money at some 2.3 Single cashflows
future time point. The ac-
cumulated value of a series 2.3.1 Future values
of cashflows is also called
the future value of the cash- The future value of a cashflow or cashflows is simply the value of those cashflows
flows. at some specified future time. In order to determine the value of some cashflow
at a future time, we accumulate the cashflow with interest.
? What annual effective in- For example, R100 invested in the South African stock market in 1996 would
terest rate did the SA stock have accumulated to R875 by 2013 - R875 is the future value in 2013 of R100 in
market deliver over the pe- 1996.
riod from 1996 to 2013?
Exercises for 2.3.1 Future values

Ex.2.5 Consider R1 000 invested at 12% p.a. for 20 years. Find the accumulated
amount at times 2, 5, 10, and 20 years.

Ex.2.6 Suppose you need to pay an amount of R10 000 in three years’ time,
and money invested at any time over the next three years earns interest
at a rate of 15% p.a. How much should you invest now to repay the
R10 000 in three years’ time?

2.3.2 Present values

In exercise 2.6 above, you needed to invest R10 000 × 1.15−3 = R6575.16 now in
order to have R10 000 available in three years’ time. More generally, if you want
to have RX in n years’ time, and the interest rate is i p.a., the amount you must
invest now is:
X (1 + i )−n .

T The present value is also re- This is known as the present value of RX , due in n years’ time. The factor
ferred to as the discounted (1+i )−n above is referred to as the discount factor and for i , n > 0, 0 < (1+i )−n < 1.
value. The term ‘to dis-
So when we find the present value of RX paid in n years’ time by multiplying
count’ refers to the action
of finding the value of a
RX by (1 + i )−n , we are effectively saying that RX paid in n years’ time is worth
cashflow at an earlier date less than RX now; it has a discounted value now.
than it occurs.
2.3 Single cashflows 27

The present value can be thought of as the amount you would invest now to
get a specific future value. In Exercise 2.6, the present value is R10 000 × 1.15−3 .
The factor (1 + i )−1 is usually denoted by v, and sometimes by v i if it is nec-
essary to emphasise the rate being used. For example, using this notation in
3
Example 2.6, the present value could be written as R10 000 × v 15% .

Example 2.8
What is the present value of R129.69 received in 10 years time at an interest rate of
10% p.a.?

Solution:

Present value = 129.69 × (1 + 0.1)−10

= 129.69 × v 10
= 129.69 × 0.385543289
= R50.00.

Exercises for 2.3.2 Present values

Ex.2.7 Suppose that you need R10 000 000 to retire.

1. What amount of money would you have to invest now at an inter-

est rate of 15% p.a. to retire in 40 years’ time?
2. If the interest rate is lower, say, 10% p.a., how much would you
have to invest now to retire in 40 years’ time?
3. Considering your answers to (1) and (2), what relationship is there
between the interest rate and the amount you need to invest now?

4. If the interest rate is 15% p.a. and you want to retire in 25 years’
time, how much would you have to invest now?
5. Considering your answers to (1) and (4), what relationship is there
between the length of investment and the amount you need to
invest now?

Ex.2.8 A zero coupon bond has a term of 13 years. The redemption amount
is R100,000. Using an interest rate of 9.5% p.a., calculate the current
value of the bond.

Ex.2.9 Another zero coupon bond has a term of 2 years 8 months. The market
is pricing this bond using an interest rate of 8.17%. Calculate the price
of R100 nominal of this bond if it is redeemed at par.
28 Chapter 2 Value

2.3.3 Relationship between interest, term and present and future val-
ues
Exercise 2.7 illustrates a very important relationship between interest, term and
the value of a cashflow.

Example 2.9
Let’s say we are thinking of investing R1 000 for 5 years at 15% p.a. interest. We
would expect to get R2 011 at the end of the term. (Can you show that this is
correct?)
If we find another investment that offers us a higher rate of interest, say 17% p.a.,
we would expect the future value to be higher too (R2 192). What is the relationship
between interest and future values?

The higher the interest, the higher the future value, and the lower the interest,
the lower the future value.
The relationship between term and future value is equally straightforward:
the longer you invest money for, the greater the future value.

Example 2.10
In the previous example, if we change the term to 6 years, the future value increases
from R2 011 to R2 313.

Now let’s look at present value:

Example 2.11
The present value of R5 000 received in 10 years’ time at an interest rate of 12% p.a.
is R1 610 (check this).
What if we increase the rate of interest to 15% p.a.? Will the present value be lower
or higher?

Solution: At 15% p.a., the present value decreases from R1 610 to R1 236.

This should make sense, since if you can earn more interest, you can invest
less to get the same final value. So present values are inversely related to interest
rates: the higher the interest rate, the lower the present value, and vice versa.
And what about the relationship between term and present value?
2.3 Single cashflows 29

Exercises for 2.3.3 Relationship between interest, term and present and future
values

Ex.2.10 In the previous example, if we shorten the term from 10 to 8 years, does
the present value increase or decrease? Can you explain why this is?

In summary, present values are inversely related with both interest rate
and term. More clearly, the lower the interest rate and shorter the period of
investment, the more money you need to start with to have the same final amount.

2.3.4 Comparing single cashflows

At the start of the chapter you were asked which one of several cashflows you
preferred. We saw that in some cases it is easy to decide, e.g. when the amounts
were the same we preferred the earlier cashflows. In other cases it was difficult to
decide between the cashflows offered. It will usually not be easy to decide which
cashflows you prefer.
To compare two cashflows of different amounts, we need to find their values
at the same time point and then compare these values.

Example 2.12
Consider a choice between receiving R100 now or R120 in one year’s time. Which
would you prefer if the interest rate you can invest at is 10% p.a.?

Solution: There are many ways to answer this question. One method is to find the
present values of each of the cashflows and compare these. The two present values
(PVs) are:
PV1 = 100v 0 = 100 and PV2 = 120v 1 = 109.09.
Here we are comparing the values of the cashflows at time 0. The R120 received
in a year’s time has a value equivalent to R109.09 now. This is because we would
need to invest R109.09 now at an interest rate of 10% to have R120 in a year’s time.
By comparing these two present values, we can see that if we can earn interest at a
rate of 10% p.a., we would prefer to wait a year and receive the R120.
Similarly, we could compare the future values (FVs) of the cashflows, which are:

FV1 = 100(1 + 0.1)1 = 110 and FV2 = 120(1 + 0.1)0 = 120.

Here we are comparing the values of the cashflows at time 1. The R100 received
now is equivalent to R110 received in a year. Again we see that we would still prefer
to wait until the end of the year to receive the R120.

The overarching principle is that when we need to compare the values of

cashflows, we must compare their values at the same time point.
30 Chapter 2 Value

Exercises for 2.3.4 Comparing single cashflows

Ex.2.11 Suppose that you need to choose between two assets, each of which
requires the same initial investment. The first investment returns a
cashflow of R1500 in 5 years’ time and the second returns a cashflow
of R2 200 in 10 years’ time. Using an interest rate of 10% p.a., which of
these cashflows would you prefer?

Until now, we have only considered single cashflows (or a few cashflows):
R5000 in 10 years’ time, or R1000 now.
But in the real world, situations often arise which result in a number of cash-
flows – for example, an investment could pay R50 a year for 3 years and then
R1000 in year 4; or customers could owe you R100 this month, R1 000 next month,
nothing in month 3 and R500 in month 4. These kinds of cashflow combinations
still need to be valued.
It is often easier to work with series of cashflows if you can represent them
visually. You can do this using a cashflow diagram:

Figure 2.7 Cashflow diagram of the customers’ payments.

Drawing a cashflow diagram helps you ensure that you have accounted for
all cashflows, and it helps you spot patterns in cashflows which may make them
easier to value (more about that later).
Another way to represent cashflows is called a time line. It’s a quicker way to
indicate cashflows (and requires less drawing skill!). The timeline for the above
cashflows is represented in Figure 2.8.

payment 100 1000 0 500

time 1 2 3 4

Figure 2.8 Timeline of the customers’ payments.

2.4 Series of regular cashflows 31

You can use whichever type of illustration that you are most comfortable with.
We will alternate between the two in these notes to make sure you are familiar
with both.

Exercises for 2.3.4 Comparing single cashflows

Ex.2.12 Draw a timeline and cashflow diagram of the following cashflows:

R1 000 in 2 years’ time, R300 received in 5 years’ time and R2 500 re-
ceived in 3.5 years’ time. What is the present value of each of these
cashflows at 10% p.a.? What is the total present value?

Ex.2.13 Suppose you do some part time work and your employer offers two
different pay options. If you choose the first pay option, you will receive
R2 000 now, R3 000 in 6 months, and R2 000 in 8 months’ time. If you
choose the second pay option, you will receive R2 000 now, R600 in
6 months, R3 000 in 8 months, and R1 500 in 12 months’ time. (Hint:
Draw a timeline!)

1. Without calculating the present or future values, which pay option

do you think you would prefer?
2. Now at an interest rate of 10% p.a., which pay option would you
prefer?
3. Now at an interest rate of 18% p.a., which pay option would you
prefer?
4. What do you notice about your answers to 2. and 3.?

Exercise 2.13 demonstrates that when we use a higher interest rate to find
the present value, the cashflows which happen further into the future contribute
less to the total present value. In Exercise 2.13.2., we would choose the second
pay option and in Exercise 2.13.3., when a higher interest rate is used, we would
choose the first pay option. The reason our choice changes is the cashflows of the
first pay option happen earlier on average than the cashflows of the second pay
option.

2.4 Series of regular cashflows

So far we have looked at single cashflows or a few cashflows together. However, a
bond or an equity also have regular payments (coupons and dividends) which
need to be valued. It would be very tedious to value these individually (a 20 year
bond will pay out 40 separate coupon payments!) In this section we develop a
formula to help us value series of regular cashflows, called annuities.
An annuity is a series of payments made at fixed intervals while a certain
‘status’ continues to be valid, e.g. for a fixed number of years, or while a specified
person is alive. Annuities can have different characteristics:
32 Chapter 2 Value

T Condition: this refers to the ‘status’ that must be true for the annuity to
be paid. For example, an annuity that is payable for a certain fixed term is
called an annuity certain. Other annuities depend on different conditions:
an annuity payable while a person is alive is called a life annuity. For now,
we will focus on annuities certain; life annuities will be covered in Module
3.

T Increases: Annuities differ in how they increase in payment. The simplest

annuity, called a level annuity, does not increase at all – the payments
are the same every time. Annuities can also increase over time – we will
consider a simple case of an annuity that increases by a certain percentage
with every payment.

T Timing of payments: annuities are also distinguished by whether they are

paid at the start of each time period (in advance) or at the end of the period
(in arrear).
? See if you can name the These terms can be used in combination to describe any series of regular
following: payments. For example, if you receive R100 000 at the end of each year for 10
years, that would be a level annuity certain payable annually in arrear for 10 years.
• R10 a year paid at the
start of each of 5 years,
increasing at 5% per Example 2.13
year.
An example of a level annuity ‘in arrear’ would be the payments you make to repay
• R50 000 paid on the a loan you take out to pay for university. For example, you may have to pay R35 804
1st of January of each at the end of each year for ten years to repay a loan of R220 000. More explicitly,
year as long as George
is alive. At the end of year one you have to pay R35 804,
• The coupons of a 10 At the end of year two you have to pay R35 804,
year bond which pays ..
R10 p.a. .
At the end of year ten you have to pay R35 804.
• The dividends of an
equity which grow by
5% per annum and
are paid annually. The
first one is R2.69.

Figure 2.9 Cashflow diagram of your student loan.

2.4 Series of regular cashflows 33

At an interest rate of 10% p.a., the present value of this series of cashflows is:

35 804(v + v 2 + . . . + v 9 + v 10 ) = 35 804(1.1−1 + 1.1−2 + . . . + 1.1−10 )

= 35 804(0.9091 + 0.8264 + . . . + 0.3855)
= 220 000.

The repayments for your loan can be called a level annuity certain of R35 804
payable annually for 10 years, in arrear.

2.4.1 Present values of level annuities certain

How do we calculate the present value of a level annuity of R1 p.a., payable
annually in arrear for n years?

Present value of a level annuity certain, payable annually in arrear:

It is just
n
vk = v + v2 + ... + vn.
X
k=1
It can be tedious to calculate the present value of this annuity by calculating each
term in the sum and adding them all up. Luckily there is a neat formula for this
sum which we derive now. Suppose that the value of the sum equals Z . That is
n
vk.
X
Z= (2.1)
k=1

Now multiply both sides of Equation (2.1) by (1 + i ) to give

(1 + i )Z = 1 + v + . . . + v n−1 . (2.2)

Then subtract Equation (2.1) from (2.2), to give

(1 + i )Z − Z = (1 + v + . . . + v n−1 ) − (v + v 2 + . . . + v n )
= 1 − vn.

Rearranging the above equation we can find

n 1 − vn
vk =
X
Z= .
k=1 i

So (1 − v n )/i is the formula for the present value of a level annuity certain of n
payments paid in arrear. We will use this formula so much that there is a special
symbol allocated to it: a n . So a n = (1 − v n )/i .
What about an annuity paid annually in advance? How do we calculate the ! Useful formula:
present value of a level annuity of R1 p.a., payable annually in advance for n years? 1 − vn
It is just an =
n−1
i
X k
v = 1 + v + . . . + v n−1 .
k=0
34 Chapter 2 Value

This can be shown (using a similar trick to above) to be

? See if you can show this.
n−1 1 − vn
vk =
X
.
k=0 1−v

We use the symbol ä n to represent the present value of a level annuity certain of n
payments paid in advance, i.e. ä n = (1 − v n )/(1 − v).
We will often come across 1 − v, and so it also has a shortcut symbol attached
to it:
1 1+i −1 i
d ≡ 1−v = 1− = = .
1+i 1+i 1+i
d is referred to as a discount rate.
! Useful formula: Using d , we can express the formula for an annuity payable in advance as
1 − vn 1 − vn
ä n = . ä n = .
d d

Exercises for 2.4.1 Present values of level annuities certain

Ex.2.14 Try using one of these formulae to find the present value of the repay-
ments in loan example 2.13.

Ex.2.15 1. Calculate the present value of each of the following, at 5% p.a.

interest: R1 000 in 1 year; R1 000 in 2 years; and R1 000 in 3 years.
Sum their values to get a total.
2. Calculate the present value, at 5% p.a., of a level annuity certain
of R1 000 p.a. payable in arrear for 3 years, using the formula we
have derived above. Did you get the same value for (1) and (2)?

You can also work with annuities priced at nominal interest rates, by convert-
ing the nominal interest rates into effective rates which match the interval of the
annuity payments:

Example 2.14
As an example, let us find the present value of R1 p.a. payable annually in arrear
for 20 years using a nominal interest rate of 12% p.a., convertible monthly.

Solution: To answer this question it is easiest to work with an annual effective rate.
The interest rate in the question is quoted as a nominal rate, convertible monthly
and is not an annual effective rate. We have to convert it to an annual effective rate.
The annual effective rate is

12% 12
µ ¶
1+ − 1 = 12.68%.
12

As we saw before, the annual effective rate is greater than the nominal rate. Why
do you think this is the case?
2.4 Series of regular cashflows 35

Now that we have the annual effective rate it is easy to answer the question. The
present value of the annuity is

PV = a 20 12.68% = R7.16.

Example 2.15

Now suppose you want to find the present value of an annuity which pays R1 per
month paid monthly in arrear for 20 years, using a nominal interest rate of 18%
p.a., convertible half-yearly.

Solution: For this example a monthly effective rate would be easiest to work with
because the payments are monthly. How do we get a monthly effective rate? One
method is to find the annual effective rate and then convert this to a monthly
effective rate. In this case the annual effective rate is

18% 2
µ ¶
1+ − 1 = 18.81%,
2

and the monthly effective rate is

1
(1 + 18.81%) 12 − 1 = 1.45%.

Now that we have the monthly effective rate we can find the present value as

PV = a 240 1.45% = R66.92.

There are a total of 240 payments as there are 12 payments per year for 20 years.

2.4.2 Accumulations of level annuities certain

How do we calculate the accumulated value at time n of a level annuity of R1 p.a.,
payable annually in arrear for n years?

Accumulations of level annuities certain, payable annually in arrear:

It is just
n−1
(1 + i )k = (1 + i )n−1 + (1 + i )n−2 + . . . + 1.
X
k=0

A formula for this sum can also be derived. Let

n−1
(1 + i )k ,
X
W=
k=0

and multiply W by (1 + i ) to give

(1 + i )W = (1 + i )n + (1 + i )n−1 + . . . + (1 + i ).
36 Chapter 2 Value

Now

(1 + i )n + (1 + i )n−1 + . . . + (1 + i )
¡ ¢
(1 + i )W − W =
− (1 + i )n−1 + (1 + i )n−2 + . . . + 1
¡ ¢

= (1 + i )n − 1.

Rearranging the above equation gives the result

n−1 (1 + i )n − 1
(1 + i )k =
X
.
k=0 i

We attach the symbol s n to this future value, i.e. s n = ((1 + i )n − 1)/i .

! Useful formula: It is also possible to find the accumulated value at time n of a level annuity of
(1 + i )n − 1
R1 p.a., payable annually in arrear for n years, by finding the present value and
sn = . then accumulating the present value to time n. More explicitly, we can find the
i
accumulated value as follows,

1 − vn
a n × (1 + i )n = × (1 + i )n
i
(1 + i )n − 1
=
i
= sn

The accumulated value of an annuity in advance can be found in a similar

way to the accumulated value of an annuity in arrear. The accumulated value at
time n of a level annuity of R1 p.a., payable annually in advance for n years can
be found as
n
(1 + i )k = (1 + i )n + (1 + i )n−1 + . . . + (1 + i ).
X
k=1

This sum can be simplified to

? See if you can show this. n (1 + i )n − 1
(1 + i )k =
X
.
k=1 d

We attach the symbol s̈ n to this future value, i.e. s̈ n = ((1 + i )n − 1)/d .

! Useful formula: For all of these annuity values, when it is necessary for clarity to identify the
(1 + i )n − 1
interest rate we can write s̈ n i , for example. So the accumulated value of R1 p.a. in
s̈ n = . advance for 6 years at 15% p.a. interest is s̈ 6 15% .
d

2.4.3 Pricing bonds

Now that we know how to calculate the present value of a single cashflow and
a series of cashflows, we can easily use these techniques to value bonds. Recall
that a bond is essentially a series of regular coupons, plus a single redemption
payment.
2.4 Series of regular cashflows 37

Example 2.16
Suppose we want to buy R100 nominal of a bond that has ten years to redemption.
This bond pays annual coupons at a coupon rate of 8% p.a., with the next coupon
due in one years’ time. How much, at an interest rate of 12.5% p.a., would you be
willing to pay now for R100 nominal of this bond?

Solution:
The price we would be willing to pay is the present value of the payments the bond
will make.
For R100 nominal, the annual coupon payments will be R8 (8% of R100 nominal)
and the redemption amount will be R100.
The price of the bond will be

Price = (R100 × 8%) × (v + v 2 + . . . + v 10 ) + R100v 10

= R8a 10 + R100v 10
= R75.09.

At an interest rate of 12.5% we would be willing to pay R75.09 per R100 nominal of
this bond. ? If that is indeed the market
price of the bond, does that
mean it is trading at a pre-
mium or at a discount to
Exercises for 2.1 Cashflow patterns of assets par?

Ex.2.16 Consider a bond which will be redeemed at par in ten years time. The
bond pays semi-annual coupons at a coupon rate of 15% p.a. and the U That means the bond pays
R7.50 per R100 nominal ev-
next coupon is due in six months’ time. At an interest rate of 5% per
ery 6 months!
half-year, what would you be willing to pay for R100 nominal of this
bond now?

Ex.2.17 Consider a bond with 5 years to redemption (which is at par). The bond
pays annual coupons at a coupon rate of 10% p.a. and the next coupon
is due in one year’s time.

1. At an interest rate of 8% p.a., what would you be willing to pay for

R100 nominal of this bond now?
2. At an interest rate of 10% p.a., what would you be willing to pay
for R100 nominal of this bond now?
3. At an interest rate of 12% p.a., what would you be willing to pay
for R100 nominal of this bond now?
4. Comparing the coupon rate and the interest rates used to value
the bond in (1)-(3), what can you conclude?
38 Chapter 2 Value

2.4.4 Perpetuities
So far, all of the annuities we have considered have had a specified term, or some
other “stopping” condition (such as a death). But there is no reason why an
annuity should not be payable forever – such annuities are sometimes referred to
as perpetuities.
Can a perpetuity be valued? It seems strange that we would be able to put a
value on an infinite series of payments. Surely the present value of that is infinite,
too?
Again, the power of time value of money comes to the rescue. The further a
payment is in the future, the less its present value. For example, at 12% p.a.:

PV of R100 today = R100

PV of R100 in 1 year = R89
PV of R100 in 10 years = R32
PV of R100 in 100 years = R0.0012

You can see that the later parts of the perpetuity are not worth very much in the
present!
We can derive a formula for the present value of a perpetuity from our normal
annuity formula.
? See if you can derive a simi-
lar formula for a perpetuity
of R1 p.a., paid in advance, For a perpetuity of R1 p.a., paid in arrear, at an interest rate of i p.a., the
using an interest rate of i present value is:
p.a.

∞
vk
X
= lim a n
n→∞
k=1
1 − vn
= lim
n→∞ i
1 vn
= − lim
i n→∞ i
1
= . As lim v n = 0
i n→∞

2.4.5 Increasing annuities certain

So far we have only considered level annuities, where the payments are all equal.
But cashflows that increase over time in a consistent way can also be valued using
formulae.
Annuities could increase in at least two ways:

T Increase by the same percentage every year.

2.4 Series of regular cashflows 39

Example 2.17
For example, R100 in year 1, R120 in year 2, R144 in year 3, and so on. These
payments can be represented in the following cashflow diagram.

Figure 2.10 Cashflow diagram of a geometrically increasing annuity.

This annuity increases at 20% p.a. Note the shape of the curve – it is concave,
meaning it increases more each year in later years than in earlier years. This
type of increasing annuity is called a geometrically increasing annuity.

T Increase by the same amount every year.

Example 2.18
For example, R100 in year 1, R120 in year 2, R140 in year 3, and so on.

Figure 2.11 Cashflow diagram of an arithmetically increasing annuity.

This annuity, called an arithmetically increasing annuity, or just an “in-
creasing annuity”, increases by R20 every year. Note how much less this
type of increase amounts to after 10 years; a geometrically increasing annu-
ity pays R619 in the tenth year, while the arithmetically increasing annuity
pays only R280 at this time.
40 Chapter 2 Value

? For each of the annual se- Both of these types of increasing annuities can be valued using a formula. The
ries of cashflows below, de- formula for arithmetically increasing annuities is fairly complex to derive, and we
cide whether it is an annu- will only look at this next year.
ity, and what type. The formula for a geometrically increasing annuity is however very useful and
(For example: the payments we can derive it as follows.
of 2, 4, 6, 8, 10, are the pay-
ments you would receive
from an arithmetically in- Formula for a geometrically increasing annuity, payable annually in ar-
creasing annuity with a rear:
term of 5 years.)
Suppose that we want to find the present value at a rate of interest i p.a. of a series
(a) 1000, 1001, 1002, 1003, of cashflows 1, (1 + j ), (1 + j )2 , . . . , (1 + j )n−1 paid in arrear at times 1, 2, 3, . . . , n. The
1004. present value is the value of the sum
(b) 200, 300, 450. n
(1 + j )k−1 v ik = v i + (1 + j )v i2 + (1 + j )2 v i3 + . . . + (1 + j )n−1 v in .
X
(c) 20, 20, 20, 30. k=1

Now this looks rather complicated but with some manipulation we can change it
into something we have seen before. Working with the right-hand-side, we can
multiply by (1 + j )/(1 + j ) to give
n 1 ¡
(1 + j )k−1 v ik = (1 + j )v i + ((1 + j )v i )2 + ((1 + j )v i )3 + . . . + ((1 + j )v i )n .
X ¢
k=1 (1 + j )

If we set v l = (1 + j )v i in the above, i.e. (1 + l )−1 = (1 + j )/(1 + i ), then we can write

n 1 ¡
(1 + j )k−1 v ik v l + v l2 + v l3 + . . . + v ln
X ¢
=
k=1 (1 + j )
1
= an l ,
(1 + j )

and which we can calculate using the formula for annuities certain in arrear.
Notice that the interest rate used for the annuity is l p.a. The value of l satisfies the
equation
1+i
1+l = .
1+ j

This should make intuitive sense: think about a single cashflow which in-
? Can you see the parallels be-
tween this present value creases over time. Say you are offered a single payment in n years’ time. The
and the calculation of l payment is R100 increased by j per year, until it is paid at time n. So the payment
above? is R100 × (1 + j )n . What is the present value of this payment at interest rate i p.a.?

PV = Payment × (1 + i )−n
= R100 × (1 + j )n × (1 + i )−n
1 + i −n
µ ¶
= R100 × .
1+ j

Returning to series of regular cashflows, suppose now that the cashflows are
in advance. How would you find the present value at a rate of interest i p.a. of
2.4 Series of regular cashflows 41

a series of cashflows 1, (1 + j ), (1 + j )2 , . . . , (1 + j )n−1 paid at times 0, 1, 2, . . . , n − 1?

The present value can be found by finding the value of the sum,

n−1
(1 + j )k v ik = 1 + (1 + j )v i + (1 + j )2 v i2 + . . . + (1 + j )n−1 v in−1 .
X
k=0

This sum can be simplified to

n−1
? Can you show this?
j )k v ik
X
(1 + = ä n l ,
k=0

where
1+i
l= − 1.
1+ j
Finally, note that if you need to calculate the future value of an increasing
annuity, you can do so by calculating the present value of the annuity at

1+i
l= −1
1+ j

and then accumulating that present value to time n at i .

Exercises for 2.4.5 Increasing annuities certain

Ex.2.18 From first principles, show that the formula for calculating the future
value, at an interest rate of i , of an annuity payable for n years annually
in advance, where the first payment is R1 and subsequent payments
increase by j p.a., is ä n l × (1 + i )n , where

1+i
l= −1
1+ j

Ex.2.19 Suppose that you want to buy an annuity certain of R15 000 p.a. payable
annually in arrear for 7 years and increasing at 5% p.a. The increases
start at the start of the second year. The rate of interest is 9% p.a.

1. What are i , j , and l in this example?

2. Calculate the present value of this annuity.
3. Calculate the accumulated value of the annuity at the end of the
7 year period.

2.4.6 Pricing equities

We can now return to equities and attempt to put a value on an equity. Recall the
cashflow diagram for an equity we developed in section 2.1.2:
42 Chapter 2 Value

Figure 2.12 Example of a cashflow pattern from a dividend paying share.

The arrow at the end shows that dividends are expected to continue indefi-
nitely – equity, unlike a bond, the redemption date is unknown and we treat it as
if it has no redemption date. Once you own the share, you can expect to receive
dividends for as long as the company is in operation (unless you sell the share to
someone else, of course).
You can see in the graph above that dividends are expected to vary from year
to year. But they are also generally expected to increase over time – at least on
average. The curved line indicates the overall dividend growth that this company
is expected to experience.
If we wanted to value a share, we would not be able to predict the fluctuations
in dividends from year to year. But it is possible to estimate the overall average
growth in the share .
U That is in fact one of the So instead of valuing the real, unpredictable cashflows that this share will
things that quants do when generate, we value estimated, smooth cashflows:
they work as investment
analysts.

Figure 2.13 Example of a smoothed cashflow pattern from a dividend

paying share.
2.4 Series of regular cashflows 43

So what we have here is a series of cashflows that is payable in perpetuity,

and that increases at a rate of g p.a. We assume that the cashflows increase at
a constant rate as this simplifies the mathematics. This series of cashflows is
therefore a geometrically increasing perpetuity.

Valuing a dividend stream with constant increases:

Suppose a company has just paid a dividend of RD and the company pays annual
dividends which increase at a rate of g p.a. The present value, at an interest rate of
i p.a., of the dividend stream is:

PV = D(1 + g )v i + D(1 + g )2 v i2 + D(1 + g )3 v i3 + . . .

= D v l + D v l2 + D v l3 + . . . (v l = (1 + g )v i )
= D lim a n l
n→∞
D
= .
l

where
1+i
l= − 1.
1+g

The present value of the dividends is the price that you are willing to pay for
the share – your evaluation of the worth that you can get from this investment.

Example 2.19 Suppose you are considering buying a share of Melon Com-
puters (an Apple competitor). Melon is trading at $980 per share, and the
next dividend, payable 1 year from now, is expected to be $60. Melon
computers pays annual dividends and you think that the company will
experience dividend growth of 9% p.a. Using 15% p.a. as the interest rate,
how much do you value the share at? And how does that compare to the
market price – should you buy the share or not?

Solution:

PV = 60v 15% + 60(1 + 0.09)1 v 15% 2

+ 60(1 + 0.09)2 v 15%
3
+...
1 ¡
60(1.09)v 15% + 60(1.09)2 v 15% 2
+ 60(1.09)3 v 15%
3
¢
= +...
1.09
60 ¡
1.09v 15% + 1.092 v 15% 2
+ (1.09)3 v 15%
3
¢
= +...
1.09
60 ¡
v l + v l2 + v l3 + . . .
¢
=
1.09
60
= lim a n l
1.09 n→∞
60 1
= × (l = (1.15)/(1.09) − 1 = 0.0550)
1.09 l
= $1 000.
44 Chapter 2 Value

So you value the share at $1 000 per share, and the market is willing to sell
it to you at $980 per share. Should you buy? Yes, it’s a good deal (if your
valuation is correct)!

Note that in the last example, we were not working with the dividend D paid
just before the valuation, but with the dividend expected in one year’s time (let’s
call that D 1 ). More generally:

Valuing an equity if D 1 is known

If D 1 , the dividend payable one year from now, is given, and the dividends are
expected to grow at g p.a. and are valued at i p.a., we can value the share as:
T Rather than memorise differ-
ent formulas, always draw a
cashflow diagram and work
PV = D 1 (1 + i )−1 + D 1 (1 + g )(1 + i )−2 + D 1 (1 + g )2 (1 + i )−3 + . . .
from first principles!
D1 1+g 1 1+g 2 1+g 3
µµ ¶ µ ¶ µ ¶ ¶
= + + +...
1+g 1+i 1+i 1+i
D1 D1 1
= lim a n l = ×
1 + g n→∞ 1+g l

Note that
1+i 1+i −1−g i −g
l = −1 = =
1+g 1+g 1+g

D1
So then PV =
(1 + g ) × l
D1
=
(1 + g )(i − g )/(1 + g )
D1
=
i −g

Most of the time, equities are valued as if you were planning to hold them in
perpetuity. But if you have a specific plan to sell an equity after a number of years,
you could value the dividends which you will receive from now until the sale date,
and then estimate the sale price.

Example 2.20
Consider a share that will pay dividends at the end of each year with a dividend of
R5 due in one year’s time. The dividends are expected to increase by 3% p.a. (after
the dividend due in a year) and the share price is expected to be R113 in ten years’
time. You want to buy the share now. How much would you pay for the share using
an interest rate of 16% p.a.?
2.4 Series of regular cashflows 45

Solution:
2
Price = 5(v + (1.03)v 16% + . . . + (1.03)9 v 16%
10 10
) + 113v 16%
= 5(1.03)−1 (1.03v 16% + 1.032 v 16%
2
+ . . . + 1.0310 v 16%
10 10
) + 113v 16%
10
= 5(1.03)−1 a 10 l + 113v 16% .

To find the value of the share we need to find the value of l . It is equal to
1.16
l= − 1 = 0.1262.
1.03
The value of share is then

S 0 = 5(1.03)−1 a 10 12.62% + 113v 10 = R52.36.

Exercises for 2.1 Cashflow patterns of assets

Ex.2.20 You are holding a share which is expected to pay a dividend of R12 in
one year’s time. Dividends are expected to grow at 5% p.a. At an interest
rate of 9% p.a., what is the value of the share? The market price of the
share is currently R350. Is the share over- or under-priced according to
your valuation?

2.4.7 Deferred series of cashflows

The term deferred is used to refer to payments which do not start immediately
but rather start some time in the future. The length of time from now until the
cashflows start is referred to as the deferred period.
Remember that so far, all of our formulas calculate the present value of an
annuity right at the start of the annuity: so for an annuity paid in advance, the
present value is calculated as at the date of the first payment. For an annuity
payable in arrear, the formula for present value is based on the time period one
year before the first payment.
But we should be able to calculate the value of an annuity even if it doesn’t
start immediately, i.e. if it is deferred for some time.
A deferred annuity is an annuity which starts at some point in the future. For
example, R1 payable annually in advance for 10 years starting exactly 7 years from T Or it could be a level annu-
now is a level annuity certain with a term of 10 years, payable in advance, deferred ity certain with a term of
10 years, payable in arrear,
for 7 years.
deferred by 6 years - can you
see why?
Example 2.21
An example of a deferred series of cashflows would be the cashflows from an
annuity of R1 p.a. paid annually in arrear for n years, deferred for m years. This
is a deferred annuity, payable in arrear, and would make payments of R1 at times
(m + 1), (m + 2), . . . , (m + n). The following diagram illustrates when the payments
occur.
46 Chapter 2 Value

payment 1 1 ... 1
... m ... m +n
time 0 1 m +1 m +2

The present value of the cashflows from this deferred annuity is

PV = v m+1 + v m+2 + . . . + v m+n−1 + v m+n

= v m (v 1 + v 2 + . . . + v n−1 + v n ).

The last equation shows that the present value is equal to v m a n . This is equivalent
to finding the value at time m of the an annuity of R1 p.a. paid annually in arrear
for n years, and then discounting that by v m . It is worth noting that m, the length
of the deferred period, does not have to be an integer.

Exercises for 2.4 Series of regular cashflows

Ex.2.21 Show that the present value of an annuity R1 p.a. paid annually in
arrear for n years but deferred for m years is equal to

PV = a m+n − a m .

Give an intuitive explanation for this result.

Ex.2.22 You are considering buying a share in a new company. From your
research you conclude that the company is only going to start paying
dividends in five years’ time. The dividends will start at R12 and will
increase by 10% p.a. for 5 years after the first dividend. After this period
the dividends will remain constant. You expect to sell the share for
R150 in fifteen years’ time. At an interest rate of 13.5% p.a., how much
are you willing to pay for the share now?

Ex.2.23 Suppose that you have a child and you want to buy a financial product
which pays you regular amounts at the beginning of each year for
the next 10 years. These amounts are meant to pay for some of your
child’s school fees. How much would you be willing to pay now for
this financial product, at 12% p.a. interest, if the product pays out the
following amounts?

1. R5 000 p.a.
2. R5 000 p.a. for the first 2 years and R6 000 for the next 8 years.
3. R5 000 p.a. which increases by 5% at the end of each year (the first
increase is applied to the second payment).

Ex.2.24 You have just won the lottery and the lottery board has offered you two
alternative ways of claiming your prize. The first option they offer you
will pay R10 000 000 now, but you will incur 40% tax now and so you
will receive R6 000 000 (= 10 000 000 × (1 − tax rate)) now. The second
2.5 Equations of value 47

option will pay you R1 000 000 at the end of each of the next 15 years,
which increases by 9% at the end of each year. If you take the second
option you will pay 35% tax on each of the payments at the time they
are received. At an interest rate of 13.5% p.a., which option would you
prefer? (Hint: draw a timeline for the second option and do not forget
to deduct the tax payments.)

2.5 Equations of value

An equation of value is simply the equating of two present values or two accumu- T An equation of value can
lated values. We have been doing this all along when calculating present values. look different depending
In the last section, we used an equation of value to find out what the annual on the cashflows involved.
repayment on a loan of R100 over 12 years will be: For example, for an annu-
ity payable in arrears, the
R100 = X × a 12 10% . following is an equation of
value:
Equations of value are useful because they allow us to calculate any missing
component of the equation by solving for the unknown element. Another example PV = C × a n i .
of an equation of value is if you have to calculate how much you should invest
now at an interest rate of 15% p.a. to have R10 000 in three years’ time. If you Note that the unknowns in
this equation could be PV,
invest RX now, the equation of value in this case is X (1 + 0.15)3 = 10 000. We have
C , n and i . If all but one of
equated the two accumulated values and we can use this equation to solve for RX . these variables are known,
Equations of value can be used to solve for other variables and this is the topic of then you can find the miss-
this section. ing value!

2.5.1 Solving for the yield (i )

Sometimes we are faced with a situation where we want to figure out the interest
rate, because it is not given.

Example 2.22
You have been offered an opportunity to invest R100 000. The promised payout
to you after 3 years is R300 000. Can you use an equation of value to calculate the
return you are being offered?

Solution: You can construct the equation of value for this transaction, with the
investment amount on the left-hand side, and the present value of the payout,
discounted for 3 years at an effective rate of interest i , on the right-hand side:
100 000 = 300 000v i3
= 300 000 × (1 + i )−3
We can now solve for i:

1
100 000 − 3
−=i
300 000
i = 44.225%
48 Chapter 2 Value

That’s almost too good to be true! Better check this out more carefully before
committing...

It’s easy to solve for i when there is only one future cashflow. But what about
a series of cashflows?
For example, let’s consider an investor who has the opportunity to pay R1 000
to receive R120 p.a., annually in arrear for 8 years, plus R1 000 at the end of year
8. In this case, it is clear that the investor gets 12% on his or her investment of
R1 000, i.e. that the yield is 12% p.a.

Exercises for 2.5 Equations of value

Ex.2.25 To see this, check for yourself that i = 0.12 is the solution of

1 − (1 + i )−8
1 000 = 120a 8 i +1 000(1 +i )−8 = 120 +1 000(1 +i )−8 . (2.3)
i

T In the case of Now consider a second investment opportunity, also available at a cost of
R1 000: R90 p.a., annually in arrear for 8 years, plus R1 250 at the end of year 8.
PV = C × a n i , In this second case, it is not so clear what the yield is. (But we can see the yield
is greater than 9%. How?)
the PV is the price you are
prepared to pay for a series By analogy with Equation (2.3) we define the yield on any transaction as
of cashflows of C for n years follows. The rate of interest i such that
at interest rate i .
the p.v. at rate i of the outlays = the p.v. at rate i of the receipts
So the PV is an outlay: what
you are prepared to pay.
is called the yield or internal rate of return on a financial transaction. (There are
And the C s are receipts:
awkward examples in which such an equation has no roots i , or more than one,
what you are getting in re-
but you can worry about those next year!)
turn.
What then is the yield to the investor in this case? The rate i such that

1 − (1 + i )−8
1 000 = 90a 8 i + 1 250(1 + i )−8 = 90 + 1 250(1 + i )−8 .
i
We can rearrange this as follows:

0 = 90a 8 i + 1 250v 8 − 1 000.

The problem is that there is no explicit way to solve this equation. It has a
solution - there is a value for i where the equation is true - but it is not possible to
find that solution by manipulating the equation.
The mathematical approach to such equations is to try different solutions
until you find one that fits. You could do this at random (but that would take
forever!), or you can use the information you get from each guess to make a better,
new guess.
2.5 Equations of value 49

Let’s try this in the above equation. We know that interest rates should be a
small number like 5% or 15%. Let’s try, as a first guess, i = 0.12: At i = 12%:
90a 8 i + 1 250(1 + i )−8 − 1 000 = −R48.06.
This is a little too low. What would be a better guess? Should we choose an
interest rate lower or higher than 12%?
A smaller i makes for a higher present value, and if our present value is smaller,
the result will be higher than -R48.06. (Can you see why?)
So we can try 11%: then the result of the calculation goes up to R5.56. This
tells us that we have overshot, and that the yield is between 11% and 12%. We
could try 11.5% next (or maybe a number closer to 11% than 12%, since the result
of 11% was closer to 0, which is what we want). So we could try 11.25%, and so on.
There is a formal mathematical technique, called linear interpolation, which
helps to find such answers with the minimum number of tries. You may learn this
technique in another course. However, it is still laborious, and nowadays such
answers are usually found by making a computer program which makes many
guesses very quickly.

Example 2.23
If R180 p.a., paid annually in arrear for 8 years, with no further payment at the end
of year 8, is valued at R1000, what is the interest rate it is valued at? Start with 8%
as your first guess.

Solution: The present value of f (i ) = 180a 8 i −1000 = 34.40 at 8%. We must choose
a higher i for our second guess, and at 9% the present value is -3.73. Hence the
answer is between 8% and 9%, but much closer to 9% than to 8%. Make another
few guesses until you get close to the right solution!

Exercises for 2.5 Equations of value

Ex.2.26 For example 2.23, show that 8.90% is a reasonable approximation of

the solution. What is the present value at i = 8.90%?

Ex.2.27 A fixed interest bond pays annual coupons, the next of which is due
in 3 months time. The term of the bond is 6 years and 3 months, the
coupon rate is 5%, the redemption is at 80% of par and the current
price of the bond is R69 per R100 nominal.

Write down an equation of value for this bond.

2.5.2 Solving for the number of payments or the term

Another use of the equation of value is the determination of the number of
payments or the term over which payments occur. Again, we use an example to
introduce the ideas.
50 Chapter 2 Value

Example 2.24
Suppose that you have R82 664.32 and want to buy an annuity that will pay you
R10 000 annually, in arrear, for as many years as possible. Let’s call the number
of years n, and assume that an interest rate of 11% p.a. is used to calculate the
annuity price. How many annual payments can you afford, i.e. what is the value of
n?

Solution:
The starting point is to write down the equation of value, which is

R82 664.32 = R10 000a n 11% .

Expanding the a n 11% term on the right-hand-side we get

1 − (1 + 0.11)−n
R82 664.32 = R10 000 .
0.11
Rearranging this equation to have (1 + 0.11)−n on the left-hand-side, we have

R10 000 − 0.11 × R82 664.32

(1 + 0.11)−n = .
R10 000
Taking logarithms of both sides of the equation we can solve for n,
¡ R10 000−0.11×R82 664.32 ¢
log R10 000
n = −
log 1.11
= 23.

So with the amount of money you have, you can buy 23 annual payments of
R10 000!

So far we have demonstrated how you can solve for the yield and the term.
Equations of value can be used to solve for other unknown values, such as the
values of payments, rates of increase, amongst other unknowns.

Exercises for 2.5 Equations of value

Ex.2.28 If you were to invest R13 866.39 at the end of each year and these
amounts earn interest at a rate of 7% p.a., how long will it take you to
save up R120 000 to buy a new car?

2.6 Risk discount rates

So far, we have assumed that when we look at future cashflows, we only need to
work out their value, and that there is no uncertainty as to whether the cashflow
will actually happen or not. But that is not very realistic – some cashflows are
definitely less likely than others. For example, when we discussed bonds, we said
that bonds can be issued by governments or companies. If you were offered two
2.6 Risk discount rates 51

bonds of the same term and coupon, one issued by the government of the UK
and the other by a small company, which would you expect to cost more?
A government bond will generally have a higher price than a bond issued by a
company (called a corporate bond). In order to understand why, we need to go
back to the concept of risk.
Recall that when it comes to assessing financial outcomes, actuaries and
quants take into account risk.
We have defined the components of risk as value and uncertainty.
So far, in our calculation of the present value of cashflows, we have only taken
into account value, but not uncertainty. We have basically assumed that there is
no risk of the other party not paying the promised cashflows over to us. But is this
always a realistic assumption? A government may be a very safe issuer, where it’s
reasonable to say there is no risk of the promised payments not being made. But
a company is a different story – a company could go under, and stop paying out
what it owes investors.
Let’s consider a 5 year zero coupon bond issued by a government, and let’s
say that the market is pricing it using an 8% p.a. interest rate. What is the market
price of R100 of this bond?

Market Price = R100 × 1.08−5 = R68.

If we want to allow for uncertainty of the issuer actually paying out in this equa-
tion, we can say there is a 100% chance of this issuer paying the R100 in 5 years
time as promised. So the present value, allowing for uncertainty, can be pre-
sented as:

Market Price = Value × Uncertainty = R100 × 1.08−5 × 100% = R68.

Now let’s look at another 5 year zero coupon bond, this time issued by a company.
What is the probability of this company going bankrupt during the 5 year term?
We don’t think it’s 0%: so our investment is not 100% safe. We need to make an
estimate of the uncertainty. Let’s say we do some investigations and estimate that
every year the company operates, there is a 1% chance it may go out of business.
OK, then what is the chance that it is still in business after 5 years?
Chance it is still in business after 1 year: 99% = (1 − 0.01)

Chance it is still in business after 2 years: 99% × 99% = (1 − 0.01)2

Chance it is still in business after 3 years: (1 − 0.01)3

Chance it is still in business after 4 years: (1 − 0.01)4

Chance it is still in business after 5 years: (1 − 0.01)5

So the expected market price of this bond which accounts for uncertainty is:

Expected Market Price = Value × Uncertainty = R100 × 1.08−5 × (1 − 0.01)5

1.08 −5
µ ¶
= R100 × .
1 − 0.01
52 Chapter 2 Value

Have a look at the expression in brackets; 1.08/(1−0.01) = 1.091. So 1.08/(1−0.01)

could be approximated by 1.08 + 0.01 = 1.09. This approximation works quite well
as long as the chance of no payment stays small, i.e. equal to no more than a few
percent. Try it with other interest rates (5%, 10%, 15%) and other small annual
risks of bankruptcy (2%, 3%, 4%). You will find that the approximation holds up
U In fact for small x, quite well, allowing us to simplify our calculations:
(1 − x)−1 ≈ (1 + x). Market Price = Value × Uncertainty = R100 × 1.08−5 × (1 − 0.01)5
So if x = 0.01, then ≈ R100 × (1.08 + 0.01)−5
= R100 × (1.09)−5
(1 − 0.01)−1 ≈ 1 + 0.01.
= R65.
This approximation can
be shown using a Taylor So you can see that allowing for uncertainty in the calculation reduces the price
expansion, which you will of an asset, which makes sense: if something is more risky, investors want to pay
meet in your maths course. less for it.
Let’s go back to how we adjusted the interest rate: for a government bond, the
interest rate was 8% p.a. For the corporate bond, we adjusted the interest rate to
allow for uncertainty: we estimated that the company had a 1% p.a. chance of
going under, so we approximately adjusted our interest rate up by 1% to 9% in
order to allow for risk. Such an increased interest rate is called the risk adjusted
rate of interest. Another way to wrap your head around this is that investors
demand higher returns from riskier assets – otherwise they will just buy safe
assets.

Example 2.25

Suppose a complete stranger asked you to lend them R1 000. Most of us would
choose not to lend R1 000 to someone that we have just met. The main reason for
this being, we are unsure when the person would repay the loan, if at all.
There is a large amount of uncertainty around the timing and amounts of the
future cashflows we would receive if we entered into the loan agreement with the
stranger. As a result, we would want to be compensated for taking on the extra risk
of giving this strange person a loan – and the best way to compensate us is to pay
us more interest!
If the person was not a complete stranger but a school friend, we may be more
inclined to lend the R1 000 to them as we know more about this friend than the
complete stranger. For example, we may know they have a job which pays them
well and so they are more likely to repay you. There is a greater amount of certainty
around the future cashflows if we entered into the loan agreement with a friend. As
a result, would not need to increase the interest rate we are expecting to account
for risk – or, at least, not as much.
Another way to see this is to scale up: say you go into business giving loans to
strangers (basically you are starting a bank!). Out of every 100 loans that are due
to be repaid each year, you expect about 3 of them to not repay you. That’s 3% of
your loans going bad! If all 100 people owe you R1000 each, the total amount you
are owed is R100 000, and you think 3% of that (R3 000) will be lost each year.
2.6 Risk discount rates 53

How can you compensate for this problem? One way is to charge everyone else
3% more for their loans. Your 97 “good loans” will pay you R30 (3% of R1 000)
more each and on the whole you will get R2 910 from these borrowers – close to
the R3000 that you lose to the bad loans. What you just did was the same thing as
increasing the interest rate by 3% – to allow for risk.

This example illustrates an important point around the certainty of cashflows.

The greater the uncertainty around a cashflow, the higher the interest rate we will
be valuing it at (to account for the uncertainty) and the less value we are willing
to attach to it now.
There are several ways that the uncertainty in future cashflows can be handled
when we are finding present values. The method we have been discussing is to use
a risk adjusted rate of interest, also called a risk discount rate. The risk discount
rate is the interest rate that we would use to discount the future cashflows but
it has been adjusted to reflect the uncertainty around the cashflows. The risk
discount rate is defined as

risk discount rate = risk free rate + risk premium.

The risk free rate is the interest rate that we would use if there was no uncertainty
around the cashflows (like the rate we use to value government bonds, where
there is no chance of the government not paying out). The risk premium is the
additional interest rate that is included when there is uncertainty around the
cashflows. The greater the uncertainty around the cashflows, the greater the risk
premium.

Example 2.26

For example, suppose you lent the R1 000 to the stranger and they agreed to repay
the loan in a year’s time along with an interest payment of R100. Further, suppose
the current risk free rate is 5% p.a. and you use a risk premium of 10% p.a. due
to the uncertainty. The risk discount rate is 15% p.a. The present value of the
cashflow of R1 100 in a year’s time is

R1 100 v 15% = R956.52.

The present value is less than the R1 000 and so this loan does not look like a good
idea. The stranger would have to make an interest payment of more than R150 for
the loan agreement to be appealing.
Now suppose that you lent the R1 000 to the friend and they agreed to repay the
loan in a year’s time along with an interest payment of R100. The risk premium
that you use for your friend is 3% as you believe that they are more likely to repay
the loan. There is still some uncertainty around whether they will repay the loan
or not and this is reflected in the 3%. Now the present value of the cashflow of R1
100 in a year’s time is
R1 100 v 8% = R1 018.52.
The present value is greater than the R1 000 and so entering the loan agreement
with your friend appears to be a good idea.
54 Chapter 2 Value

You can use risk discount rates in any present calculation where you feel
that there is a likelihood of the cashflow not happening in the future. The risk
premium should be roughly in line with the annual chance of the payment not
happening: there is an implicit assumption that the further in the future the
cashflow takes place, the less likely it is to happen (so the cumulative effect of
the risk premium is stronger for cashflows further in the future). You can use
this method when calculating values of single cashlfows, annuities and any other
combinations of cashflows.

Example 2.27 Consider the following two bonds:

Bond A has a coupon of 5% p.a. payable annually in arrear, a term of 5 years,
and was issued by a large international corporation. We estimate that the
probability of default in any year for this bond is 0.005.
Bond B also has a 5% p.a. coupon payable annually in arrear and a 5 year
term, but it was issued by a small travel agency. We estimate the probability
that the travel agency may go under at 0.04, in any year.
The prevailing interest rate applicable to 5 year bonds is 10% p.a. Before we
calculate the price for R100 par value of these bonds, which of them do you
think is likely to be cheaper?

Price = R5 × a 5 + R100 × (1 + i )−5

The only difference between the 2 bonds will be the value of i . For Bond A,
i will be 10% + 0.5% = 10.5%, and for bond B it will be 10% + 4% = 14%.

Price at i = 10.5% is R79.41;

Price at i = 14% is R69.10.

So the additional 3.5% p.a. in risk results in about R10 difference in price
between the 2 bonds.

Exercises for 2.6 Risk discount rates

Ex.2.29 Let’s say the following people owe you R1000, due back in 1 year:

1. Your dad (probability of not paying: 0)

2. Your sister (probability of not paying: 0.03)
3. Your best friend (probability of not paying: 0.05)
4. That guy you met at that party (probability of not paying: 0.07)

For each of them, work out the present value of the debt, using a 10%
p.a. risk free rate and adjusting for risk. Which one do you think will
have the lowest value?
2.7 Practical applications 55

Ex.2.30 You are thinking of buying two different shares: Share A is for a large
multinational car company, and Share B is for a local florist. Both
shares are expected to pay a dividend of R5 in one year’s time, and
both are expected to grow at 3% p.a. The chance of Company A going
bankrupt in any year is 0.5% p.a. The chance of company B going
bankrupt in any year is 10% p.a. Calculate the price that you would pay
for each of the shares, using a risk adjusted interest rate and a risk free
rate of 8% p.a.

2.7 Practical applications

2.7.1 Comparing assets to each other using present values
Let’s say that you are an investment trader, and you are presented with two
options:

Asset 1 pays R100 per year for 5 years, and you can sell it for R1500 in 5
years’ time; and

Asset 2 pays R150 per year for 15 years.

Both of those assets are for sale for R1200. The risk free interest rate is 10% p.a.
Which asset would you buy? What is better value?
Let’s have a look at the cashflow patterns.

Figure 2.14 Cashflow patterns for the two assets.

From looking at the graph, it’s hard to see which has a higher total value. You
could try just adding up all the money you are going to receive from each asset:
For asset 1, that’s R100 × 5 + R1 500 = R2 000. For asset 2, that’s R150 × 15 = R2 250.
Based on this, asset 2 looks better . . . but recall our work on time value of money.
Cashflows far in the future are worth less to you now than cashflows payable
sooner. So just adding them is not giving you a true comparison. It would be
more meaningful to adjust for time value of money.
56 Chapter 2 Value

In order to compare any offers of money at different times, you will need
to bring the cashflows to the same point in time. You can use any time as a
benchmark, but present value is most common and often easiest. So in order to
be able to solve the above problem, you will need to calculate the present value of
both of these assets. For asset 1, the equation can be set up as follows:
? Check this calculation.
R100 × a 5 + R1 500 × (1 + i )−5 ,
where i is equal to 10% and so the value of asset 1 is then R1 310.
Let’s do the same for asset 2. Can you work out the equation for asset 2? The
? Check this value. present value, calculated at the same rate of interest, is R1 141.
So that gives us:

Asset 1 is worth R1 310, and can be bought for R1 200.

Asset 2 is worth R1 141, and can be bought for R1 200.

Which one would you buy? Why?

You can see from our calculations that asset 1 is a bargain: you can buy it for
less than what you estimate it is worth. Investment traders are always looking for
bargains, and asset 1 is very appealing in this respect.
Asset 2, on the other hand, is overpriced according to our calculations. So you
would not buy it – you would be losing money!

2.7.2 Accounting for risk

In the above example, we assumed that both assets are risk-free: maybe they were
issued by a government. But in real life, we would need to adjust the interest rate
to allow for risk.
Say asset 1 was issued by a pharmaceutical company in Australia. The com-
pany has been in the press lately because their drugs are causing some strange
side effects. . . you think you should add a 4% risk premium when valuing this
asset, so the total interest rate you want to use is 14% p.a. (risk free rate + risk
premium).
Asset 2 was issued by a government in East Africa. You attach a 1% risk
premium to assets issued by this government, so you will value this asset at 11%.
Now the present values are:

PV(Asset 1) = R100a 5 14% + R1 500 × (1 + 0.14)−5 = R1122;

PV(Asset 2) = R150 × a 15 11% = R1 079.

Asset 1 is still worth more than asset 2. But both of them are now valued at less
than the price they are selling for – R1 200. Neither of them is a good deal if risk is
accounted for.
Present values are one way to compare cashflows from assets. Another way is
to compare the yields that the assets offer, i.e. the interest rates at which they are
priced.
2.7 Practical applications 57

2.7.3 Solving for interest rates

Going back to the previous example, we valued asset 1 at R1 122 using an interest
rate of 14% p.a. But the seller was selling it at R1 200 – how did they get that
number?
The seller valued the same cashflows at a different rate of interest. If we can
figure out what this rate of interest is and what the rate of interest is that the
seller of asset 2 is using, we should be able to see which of assets 1 and 2 is more
attractive.
Let’s do the calculations. For asset 1,

R1 200 = R100 × a 5 i + R1 500 × (1 + i )−5 .

We can solve for i using guesses, linear interpolation or Excel. Check that you
agree that i = 12.25%.
For asset 2, the equation of value is as follows:

R1 200 = R150 × a 15 i .

Given that we already know this asset is more expensive than asset 1, are you
expecting the interest rate we calculate here to be higher or lower than 12.25%?
So asset 2, when sold at R1 200, has a much lower yield then asset 1 sold at the
same price. Buying asset 1 means you are earning 12.25% interest – much more
than what you could get from asset 2. But your evaluation of the risk of this asset
says that you are looking for 14% returns to compensate you for the risk of the
pharmaceutical company’s problems; 12.25% is not quite enough – it seems like
you will have to keep looking for another investment opportunity.

2.7.4 Identifying cashflow patterns

A big component of asset valuation is to identify series of cashflows to make the
valuation easier.
Sometimes many different series will be involved, plus some single cashflows.
Cashflow diagrams can be really useful in this work. Consider this schedule of
cashflows:
58 Chapter 2 Value

Time Amount
1 R1 100
2 R100
3 R100
4 R100
5 R100
6 -
7 R100
8 R100
9 R100
10 R100
11 R300
12 R360
13 R432
14 R518
15 R622
16 R1 246
17 R1 396
18 R500
19 R500
20 R500
21 R500

A good way to work with this kind of information is to have a look at the cashflow
diagram:

Figure 2.15 Cashflow diagram.

2.7 Practical applications 59

One way to value these cashflows is just to value each one of them as a single
cashflow:
PV = 1 100v 1 + 100v 2 + 100v 3 + . . . + 500v 21 .
But this is a lot of work unless you can do it on a computer. If we can identify
groups of regular cashflows, the exercise becomes easier:

Figure 2.16 Cashflow diagram with all cashflows classified as belong-

ing to a particular series.
Most of these are easy. Note that we are treating the R100 at times 2 – 10 as
one stream of cashflows with a cashflow missing at time 6 – we could just as easily
treat this as two streams of R100, one from time 2 to 5 and another at time 7 to 10.
There are other ways to group them still – see if you can think of any more!
But what is going on at time 16 and 17? That R1 246 and R1 396 look a bit
strange, and hard to figure out. At times like this, it can be useful to look at the
adjoining cashflows: in this case, the increasing series starting with R300 in year
11, and the R500 starting in year 18. It looks like maybe these 2 series overlap in
years 16 and 17.
If we subtract R500 from R1 246 and R1 396 respectively (effectively assuming
that the R500 series starts at time 16 already), then what remains is R746 and
R896 at times 16 and 17. This fits in with the other series: 746/622 = 1.2, and also
896/746 = 1.2. This gives us the solution:
60 Chapter 2 Value

500

500
1100

896

746
622
500 500 500 500
518
432
360
300
100 100 100 100 100 100 100 100
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Series 1 Series 2 Series 3 Series 4

Figure 2.17 Cashflow diagram with some cashlows classified as be-

longing to a particular series.
This gives us 4 “series” we can sum to get the total present value.
Series 1 is a single cashflow of R1100 at time 1:

PV(Series 1) = R1 100v.

Series 2 is a level annuity of R100 payable for 9 years annually in arrear and
deferred for 1 year (it only starts at time 2), minus R100 at time 6:

PV(Series 2) = R100a 9 × v − R100v 6 .

Series 3 is an annuity increasing geometrically at 20% p.a., starting at R300,

payable for 7 years, and deferred for 10 years (it only starts at time 11):

R300
PV(series 3) = a × v 10 ,
1.2 7 l
where the interest l = (1 + i )/(1 + 0.2) − 1.
And series 4 is a level annuity of R500 payable for 6 years and deferred for 15
years:
PV(Series 4) = 500a 6 × v 15 .
The combined present value can be found by adding the present values of the
four series we have identified.
? If i = 0.12, see if you can Note that the grouping we have used is not the only possibility, there are
calculate the present value alternative groupings which you could have used and they would give the same
of the entire series of cash- answer. Can you think of any of these alternatives?
flows.
Exercises 61

Exercises

Exercises for 2 Value

Ex.2.31 What is the present value of an amount of R10 000 payable in 10 years
time, using an interest rate of 10% per annum effective?

Ex.2.32 Show that ä n = 1 + a n−1 . Explain this result by general reasoning.

Ex.2.33 Show that

1 1
= +i.
an i sn i

Ex.2.34 Mr Christianson is targeting accumulated savings of R10 000 and can

afford savings of R1000 per year, annually in advance.

1. Calculate for how many years Mr Christianson must save to reach

his target if he earns interest at the rate of 10% per annum effec-
tive.
2. Calculate the last annual savings amount that Mr Christianson
should invest (less than R1000) in order to meet his target exactly
at the end of that year.

Ex.2.35 A commonly used ‘rule of thumb’ for the time in which money doubles
itself under compound interest is as follows: if the rate of interest is
K % p.a., money doubles itself in approximately 70/K years. Check
the numerical accuracy of this formula for K = 2, 4, 5, 10, 15. (Some-
times 72/K is used rather than 70/K .) Can you figure out where these
approximations come from?

Ex.2.36 You borrow R10 000 from a friend who needs you to pay back this
amount (without interest) in 1 years time. As a student working a bar
tending job on weekends you earn R1 500 a month. What percentage
of your salary will have to invest at the start of each month for the next
12 months to pay back the loan? You are investing your savings in a
bank account that earns interest at a rate of 2.4% p.a.

Ex.2.37 Consider the R2023 government bond described in Section 2.1.1. You
are told the bond was issued on the 22nd of June 2012 at a price of
R99.81 per R100 nominal and will be redeemed at par at the end of
February 2023. The bond pays semi-annual coupons at the end of
August and February, with an annual coupon rate of 7.75%. The first
coupon is paid at the end of August 2012. Show that the annual yield
on this investment is 8.28%.

Ex.2.38 An investment project has set up costs of R20 000 at the start of the
project, R20 000 two years later and R20 000 after another year (i.e. at
62 Chapter 2 Value

the end of the third year of the project). Income is expected to start
coming in during the fourth year of the project. Income is expected to
be R8 000 per year, received in equal monthly amounts at the end of
each month, for 15 years.

1. Draw a cashflow diagram of the income and outgo of this project

showing the amount of the cashflows, whether it is income or
outgo, and the time of each cashflow.
2. Write down the equation of value that you would use to calculate
the yield for this project.

Ex.2.39 An investor wants to buy a business that is for sale. She estimates that
the business will generate an income of R1 000 per month (at the end
of each month) for the first two years. She will then spend R50 000
on marketing and advertising which she estimates will increase the
monthly income to R2 000 per month for the next 7 years. She thinks
she can sell the business for R100 000 after nine years from now.
Based on the above information, how much will she be willing to pay
for the business now if she wishes to get a yield of 12% p.a. on this
investment?
Solutions 63

Selected Numerical Solutions for Part 1

Chapter 2 Solutions
Ex. 2.1 Simple interest: R130
Compound interest: R133.10

Ex. 2.2 Nominal 12% - 3 months: R103.03

Nominal 12% - 3 years: R143.08
Effective 12% - 3 months: R102.87
Effective 12% - 3 years: R140.49

Ex. 2.3 1. 6.090%

2. R1 806.11

Ex. 2.4 1. R145

2. R153.56
3. 156.54

Ex. 2.5 Values at different times:

Time 2: R1 254.50
Time 5: R1 762.34
Time 10: R3 105.85
Time 20: R9 646.29

Ex. 2.6 R6 575.16

Ex. 2.7 1. R37 332.44

2. R220 949.28
4. R303 776.37

Ex. 2.8 R30 733.81

Ex. 2.9 R81.11

Ex. 2.11 PV of first investment: R931.38

PV of second investment: R848.20
Prefer investment 1.

Ex. 2.12 Total PV using 10% interest: R2,803.60

64 Chapter 2 Value

Ex. 2.13 2. First option: R6 737.26

Second option: R6 751.02
Prefer second option.
3. First option: R6 552.78
Second option: R6 510.11
Prefer first option.

Ex. 2.15 1. R2 723.25

Ex. 2.16 R131.16

Ex. 2.17 1. R107.99

2. R100.00
3. R92.79

Ex. 2.19 1. i = 9%, j = 5% and l = 3.81%

2. R86 350.46
3. R157 852.01

Ex. 2.20 R300 - overpriced.

Ex. 2.22 R76.77

Ex. 2.23 1. R31 641.25

2. R36 076.64
3. R38 043.16

Ex. 2.24 Option 1: R6 000 000

Option 2: R6 571 079.38
Option 2 is better.

Ex. 2.26 The present value is 999.98.

Ex. 2.28 7 years.

Ex. 2.29 1. R909.09

2. R884.96
3. R869.57
4. R854.70

Ex. 2.30 A. R90.91

B. R33.33
Solutions 65

Ex. 2.31 R3 855.43

Ex. 2.34 1. 7 years.

2. R603.74

Ex. 2.36 Repayments are 822.6773 Percentage of salary is 54.85%

Ex. 2.39 R109 595.48

 Module 2

Uncertainty and Risk

Management

67
Chapter 3

Uncertainty

You are now familiar with the concept of risk and its components, uncertainty
and value. Module 1 was focussed on value—you learnt how to think of value as
time-value of money, how important interest was in valuing future cashflows,
and how actuaries and quants always take into account the timing of cashflows,
not just the amount paid.
We also started to think about uncertainty, and how uncertainty can be
brought into valuation of cashflows by increasing the risk-free interest rate by a
“risk premium”. This allowed us to put a value on various assets, such as bonds
and equities.
In this module we will think more about how to put a value on liabilities. And U Liabilities:
the liabilities that we are thinking of are generally quite unpredictable—that is Liabilities are the debts or
the nature of the business of insurance companies and pension funds. obligations that arise for
a company during their
business operations. They
Risk as a threat and as an opportunity are settled by transferring
some form of economic
All companies face risks: the risk that they don’t make enough sales, the risk value.
that employees are unproductive, the risk that people steal from them or commit
In insurance, liabilities typ-
fraud, the risk that perishable supplies go off, the risk of not having enough cash
ically refer to the obligation
to meet payments due—and many other risks associated with the operation of
an insurer has taken on to
the company. We refer to those as “threats”—risks that are bad for the company.
pay benefits to policyhold-
Much work goes into managing these types of risk—the whole topic of “Enterprise
ers sometime in the future.
Risk Management”, that is now one of the actuarial specialist subjects, is all about
helping companies quantify and reduce such risks.
But in addition to having those internal risks (everybody does!), some companies
see other people’s risks as an opportunity to do business. The main examples
of this are insurance companies. Insurance companies are created specifically
to help people manage the risk of death, disability, damage to property, or the
inability to work, amongst other risks. Such companies see risk as an opportunity,
and that is why those companies attract actuaries, who are experts in risk. Further
examples of companies that see risk as an opportunity are asset managers. For
asset managers, the opportunity is to be better at assessing market risk than the
average investor, and because of this to make better buying and selling decisions
in their asset portfolios.

69
70 Chapter 3 Uncertainty

In order to be able to value uncertain cashflows, we need to talk about the

concept of “uncertainty” which is one of the main components of risk.

3.1 How do we measure uncertainty?

There are outcomes in life that are uncertain. Consider the following examples.

Example 3.1
What is the probability of throwing a 6 when using a fair 6-sided die?
Solution: One in six or 61 .

What is the probability of pulling a heart card out of a standard deck of cards?
Solution: 1 in four or 14 , unless there are jokers—in which case, 13 out of 54!

What is the probability of tossing a coin three times and getting 3 heads in a row?
Solution: 1 in 8 or 18 (because there are 8 possible combinations of 3 coins: TTT,
TTH, THT, HTT, HHT, HTH, THH, HHH, and you only want one of them).

All of the above questions are answered by considering the following intuitive
reasoning:
NB: This formula works when
all of the outcomes have the The probability of an event =
same chance of happening.
The number of outcomes which fall into the definition of our event
. (3.1)
The total number of possible outcomes
For example, when identifying the probability of seeing three heads in row, there
is only one outcome which falls into our definition of the event and there are a
total of eight possible outcomes. This gives us the probability of 18 .
Your study of probability and statistics over the next few years will enable you
to make ever more accurate calculations of probability.

Exercises for 3.1 How do we measure uncertainty?

Ex.3.1 What is the probability of the following:

1. Drawing a queen out of a shuffled deck of 52 cards?

2. Winning when betting on “black” in roulette (roulette has 18
black, 18 red and 1 green field)?
T Find out more about roulette
here: [Link] 3. Guessing someone’s star sign correctly? (There are 12 star signs
org/wiki/Roulette. in a year, and a person can be assumed equally likely to have any
Find out more about star one star sign.)
signs here: [Link]
4. Guessing someone’s birthday correctly?
[Link]/wiki/Zodiac.
5. Getting 2 sixes when throwing two 6-sided dice?
3.2 Elementary Probability 71

3.2 Elementary Probability

You will be used to using a variable, such as x, to denote an unknown quantity in
a mathematical formula. We use a capital X to denote the outcome of a particular
event; for example, the outcome of rolling a die. The outcome of rolling a die will
only be known once the die is rolled, and so before it is rolled you can think of X
as eventually taking on a value in a range of possible outcomes, but its current
value is unknown. Such a variable X is known as a random variable.

Exercises for 3.2 Elementary Probability

Ex.3.2 For each of the following events, list all the possible outcomes:

1. A soccer game between team A and team B

2. The results of a university course
3. Taking a flight to Johannesburg

Note for the above exercise: there are many ways to group outcomes—the
results of a university course could be as simple as “PASS” or “FAIL” or they could
be “PASS”, “SUPP”, “FAIL” and “DP REFUSED”, or the full range of percentage
marks.
All of those are correct—we will tend to group outcomes into categories which
make sense given what we are trying to work out. So if a bursary is withdrawn
if you fail a course, the only groupings that are necessary are “PASS” and “FAIL”.
But if a bursary also pays a bonus if you get 75% or more, then “PASS>=75%”,
“PASS<75%” and “FAIL” are more useful categories.
For our purposes, we will confine ourselves to discrete random variables,
i.e. quantities X which can take on one of a finite or countably infinite number
of values, and do so with probabilities that add up to 1. A probability, for our
purposes, is a number between 0 and 1 indicating how likely a particular event is.
We will also define the set of possible outcomes for a random variable X as U If you list all the possible
the sample space of X . Consider the following example. outcomes to an event, their
probabilities must add up
to 100% - since when the
Example 3.2 Suppose that you roll a fair six-sided die, and define X to be event happens, it is 100%
the number of dots showing on the face of the die when it comes to rest. X likely that one of those out-
may therefore take on any one of the values {1, 2, 3, 4, 5, 6}; this is the sample comes comes to pass.
space of X and we can use Ω to denote this set. Since the die is a fair die, the
probability of each of these outcomes is equal. Since there are six equally
likely outcomes, the probability of each is 16 . This can be written as:
½ 1
for x = 1, 2, . . . , 6
Pr(X = x) = 6 .
0 otherwise

This is known as a probability distribution, and we can use it to answer

questions about the probability of more complicated outcomes.
72 Chapter 3 Uncertainty

Example 3.3 The probability that the die will roll a number of three or
above can be calculated as follows:

Pr(X ≥ 3) = P(X = 3) + P(X = 4) + P(X = 5) + P(X = 6)

1 1 1 1
= + + +
6 6 6 6
2
= .
3

T The term probability distri- With that background, we can introduce the probability mass function more
bution is a general term, formally.
and we will use it inter-
changeably with the term
probability mass function. Probability mass function

The probability mass function of the random variable X is the function P(X = x),
T The mathematical symbol often written as p(x), and sometimes referred to as the probability function. The
“∈” means “is an element probability mass function has the following properties:
of”. For example, 1 is an ele-
X
ment of the set A = {1, 2, 3}, p(x) = 1 (3.2)
and this can be written as x∈Ω
1 ∈ A. The term element is and
used to refer to an individual 0 ≤ p(x) for all x ∈ Ω, (3.3)
element of a set.
where Ω is the set of all outcomes of X .
? Can you see if these condi-
tions hold, then p(x) ≤ 1?
Equation (3.2) states that the total probability of all possible outcomes of X
must sum up to one, and Equation (3.3) states that the probability of any outcome
of X must be greater than zero.

Example 3.4
Say you are participating in a lottery where the chance of winning the first prize is
1/1000, the chance of winning one of the second prizes is 5/1000, and the chance
of winning a consolation prize is 1/10. These are the only prizes.

(a) If we define the random variable X to be the outcome of the lottery for you,
what is the sample space of X , i.e. what are the possible outcomes of X ?
(b) What is the probability of not winning anything?

Solution:

(a) The random variable X is the outcome of the lottery for you. So X can be
any one of the outcomes in the set

Ω = {first prize, second prize, consolation prize, no prize}.

3.2 Elementary Probability 73

(b) From the question we know the probabilities of each of the outcomes of X
are

P(X = first prize) = 1/1000 = 0.001

P(X = second prize) = 5/1000 = 0.005
P(X = consolation prize) = 1/10 = 0.1,

and we want to find P(X = no prize).

We know that all the probabilities of all the possible outcomes for playing
this lottery must sum to one. So

1 = 0.001 + 0.005 + 0.1 + P(X = no prize)

and rearranging this we have

P(X = no prize) = 1 − 0.106 = 0.894.

Exercises for 3.2 Elementary Probability

Ex.3.3 List all of the possible outcomes for the following events:

1. The weather at 8am tomorrow morning;

2. Calling a number from a list to sell someone medical aid;
3. A rugby game between the Springboks and the All Blacks;
4. The stock market results by the end of tomorrow

Ex.3.4 To play a game involving dice you have to roll a six on a fair die before
you can start. Let the random variable K be the number of times you
have to throw before you start. What are all the possible values of K , i.e.
the sample space of K ?

Ex.3.5 A company buys a large number of similar machines at time 0. At times

1, 2, 3, 4 and 5, these machines are carefully examined, and 5%, 10%,
15%, 20%, and 25% of the original number of machines are scrapped.
All the remaining machines are scrapped at time 6. Define X as the
number of years until a particular machine is scrapped (the ‘useful life’).
Then the probability mass function of X can be expresses as follows:

x 1 2 3 4 5 6
P(X = x) 0.05 0.10 0.15 0.20 0.25 0.25

1. Find the probability that the useful life of such a machine is at

least 3 years (i.e. 3, 4, 5 or 6 years).
2. Find the probability that the useful life is at most 2 years (i.e. 1 or
2).
74 Chapter 3 Uncertainty

3.3 Intrinsic vs. estimated probability

Throwing dice or drawing cards out of a deck is subject to an intrinsic probability.
This means that we don’t have to do any experiments with actual die throws to
find out what the probabilities of the possible outcomes are—we can find the true
probabilities using logic and science. So for a die, you could measure and weigh
and x-ray the die and be able to figure out how many sides it has, and whether all
the sides are equally likely to come up. With a deck of cards, we can check that
the cards are really indistinguishable from the back, and mix them in a way that
makes their distribution truly random—and then we know each card is equally
likely to come out, we don’t need to check it via experimentation.
But most things in life are not that simple. Think about the probability that the
Stormers win their next match, or the probability that the stock market is going
to go up tomorrow. You can weigh and x-ray the Stormers and their opponents
all you want (they might not like it though!), and you won’t be able to say what
the true underlying probability of a win is. You’ll find that most uncertain events
are like that—there is no way to measure the true underlying probability exactly.
The best we can do is to estimate it.
Usually, the best way to estimate the probability of an event is to look at the
past and find out what has happened before in similar circumstances. How many
matches have the Stormers won in the past? How many have they lost? How many
times have they played against these particular opponents? Did they win then?
What are the other factors that may play a role? Who is the coach and what is his
track record? Have the players changed recently? And so on—it is really hard to
predict the outcome of sporting events—if it was easier, all actuaries would go
into sports betting instead of finance!
So when we are asked to estimate a probability that cannot be measured
intrinsically, usually the best we can do is to look at the past.
Most of the time, we need to be able to answer a very specific question:
What is the probability that x happens? The answer will be estimated by using
probabilities calculated as
T Note the similarity to (3.1).
They are not exactly the The number of times x happened
. (3.4)
same: (3.1) is for the intrin- The number of opportunities for x to happen
sic probabilities while (3.4)
is for the estimated proba-
bilities. Exercises for 3.3 Intrinsic vs. estimated probability

Ex.3.6 Here is a table of male deaths recorded in developed countries in 2000.

The number of deaths attributed to smoking is shown separately.
Looking at the age bracket 35–69, answer the following questions.

1. What would you estimate is the probability of dying from lung

cancer?
2. If someone dies from lung cancer, what do you estimate is the
probability that the death was smoking related?
3.3 Intrinsic vs. estimated probability 75

Male smoking-attributed and total deaths, 2000, developed countries

Age All Causes All Cancer Lung Cancer Population
0–34 Smoking - - -
Total 424 859 20 691 849 305 845 000
35–69 Smoking 805 750 309 897 198 814
Total 2 704 516 708 595 216 780 282 258 000
70+ Smoking 634 106 261 517 176 258
Total 3 322 202 756 200 195 771 24 987 000
Total Smoking 1 439 856 571 414 375 072
Total 6 451 577 1 485 486 413 400 613 090 000
Source: Peto, Richard, et al. (2006) Mortality from smoking in developed countries: Detailed estimates of smoking-attributed deathsby age,

sex and disease, URL: [Link]

3. What is the chance that someone dies from cancer which is smok-
ing related?
4. What percentage of all deaths in this age group are due to smok-
ing?

Ex.3.7 Using the table in the previous exercise, answer the following questions.

1. Can you estimate the probability that someone aged 70 or more

will die from cancer?
2. Could you estimate the probability that a smoker aged 70 or more
will die from smoking-related cancer? What information is miss-
ing from the table?
3. If you were told that 25% of men over 70 are smokers, could you
answer the above question? You can assume only smokers can
die of smoking-related cancer. (Is this a reasonable assumption?)

Ex.3.8 Consider the following sets of data. Can these sets be used to estimate
the probability of the given event happening in the future?

1. Event: A farm being flooded;

Data set: Number of farms flooding in KwaZulu Natal and total
number of farms in South Africa.
2. Event: You crashing your car;
Data set: Number of car accidents registered by the Cape Town
traffic police and number of cars with a CA license plate.
3. Event: Dying of lung cancer;
Data set: Number of deaths in South Africa with “lung cancer”
on the death certificate during 1 year compared to population of
South Africa at the start of the same year.
76 Chapter 3 Uncertainty

3.4 Expected value

How can we put a value on an uncertain event?
Say you are trying to decide between buying one of two lottery tickets. They
both cost R20 and have only one prize each. For the first lottery, the probability of
winning is 0.01 and the prize is R2 500. For the second lottery, the probability of
winning is 0.000035 and the prize is R500 000. Which lottery is the better to play
in?
It is hard to compare these—both of the prizes and the probabilities of win-
ning are different. We can however work out the expected value of each of the
outcomes and use this to help us make a decision.
The expected value is the mean of the probability mass function, and is
defined as:
E(X ) = µx =
X
x × p(x),
x∈Ω

with the summation, as before, being over all possible values X , i.e. the set Ω.

Example 3.5 Returning to the die example where the random variable X is
the number of dots on the face of the die after a roll. We can think of the
expected value or the mean as being the average value of X , if we threw the
die infinitely many times. In this case, the expected value is

6
X 6
X 1 1 2 3 4 5 6
x × p(x) = x× = + + + + + = 3.5.
x=1 x=1 6 6 6 6 6 6 6

This may look odd as intuitively we anticipate that the expected value is the
outcome we would expect to get on a single throw, but clearly no standard
six-sided die has an outcome of 3.5 on one of its sides!
To avoid being confused by this, it may be helpful to remind yourself that
expected value really means average or mean outcome, and to think in
terms of the long-term average outcome rather than the outcome on any
one throw.

For many of the examples that we consider, there is only one outcome where
x and p(x) are non-zero. For example, if we define the random variable Y to be
the random variable representing the financial consequences of you passing or
failing this year, then if you pass Y = 0, i.e. there is no additional cost, but if you
fail Y = R30 000, i.e. you will have to pay a cost equal to R30 000. So the expected
value of the financial consequences of passing or failing the year, ignoring interest,
is: X
E(Y ) = y × p(y) = R0 × P(Y = 0) + R30 000 × P(Y = R30 000).
y∈Ω

But as Y is determined by you passing or failing, we can swap out P(Y = 0) with
P(pass) and P(Y = R30 000) with P(fail) as P(Y = 0) = P(pass) and P(Y = R30 000) =
3.4 Expected value 77

P(fail). The expected value can be rewritten as

E(Y ) = R0 × P(pass) + R30 000 × P(fail).

We can ignore the “pass” scenario because there are no consequences, i.e. Y = 0.
This means we can simplify the expected value calculation in this case to:
U Doesn’t this look familiar?
Expected Value(X ) = Probability(X ) × Value(X ). From the start of the course,
we have been calling uncer-
The expected value is basically a way to measure the value of uncertain future tainty and value the compo-
events taking both uncertainty and value into account. nents of risk. This equation
For each of our two lotteries, there two possible outcomes, either you win, for expected value finally
and get the prize money, or you lose, and get nothing. So if we define L 1 to be combines these compo-
the random variable representing the pay-off of the first lottery to you, then the nents into one number—a
simple way of putting a
sample of space of L 1 is {0, R2 500}. The expected value of L 1 , the pay-off from the
value on risk!
first lottery, is
E(L 1 ) = 0 × 0.99 + R2 500 × 0.01 = R25.
The 0.99 is the probability of not winning the first lottery. Similarly, if we define
L 2 to be the pay-off to you from the second lottery, then its expected value is ? Can you show how the 0.99
was calculated?
E(L 2 ) = 0 × 0.999965 + 0.000035 × 500 000 = R17.50.

Which lottery is better value? The first lottery only offers R2 500 as a prize, but
the probability of winning is so much higher than the second lottery that the
expected value is much higher. Not only that, the first lottery costs less to enter
than the expected value that you can get out of it, i.e. R20 cost vs. R25 expected
benefit—even if you are against lotteries normally, this is a no brainer!
Another way to think of expected values is as the outcome of a particular
scenario, if the scenario is repeated lots of times. So for example, if 100 people
played the first lottery in the example above, 1 person would win R2 500 and
99 others would get nothing. So you can think of the expected value as what
everyone would get on average: R2 500 averaged between 100 people is R25.
More generally, it can (but won’t here) be shown that for any function g (X ) of
a discrete random variable X :
X
E(g (X )) = g (x)p(x).
x

This result is used so often that it is sometimes called The Law of the Unconscious
Statistician. If you take g (X ) to be just a X + b, a and b being some constants,
you should be able to show that E(aX + b) = aE(X ) + b.

Exercises for 3.4 Expected value

Ex.3.9 For each of the games below define the random variable to be the net
winnings from the game, i.e. the winnings after any cost to play, and
78 Chapter 3 Uncertainty

1. identify the sample space for the random variable, supposing you
only play one round of each game,
2. write down the probability distribution for the random variable,
and
3. find the expected value of the random variable.

Game 1: A game, with no cost to play, where you draw a card and win
Rands equivalent to the face value of the card. For reference, a
deck of cards consists of 4 suits of 13 cards each, i.e. 52 cards
altogether. The cards in each suit are: Ace, 2,3,4,5,6,7,8,9,10, Jack,
Queen and King. In this game, an Ace wins R1, a 2 wins R2, a 3
wins R3, and so forth. Jack, Queen and King all win R10 each.
Game 2: The same game as above, but you have to pay R5 for every card
drawn.
Game 3: A game of chess against a chess grandmaster, where if you win,
he pays you R100 and if you lose, you pay him R10. You are
reasonably good at chess, and estimate the probability that you
will win to be 0.05.

Ex.3.10 In roulette, there are 37 slots on the wheel. 18 numbers are coloured
black and 18 numbers are coloured red. The number 0 is coloured
green—no-one can bet on this number and so when the ball lands on
0, everyone loses and the casino wins. If you bet on “black” and the
ball lands on black, you will double of your money, otherwise you will
lose your bet.

1. Suppose that you bet R100 on black. Define the random variable
X to be the pay-off from this bet. Write down the probability mass
function for X .

Figure 3.1 Traditional roulette

2. Calculate the expected value of X .
wheel.
Ex.3.11 Referring back to Exercise 3.5, what figure do you think best represents
the average useful life?

3.5 Expected values of cashflows at different times

It’s time to combine what we have learnt about time-value of money and un-
certainty. We have been working with expected values as the probability of X
multiplied by the amount X without considering the time-value of money. This
effectively means that we have been assuming that all of the events are happening
today, and the payments will also take place today.
In reality, we will mostly be looking at payments and events which will be
taking place in the future. So the expected value will need to incorporate the
3.5 Expected values of cashflows at different times 79

concept of time-value of money. We will refer to these expected values which

allow for the time-value of money as expected present values. We introduce the
ideas using an example.

Example 3.6
Suppose you have a bursary from an institution that pays out a bonus of R5 000 at
the end of the year if your average mark at the end of first year is greater than 80%,
and nothing otherwise. You think that you have a 1 out of 3 chance of getting this
bonus. It is now the end of April. Let Y be the random variable representing the
cashflow from the bonus, i.e.
½
5 000 if you get an average greater than 80%
Y = .
0 otherwise

The expected value of Y allowing for time-value of money is called the expected
present value of Y .
Using an interest rate of 11% p.a., what is the expected present value of this bonus
to you?

Solution: The expected present value (EPV) of Y is

EPV(Y ) = P(average > 80%) × 5 000 × v 8/12 + P(average ≤ 80%) × 0 × v 8/12

Note we are taking the time value of money into account.
1
= × R5 000 × v 8/12
3
= R1 554.65.

The bursary company has announced that it is changing this scheme so that you
get a bonus of R2 500 if your average is 70% or more at the end of the first semester,
i.e. at the end of June. The bonus is paid at the end of June. You think your chances
of getting an average of 70% or more in this semester are 3 out of 4. What is the
expected present value of this new bonus? If you have to choose between the new
and old scheme, is it in your advantage to switch to the new scheme?

Solution: We first need to define the random variable W to be the bonus payment,
i.e. ½
2 500 if you get an average greater than or equal to 70%
W= .
0 otherwise
The expected present value of W is

EPV(W ) = P(average ≥ 70%) × 2 500v 2/12 + P(average < 70%) × 0 × v 2/12

3
= × R2 500 × v 2/12
4
= R1 842.67.

The expected value, allowing for time value of money, is greater in the second
case—so it looks like the switch will be beneficial to you.
80 Chapter 3 Uncertainty

We can use the concept of expected present values in a situation with more
than one payment, as long as we know the probability of each payment happening.

Example 3.7

Say at the start of the year, you are playing in an on-line game where at the end of
calendar month n you have a chance to win R100n with a probability of (1−0.05n).
So in January, you can win R100 with probability 0.95, while in June you can win
R600 with probability 0.70, and so on. Let’s assume that you use an interest rate of
3% per month to value any pay-off.

(a) If you can only choose to play in one month, and you have to pay for your
ticket today, which month is the best? Assume the cost to play for all months
is the same.
(b) If you can buy a ticket to play for the whole year, what would be a fair price
to pay for the ticket?

Solution:

(a) For this example, we define the series of random variables X n for n = 1, 2, . . . , 12,
where X n is the winnings in the nth month, i.e.
½
R100n with probability (1 - 0.05n)
Xn = .
0 with probability 0.05n

The expected present value of the winnings in the nth month is:
? Can you show where this
formula comes from? EPV(X n ) = (1 − 0.05n) × R100n × v n .
T Build a spreadsheet which
For each of the months, we have the following:
does this calculation for you!
January: EPV(X 1 ) = R100 × 1 × v 1 × (1 − 0.05) = R92.23,
February: EPV(X 2 ) = R100 × 2 × v 2 × (1 − 0.10) = R169.67,
March: EPV(X 3 ) = R100 × 3 × v 3 × (1 − 0.15) = R233.36,
April: EPV(X 4 ) = R100 × 4 × v 4 × (1 − 0.20) = R284.32,
May: EPV(X 5 ) = R100 × 5 × v 5 × (1 − 0.25) = R323.48,
June: EPV(X 6 ) = R100 × 6 × v 6 × (1 − 0.30) = R351.74,
July: EPV(X 7 ) = R100 × 7 × v 7 × (1 − 0.35) = R369.96,
August: EPV(X 8 ) = R100 × 8 × v 8 × (1 − 0.40) = R378.92,
September: EPV(X 9 ) = R100 × 9 × v 9 × (1 − 0.45) = R379.38,
October: EPV(X 10 ) = R100 × 10 × v 10 × (1 − 0.50) = R372.05,
November: EPV(X 11 ) = R100 × 11 × v 11 × (1 − 0.55) = R357.60,
December: EPV(X 12 ) = R100 × 12 × v 12 × (1 − 0.60) = R336.66.

It looks like the best month to play is September—it has the highest expected
present value.
3.6 Case studies 81

(b) If you wanted to buy a ticket for all of the months together, a fair price would
be the sum of all the expected values:

12
X
EPV(X n ) = R3 649.36
n=1

So if the game was being offered for less than that, it would be worthwhile!
U A game or an exchange is
termed fair when the price
we are paying is equal to the
expected present value of the
3.6 Case studies benefits we receive from the
game or exchange.
The valuation of uncertain cashflows is at the core of actuarial work. Here are four
case studies presenting different types of risk that people are exposed to and the
resulting uncertain cashflows.

3.6.1 Case Study 1: Death

Sizwe is married. He and his wife both work—he is a reporter and his wife, Mary,
runs a catering business.
Sizwe has been offered a 3 month contract reporting from Afghanistan. It’s
a great offer, and it means a promotion at work. But the job is a bit dangerous,
and the company has offered to give them some peace of mind by offering them
R1 000 000 if Sizwe dies while reporting in Afghanistan. The R1 million will be
payable at the end of the 3 months, and is referred to as a death benefit.
What is the expected value of this R1 000 000 death benefit? What information
do we need?
First we need to define a random variable. Let Y be the random variable
which takes on the following values
½
1 000 000 if Sizwe dies while in Afghanistan
Y = .
0 if Sizwe survives his trip to Afghanistan

I.e. Y represents the pay-off of the death benefit.

The expected present value of Y can be written as

EPV = P(Sizwe dies in Afghanistan) × PV(1 000 000)

+ P(Sizwe survives Afghanistan) × 0.

So to answer the question we need P(Sizwe dies in Afghanistan) and an interest

rate i for our present value calculation.
In Module 3, we will talk about how to calculate a probability of death in much
more detail, but for now let’s assume that the probability of a reporter dying over
3 months in Afghanistan is 0.00025. And let’s say that the interest rate is 15% p.a.
Then
3/12
EPV = 0.00025 × 1 000 000v 15% = R241.42.
82 Chapter 3 Uncertainty

Exercises for 3.6.1 Case Study 1: Death

Ex.3.12 If Sizwe is sent to Afghanistan for 6 months instead of 3 months and the
probability that a reporter dies during a six month period in Afghanistan
is equal to 0.0007, what is the expected present value of the the R1 mil-
lion death benefit? The benefit will be paid at the end of the six months.
Use an interest rate of 15% p.a.

3.6.2 Case Study 2: Property damage

Natasha lives in a house in Constantia with her cats. The building is worth
R8 million and all her furniture and possessions are valued at R1 million. She has
just received a newsletter from the neighbourhood watch saying that based on
crime and incident statistics, the probability of any house in Constantia being
burgled is 0.007 in any year. When houses are burgled, the thieves take on average
40% of the furniture and possessions. The same newsletter also states that the
probability of a house burning down is much lower, and estimated at 0.000005 in
any year. When a house burns down, on average 90% of the value of the building
and contents are destroyed.
Natasha did first year Actuarial Science before she became a professional cat
therapist, so she wants to work out the expected present value of being burgled or
her house burning down, based on these statistics.
Let’s look at the burglary first. We need to define a random variable. Let L b
be the random variable representing the value of the possessions that Natasha
loses during a burglary. We will assume that Natasha can only be burgled once
a year and if she is burgled she will lose 40% × R1 000 000 = R400 000 worth of
possessions and furniture. So
½
400 000 if Natasha is burgled
Lb = .
0 if Natasha is not burgled

Now the expected present value of L b is

EPV(L b ) = 400 000v n × P(L b = 400 000) + 0 × P(L b = 0)

= 400 000v n × P(burglary) As P(L b = 400 000) = P(burglary)

We know that the P(burglary) = 0.007. But we need an n and an interest rate i for
our discount factor v n .
For v, what interest rate should we use? Let’s leave that up to Natasha, who
decides to use an interest rate of 7% p.a.
Finally, what should we use for n? The first thought is 1 year, because the
probabilities are being quoted for 1 year. But if we do that, we are saying we
expect the burglary to happen on average on the very last day of the year. This
is unlikely. It’s more likely the thieves work all year round, just like everybody
else. So if we assume that the burglary could happen at any time during the year,
what is the “average” time it will happen? We can assume that on average, all the
3.6 Case studies 83

burglaries that do happen over the next year will happen in the middle of the
year, i.e. 6 months from now. Some will happen sooner, some later, but if they are
uniformly distributed over the year, which we are assuming here, then they will
happen on average 6 months from now. So let’s make n = 0.5.
Now we can work out the EPV: U This approach to approx-
imating the timing of a
0.5 payment as happening
EPV = 400 000v × 0.007
half-way through a time
= R2 706.86. period is very common. It
makes the assumption that
In passing, we note that the EPV can be decomposed into two parts, one relates incidents are distributed
to the value and the other the uncertainty, i.e. uniformly over the time pe-
riod. Any incident that is
0.5
EPV = 400
| 000v
{z } × 0.007
| {z } . not distributed uniformly
Value Uncertainty should not be valued in this
way—think for example of
Now for the fire. Again we start by defining the random variable. In this case snow damage in cold coun-
let L f be random variable representing the value of property and possessions that tries. The distribution of
Natasha loses during a fire. We will assume that Natasha’s house can only burn incidents of damage from
down once in a year. If it does burn down, what will be the cash value she loses? snow would not be uniform
over the year—you would
The cash value of the damage will be 90% of the house plus 90% of the contents
expect the distribution to
— everything inside will also burn! So that’s 90% × (R8 000 000 + R1 000 000) =
be concentrated in the win-
R8 100 000. So ter months. You may then
½ use a different assumption;
8 100 000 if Natasha’s house burns down
Lf = . for example, that the inci-
0 if Natasha’s house does not burn down dents occur on average in
the middle of winter and
The EPV of L f is not the whole year.

EPV(L b ) = 8 100 000v n × P(L f = 8 100 000) + 0 × P(L f = 0)

= 8 100 000v n × P(fire) As P(L f = 8 100 000) = P(fire)

Again n will be 6 months, since we expect fires to happen uniformly over the next
year and i is still 7% p.a. So the EPV is

EPV(L b ) = 0.000005 × R8 100 000 × v 0.5 = R39.15.

Fires seem to be a much smaller risk than burglaries for Natasha! Maybe it’s time
for Natasha to improve the burglar bars. . .

Exercises for 3.6.2 Case Study 2: Property damage

Ex.3.13 For the situations below, state when you would expect the event to take
place on average:

1. A car accident that may happen during the course of next calendar
year;
2. A tax refund, which will happen some time in March;
84 Chapter 3 Uncertainty

3. A drought in the Western Cape, which has winter rainfall (i.e.

assume that droughts are uniformly distributed over the summer
months, being November–March).

Ex.3.14 Calculate the expected present value of the following medical pro-
cedures using 13% interest per annum and assuming the events are
uniformly distributed over the given periods.

1. Removal of wisdom teeth: probability you will need your wisdom

teeth removed over the next 10 years is 0.60, and the cost of the
procedure is R5 000 now, but the cost is expected to increase by
5% p.a. (Hint: first calculate the expected time of the procedure,
and the amount it will cost at that time.)
2. Caesarean section (C-section) for a woman currently 2 months
pregnant. The likelihood that a C-section will be needed is 45%,
and the expected timing of the C-section is sometime during the
9th month of pregnancy. The cost of the procedure is R8 500.
3. A dislocated knee for a skier: the person plans to go on a two week
skiing holiday 6 months from now. The chance of dislocating a
knee is 0.02, the cost of medical treatment is R3 000. Assume all
months are of equal length and each have 4 weeks.
Chapter 4

Managing Risks

4.1 Types of risks

While all risks have a component of uncertainty and value, risks may differ from
each other in many ways. In order to think about ways to manage risks, we need
to start by classifying risks into different categories.

Example 4.1 Think about the differences between these risky events:

• Playing the lottery

• House fire
• War
• Tsunami
• Increase in food prices
• Gambling at a casino
• Car accident
• Unemployment

Can you think of ways they differ from each other?

85
86 Chapter 4 Managing Risks

4.2 Pure vs. speculative risks

Consider the following two lists of risks:
List one: List two:

1. Gambling in the casino, 1. House burning down,

2. Playing the stock market, 2. Death,

3. Applying for a new job, 3. Car Accident,

4. Betting on horse races, 4. Unemployment,

5. Playing the lottery, 5. Illness,

6. Proposing marriage. 6. Robbery.

Can you identify how the risks in the first list differ from the risks in the second
list?
The first list contains examples of what we call “Speculative Risks”. Specula-
tive risks have at least one outcome which improves your circumstances, e.g. you
can win at a casino or on the stock market, you can get a job that is better than
your current one, and hopefully someone agreeing to marry you can be seen as
a positive outcome. Note that speculative risks can have a downside—you can
certainly also lose at the casino! But there is at least some chance of winning.
In contrast, the risks in the second list are called “Pure Risks”. Pure risks have
no “upside”. The best you can hope for when you are facing a pure risk is that
nothing bad happens—that the status quo continues. The “good” outcome is that
your house does not burn down, you don’t have a car accident, and you keep your
job.
T Pure risks are the depress- When we did calculations of expected present values in the last chapter,
ing kind of risk—you can’t pure risks were easier to work with because we only needed to worry about the
win, at best you come out
probability of the “bad” event happening, and the value incurred by the bad
even.
event—the loss.

Example 4.2 So for example, the expected present value of the loss due to
a car accident could be calculated as the probability of having an accident
over some future period of time, multiplied by the present value of the
anticipated cost of the damage to cars and people. We would not need
to add the probability of not having an accident, because the financial
implications of not having the accident are 0:

EPV(loss due to car accident)

= P(car accident) × PV(damage) + P(no accident) × 0
= P(car accident) × PV(damage).
4.3 Dynamic vs. static risks 87

Note that when talking about pure risks, the components of the risk can be
seen as
Pure risk = Uncertainty & Loss,
instead of uncertainty and value—that is because pure risks are all about losses
and there are no positive outcomes.

Exercises for 4.2 Pure vs. speculative risks

Ex.4.1 For the list of speculative risks at the start of Section 4.2, state the
potential positive, negative and neutral outcomes.

Ex.4.2 Here is a list of risks that a supermarket owner faces:

1. The risk of stock being stolen from the shelves,

2. The risk of selling genetically modified foods,
3. The risk of selling vegetables without packaging, i.e. loose,
4. The risk of an increase to the price suppliers charge for bread,
5. The risk of staff being absent from work due to illness.

6. Which ones are speculative risks and which ones are pure risks?
7. What is the potential downside (and, if relevant, upside) of each
risk?

4.3 Dynamic vs. static risks

Have a look at these two lists and try to spot the common features:

List one: List two:

1. Risk of an increase in mort- 1. Risk of having a car accident,

gage repayments,
2. Risk of a house fire,
2. Risk of an increase in food
prices, 3. Risk of a flood,

3. Risk of playing the stock mar- 4. Risk of an earthquake in Los

ket, Angeles,

4. Risk of becoming unem- 5. Risk of having a heart attack,

ployed, 6. Risk of becoming disabled.
5. Risk of not receiving an iPad
for Christmas.
88 Chapter 4 Managing Risks

The risks in the first category are called “Dynamic Risks” because they change
as the economy changes. This may be because the chances of the event, or the
value of the event, will be different under different economic conditions in the
short term, e.g. 1–5 years. Other trends, such as changing consumer preferences
or advances in technology, are often also considered when evaluating whether a
risk is dynamic or not.
T It is worth noting that most “Static Risks”, then, are risks that are considered stable over the short term—
risks are dynamic over the you are probably just as likely to have a car accident whether the economy is
very long term—mortality
doing well or not, and chances of your house burning down do not significantly
rates reduce over time as
people become healthier
change from year to year.
and living conditions im-
prove, and even the chance Exercises for 4.3 Dynamic vs. static risks
of natural catastrophes
such as floods and earth- Ex.4.3 For each of the dynamic risks in the list at the start of Section 4.3, list
quakes may change over some reasons why the level of risk may fluctuate from time to time.
time; for example, due to
global warming or shifts Ex.4.4 For each of the “static” risks in the list at the start of Section 4.3, can you
in the earth’s crust. So the think of something that may change in the long term that will make
term “static” is often ap- these risks change, i.e. something that will make them dynamic in the
plied over the short term, long term?
to denote risks that do not
change very rapidly and un-
predictably over time. The 4.4 Fundamental vs. particular risks
boundaries in terminology
can be blurry sometimes! Again, consider the two lists of risks:

List one: List two:

1. Risk of war, 1. Risk of a plane crash,

2. Risk of tsunami, 2. Risk of heart disease,

3. Risk of flood, 3. Risk of becoming disabled,

4. Risk of economic downturn, 4. Risk of theft,

5. Risk of pollution, 5. Risk of a house fire,

6. Risk of inflation. 6. Risk of the CEO dying.

“Fundamental Risks” are risks which are not independent from each other—
they are likely to affect large numbers of people when they occur. The first list
shows some risks that can be considered fundamental; e.g. a war will mean that
many people are killed, displaced or lose their possessions; inflation means that
prices go up for everybody and the cost of living increases for everyone all at once.
When the economy is doing poorly, the number of jobs reduces and thousands of
people become unemployed.
4.5 Risk management 89

The other types of risks are “Particular Risks”. Particular risks affect individ-
ual people randomly and independently; e.g. one house may be more likely to
burn down than another, but it’s unlikely that thousands of houses will catch fire
all at once. Particular risks arise from the personal circumstances of each person,
not from large scale events. They are independent from one another. This makes
their occurrence random, while incidents of fundamental risks tend to cluster
together.
Note that the line between fundamental and particular can be blurry too: it
often depends on the size and diversity of the group that is being considered.
For example, consider the house fire example above. Generally, fire is seen as a
particular risk as generally houses burn down randomly and independently from
each other. But think of an informal settlement where houses are close together
and there are few precautions against fires. In those circumstances many houses
may burn down at the same time—here fire is more of a fundamental risk.

Example 4.3 A small fishing community situated on the coast of Mozam-

bique suffers from frequent flooding. For the members of this community,
flooding is a fundamental risk: when it happens, it happens to just about
everybody. But a researcher studying the risks affecting people living in
Africa may not consider flooding as a fundamental risk; only a small propor-
tion of people in Africa will ever experience a flood, and even those who are
at risk of flooding will not all get flooded at once. The incidence of rainfall,
tidal waves and human settlements across Africa make flooding sufficiently
random to be considered as a particular risk.

Many of the risks that actuaries deal with are particular, static and pure. This
is because such risks are most suitable for insurance purposes. We discuss this
more in Chapter 5. Quants, in their asset management role, are also interested in
dynamic, speculative risks—those are the risks of the stock market.

Exercises for 4.4 Fundamental vs. particular risks

Ex.4.5 For the fundamental risks at the start of Section 4.4, think of a context
where these risks would cease being fundamental risks and become
particular risks.

4.5 Risk management

We are all exposed to risks all of the time. Depending on the nature of the risks,
there are different things we can do to manage them. Risk management ap-
proaches include:

• Acceptance,

• Avoidance,
90 Chapter 4 Managing Risks

• Reduction,

• Self-funding,

• Information,

• Risk sharing (also called risk pooling),

T Risk sharing, together with
risk transfer, are also known • Risk transfer.
as insurance.
These approaches are not all equally suitable for all risks. Let’s look at them in
turn.

4.5.1 Acceptance
One way to deal with risk is to simply accept it. There is no way that we can
completely remove the risks from our lives—no matter how careful you are, you
may still contract a disease or be in an accident. It is necessary to accept some
risks in life. Risks you may consider accepting include:

• Risks that have consequences that you can bear (see self-funding in Section
4.5.4)

• Risks that have the potential of upside, i.e. speculative risks. This is the
“nothing ventured, nothing gained” type of situation where trying to avoid
risks which can result in a positive outcome may rob you of opportunities.

• Fundamental risks—those are often difficult to avoid or manage, as they

affect everybody and are outside your control.

Exercises for 4.5.1 Acceptance

Ex.4.6 List some risks that you are exposed to and that you have accepted.

Ex.4.7 List some risks that you are exposed to and do not want to accept.

4.5.2 Avoidance
Some risks are avoidable. This involves not putting yourself in a situation which
involves the risk; e.g. there is no need to gamble in a casino, or fly somewhere in a
plane. Many risks are completely avoidable, but before you decide to avoid them,
you should consider:

• Are you foregoing opportunities by avoiding the risk? This applies to spec-
ulative risks in particular. For example, not investing in the stock market
means that you will not incur losses when there is a stock market crash, but
you will also not benefit from investment growth.
4.5 Risk management 91

• What are the costs/consequences of avoiding the risk? For example, avoid-
ing planes means that you have to drive instead. The consequence is that
you will take much longer to arrive at your destination. You are also expos-
ing yourself to a different risk—the risk that you have a car accident. All of
the options and their consequences need be weighed up.

Example 4.4
Let’s think about the flying vs. driving example a bit more. Suppose that you need
to go to Johannesburg in two weeks’ time for a meeting, and further suppose you
can either fly or drive. You need to choose one of these methods, how do you
choose? What are the costs and risks you are faced with?

Solution: You need to weigh up the costs of each method:

The main costs of driving include: The main costs of flying include:
(a) petrol, (a) the plane ticket,
(b) wear and tear on your car, (b) transport to the airport,
(c) accommodation on your journey, (c) transport to the meeting,
(d) the cost of losing out on two days’ work (d) the cost of losing out part of a day’s
while you are driving, work while you are flying,
(e) and some food. (e) and maybe some food for the trip.

You will also need to weigh up the risks of each method:

The main risks of driving include: The main risks of flying include:
(a) risk of a car accident and the conse- (a) risks of delays,
quences, (b) risk of a plane crash (this is actually a
(b) risk of car trouble and the conse- small risk compared to the risk of a car
quences, such as delays, accident),
(c) and the risk of getting lost. (c) and the risk of the taxi to the airport
and the meeting being late or getting
lost.

Once you have considered these costs and risks, you can decide on the best way to
get to your meeting. It is likely that flying is the better option.
92 Chapter 4 Managing Risks

4.5.3 Reduction
Reducing a risk refers to either reducing the probability of the event occurring, or
reducing the loss associated with the event.
For example, hedging is a technique used by asset managers to reduce their
exposure to the market. When a manager uses hedging and the market falls, the
value of their portfolio does not fall as much as the market value. However, this
comes at the expense of losing out if the market goes up.
Another example of risk reduction are preventative medical interventions—
regular health check-ups, dental check-ups, a healthy diet and exercise—which
are great at reducing the risk of illness.

Case study 2 cont.

Consider the case of Natasha and her house in Constantia. What can Natasha do
? What about the risk of fire? to reduce the risk of burglary?
How can she reduce the risk
She can:
of fire?
• install burglar bars, alarm, proximity sensors,
• hire a security company to protect her house, or
• get dogs instead of, or in addition to, cats.

4.5.4 Self-funding
Whether you accept all of a risk, or reduce the risk but retain some of it, you may
need to self-fund the risk. Self-funding means that you accept the full financial
burden of the risk—when a negative event happens (for example, you have to
have an operation), you pay the costs associated with the outcomes (the cost of
hospital and doctor bills).
Self-funding is not for everyone—you have to have access to a pool of savings
or a line of credit to consider it. The decision would be based on:

• The likelihood of the risk occurring;

• The amount that you may have to fund, compared to your level of financial
resources;

• The availability and cost of other methods of funding, for example insur-
ance;
? Say you decide to self-fund
the risk of disability as well • What other risks are you carrying? Are they correlated with this risk?
as the risk of illness. Are
these correlated with each
other? What does that mean Exercises for 4.5.4 Self-funding
for you?
Ex.4.8 Consider your parents. What risks are they subject to? What risks do
you think they might be self-funding?
4.5 Risk management 93

4.5.5 Information
“Information is power” in every aspect of life, and risks are no exception. Risks ? Think about government
can be managed by becoming more informed about them. For example, under- posters or adverts you may
standing how HIV is transmitted allows people to significantly reduce the risk of have seen recently. Are any
becoming infected with HIV. Learning how to drive more defensively reduces the of them educating the pub-
lic about risks? What is the
risk of having a car accident, and so on.
interest of the government
Almost every risk is worth knowing more about—this knowledge will also in educating the public, par-
assist you with making decisions about the most suitable risk management ap- ticularly about health risks
proach. such as HIV/AIDS? Find out
more about HIV/AIDS in
4.5.6 Risk sharing (or pooling) South Africa here: http://
[Link]/[Link].
Risk pooling is a powerful way to reduce risk. It involves a group of people agreeing
to share a risk that they are subject to. Instead of having a single person in the
? Risk pooling does not work
group being affected by the risk while the others are lucky enough to avoid it, the well on fundamental risks.
group agrees that everyone in the group will share the costs of the events that Can you think why?
happened to a few of the members.
Risk pooling relies on the statistical concept of the “Law of Large Numbers”.
We will look more closely at risk pooling and the law of large numbers later in this
chapter.

4.5.7 Risk transfer

Risks can be transferred to a third party. This means that if the risky event takes ? Who would do that? What
place, the third party will compensate you for the losses that you incur. The third institutions or people can
party is effectively taking on the risk. you think of that may as-
sume the risks from others?

Example 4.5 At the moment, it is likely that most of your serious financial
risks have been transferred to a third party. If something happened to you, ? Have you asked yourself
or your car, or your Blackberry, would your parents assist you with meeting what you are paying in ex-
change for this risk transfer?
the financial costs? If so, well done, you have transferred your risks to your
parents.

In exchange for accepting your risk, the institution that you are dealing with
will probably require something from you. An insurance company will charge
a premium for their services, while the government will collect taxes for their
services. However it may happen, the risk transfer is generally an exchange of
the risk of a large uncertain financial loss for a small certain financial loss, e.g.
exchanging the costs of a car accident which may happen in the future for the
premiums you have to pay to the insurer.

Case Study 1 cont.

94 Chapter 4 Managing Risks

Sizwe is heading to Afghanistan. He and Mary are aware of the increased risk to
his life during the next three months, and they are considering what they can do to
protect themselves:

T Acceptance: there is a part of this risk that will be impossible to manage

away. Sizwe has to accept that he might die. Mary has to accept that there
is a chance that she may lose her husband. The emotional and other non-
financial consequences of most risks cannot generally be taken away.
T Avoidance: Sizwe could still say no to the job. He would lose his promotion,
but avoid the risk of being shot. Sizwe and Mary should weigh up the likely
benefit of the promotion and the increased career prospects that go with
it against the increased risk that they are temporarily taking on. They pre-
sumably did this, and have decided that the risk was acceptable given the
upside.
T Reduction: is there anything that Sizwe could do to be safer? There are
probably precautions one can take to increase safety. He could wear a bullet
proof vest in the field, and he could talk to locals to understand better how
to behave and where to go to be safe. Risk reduction is often linked with
increasing knowledge—understanding the environment and the risks better
will definitely reduce the risk in this situation.
T Self funding: Mary is employed and would still be able to provide for some
of the family’s needs.
T Risk pooling and risk transfer: by offering Sizwe’s family a R1 000 000 death
benefit, Sizwe’s employer has already accepted some of the risk of this trip.
But the family could take out a 3 month term assurance policy (we will learn
more about different types of life assurance in Module 3) for Sizwe over this
period. However, the insurance company may refuse to cover Sizwe—can
you think why?
Sizwe is heading into an area that may be considered at war. War is a funda-
mental risk, and everyone in Afghanistan is exposed to this risk. Risk pooling,
which is one of the principles underlying insurance, does not work for fun-
damental risks—we will find out why in Chapter 5—Insurance Principles.

Exercises for 4 Managing Risks

Ex.4.9 Consider the other case study. How can each of the risk management
methods discussed above be applied to Natasha and the risk of burglary
and house fire?

4.5.8 Risk pooling in detail

Risk pooling is based on the fundamental idea that the larger the group exposed
to a low probability risk, the more predictable the risk becomes for the group as a
whole. In fact, the results for the group become closer and closer to the expected
value of the risk. This is called the “Law of Large Numbers”.
4.5 Risk management 95

The Law of Large Numbers

This law is actually a statistical theorem. We won’t prove it here, but let’s see if it
intuitively makes sense. The law states that as the number of times you perform an
experiment increases, the average of the outcomes becomes closer to the expected
value of a single outcome.
Let’s consider tossing a fair coin as an example. We will play a game where you
win R1 if the coin lands on heads. If it comes up tails you win nothing. What’s the
expected value of the game?
? Is this a speculative or pure
EV = P(heads) × R1 + P(tails) × R0 = R0.50. risk?

So the expected value is 50 cents.

So what happens if you play this game many times? Simulations show the following
results:
U The simulations were gen-
erated using wolframalpha.
1.0 com, which is a website that
can calculate just about any-
0.8 thing. The request used to
generate this graph was
“Expected value of X if X is
0.6 binomial with n=1 and p =
0.5”.
0.4

0.2

0.0
200 400 600 800 1000

Figure 4.1 Simulated mean for three different fair coins.

The above graph shows the results of 3 coins being tossed 1 000 times. After each
throw, the result is recorded and the average is calculated. At the beginning, the
average varies quite a lot. But when the number of tosses is in the hundreds, the
average stabilises around R0.50, which is the expected value.
This is a very useful result. It means that while it is very difficult to predict what will
happen to any individual person or small group (the variability is just too high),
once we are dealing with groups of tens or hundreds of thousands, it becomes
much easier to predict the number of events that such a group will experience.
So for example, it would be very difficult to predict how many of your friends will
be involved in a car accident this year. But if we have a group of a ten thousand
drivers, the number of accidents in that group becomes quite predictable—and not
that different from year to year. This property of large groups—the predictability
of events for large groups—is at the core of risk pooling.
96 Chapter 4 Managing Risks

Example 4.6
The concept of risk pooling can be employed to reduce the impact of risks on a
group. Let’s consider a fishing community. In this community of 1 000 families,
every family owns a small fishing boat which is the sole source of their livelihood.
If a boat sinks, as occasionally happens, it costs R200 000 to replace. A fisherman
makes that much in a year, and almost all of it is needed to cover costs of living.
If one family’s boat sinks, it’s a huge disaster—a disaster that they are unlikely to
recover from.
The estimated probability of a boat sinking in any particular year is 0.002. That
? How do you think we may means that we expect 2 boats out of the 1000 to sink per year on average (this will
have estimated this probabil- of course vary from year to year, but on average over the years we expect it to be 2
ity? boats).
The community has decided that they should help each other by pooling the risk.
They can do this in different ways, but let’s say that every time a boat sinks, every
family contributes 1/1000th of the cost of a new boat, i.e. R200 000/1000 = R200.
This means that if the expected number of boats has to be replaced, each family
will have to give R400 to the risk pool each year.
Note that the contribution is not guaranteed to be R400—it may be R600 or R800
or more in some years, and R200 or nothing in some other good years. But still, the
cost per family will be much more predictable than without risk pooling.
As an example consider the Smit family whose boat sinks in year 6. The cashflows
that they would have experienced if they did not pool the risk are:

Yr1 Yr2 Yr3 Yr4 Yr5 Yr6 Yr7 ...

0 0 0 0 0 200 000 0 ...

The cashflows that they actually experienced with risk pooling were:

Yr1 Yr2 Yr3 Yr4 Yr5 Yr6 Yr7 ...

400 200 400 600 0 200 800 ...

? Can you tell how many boats The costs under risk pooling are more predictable. On the other hand, without risk
sank in each year from this pooling there were many years where there were no costs at all, but the Smits carry
table? the risk that their boat sinks in the future and they are financially ruined. Under
risk pooling, the Smits are almost always helping someone out, but when their
boat sinks the other families help them out.

Case study 2 cont.

Let’s apply the principle of risk pooling to the case of Natasha, the Constantia
home owner and cat lady. In the previous chapter we calculated that the expected
value of the risk of burglary for Natasha was R2 706.86, and the risk of fire had an
4.5 Risk management 97

expected value of R39.15. Both of these figures refer to a risk of the event happening
over 1 year.
Let’s say that Natasha joins a risk pool of home owners with similar properties.
If the risk pool operates the same way as the fishing boat one, members would
be required to pay in every time that someone suffers a loss; for example, if Bob’s
house in Newlands burns down, and the damage is R5 000 000—all risk pool
members would need to be contacted to contribute a portion of the R5 000 000 as
soon as possible.

Example 4.7
What do you think is the problem with this kind of scheme, where you are asked to
pay in after someone has already experienced a loss?
Solution: What if people don’t pay up? People could belong to the scheme “for
free”, and if they had incurred a loss, they would have been covered by the pool—
but if someone else claims, they could resign from the pool or just refuse to pay.
That is why most such schemes are payable in advance, before any losses are
incurred.

What is the problem with paying in advance?

Solution: If payments are required in advance, it is not known what amount will
be required. The number and size of the claims will vary. Predicting claims in
advance is bound to lead to some mismatch between the contributions and the
claims.

Case study 2 cont.

Let’s say that Natasha’s risk pool decides to charge the contribution in advance, at
the start of each year. That means that someone (likely an actuary!) will be asked
to work out how much is needed in contributions, i.e. what is the expected cost
of claims. The actuary will gather data about fires and burglaries (hopefully from
a better source than a newsletter!), and work out the uncertainty and the value
of potential claims from all the people in the risk pool. This will allow them to
calculate the expected value of all the claims. The contributions should then be
set to be such that the contributions cover all the expected claims.

Of course, you cannot predict the exact number and size of claims, since
predictions are just good guesses. If the pool is very large, and the research was
good, the guess should be a good one, but there will still be a difference between
claims and contributions.
If the prediction was too pessimistic, contributions will be higher than claims.
That is not a big problem - the difference can be put into an account for the next
year.
98 Chapter 4 Managing Risks

If, however, we underestimated claims and they are much higher than con-
tributions, the final responsibility over the risk is still with the members of the
pool. So if more money is needed, those pool members would need to be ap-
proached for a special contribution. This is a feature of risk pooling: the risk is
not transferred away from the pool members. It is only shared between them.

Case study 2 cont.

How much of the contribution should each pool member pay?

Solution: This is an interesting question. Say the total expected value of claims
for the pool is R15 million, and there are 6 000 people in the pool. One easy way
to calculate the contribution would be to divide the R15 million needed evenly
between all the members—everyone pays R2 500.
But is that fair? Natasha has a R9 million house. Steve, who lives in Kenilworth,
has a R2 million house. If they are burgled or experience a fire, the likely loss
incurred by these two will be very different. It may be more fair to apportion the
contribution based on some other characteristic; for example, the property value.
The total value of all the properties in the pool is R18 billion. If we are to apportion
the R15 million contribution required by property value, then Natasha will need
to pay 9 000 000/18 000 000 000 × R15 000 000 = R7 500, and Steve will only have to
pay 2 000 000/18 000 000 000 × R15 000 000 = R1 667. This makes sense, as Natasha
is exposed to more risk than Steve, so it is fair that she pays more.
T We are assuming that the
chances of burglary and This approach of using some sort of measure to standardise premiums is called the
fire are the same for all measure of exposure. The measure of exposure is something that can be used to
participants—this may not standardise premiums because it is proportional to risk. In the example above, the
be the case. We will consider measure of exposure is “property value”. Using property value, we can standardise
more complex cases when the premium to R833.33 per R1 000 000 of property. Then Natasha can work out
we discuss insurance later. her premium just by multiplying the standardised premium by the number of
millions of Rands of property she owns: 9 × R833.33 = R7 500.

Exercises for 4.5.8 Risk pooling in detail

Ex.4.10 Consider the other case study: can you think of a way that Sizwe may
manage his risk of death in Afghanistan by using risk pooling?
Chapter 5

Insurance Principles

5.1 What is Insurance?

In Chapter 4, we discussed ways of managing risk, and in particular the concepts
of risk pooling and risk transfer. For example, Natasha and her housing risk pool
were able to share the risks of fire and burglary, but were still exposed to the risk
in total—a year with a high number of incidents would mean that the cost to
everyone in the pool would increase. In order to avoid this, the risk pool could
transfer the risk to a third party. This combination of risk pooling and risk transfer
is the basis of insurance.
Insurance developed as a way to pool risks and transfer them to some entity.
It has become a very formal arrangement in most countries, with its own laws
and terminology. Insurance may generally be described as:

A contract binding a party to indemnify another against a specified

loss/es in return for premiums paid.
We look at each of these terms below.

T Contract: Insurance is conducted by means of a legal contract.

The contract specifies the rights and obligations of each party.
In an insurance contract the insurer promises to pay a specified
benefit if a specified event occurs and the insured promises to
pay regular premiums. This contract is called “the policy”, and
the insured is referred to as the “policyholder”.
T Indemnify: The concept of indemnity is important in insur-
ance. To indemnify means “to put back in the original position”.
So most insurance policies aim to pay enough to get the policy-
holder back to the position they would be in if the insured event
had not happened. For example, after a burglary, the insurer
aims to pay out enough for you to be able to buy the same TV,
computer and collection of DVDs as the ones that were stolen
from you. Indemnity is not always possible; for example, if the

99
100 Chapter 5 Insurance Principles

insured event is a death, the insurance company can only pay

out a sum of money, which clearly does not put you back in the
original position. But the principle of indemnity says that the
amount paid out should at least be related to the amount lost,
so for a death, the amount insured may be in line with the lost
future earnings of the person who died.
T Specified loss: Every insurance policy will specify exactly which
risks are and are not covered by the policy. Many insurance
policies cover a set of related losses. For example, car insurance
often covers the car against fire, theft, and damage to third
parties (other cars, pedestrians, etc.).
T Premiums paid: Insurance policies usually require a regular
? Why are the premiums paid premium to be paid to the insurer monthly in advance. The pre-
in advance?
mium is calculated based on the expected value of all the events
that are specified in the policy, plus allowance for the insurer’s
costs and profits. Insurance is basically a system for exchanging
small, predictable, regular premiums for large, unpredictable
losses.

Exercises for 5.1 What is Insurance?

Ex.5.1 What could someone do to indemnify you from the follow-

ing? Think about different ways that you could be put back
into the same state as if the incident had never happened.
1. The loss of your laptop;
2. Your TV being damaged by electrical surge;
3. Losing your lunch money;
4. Having to cancel an interview with a recruitment com-
pany;
5. Being in hospital for 4 weeks and unable to work as a
result.

5.2 Why Use Insurance?

Insurance is a powerful concept that appeals to people for many

different reasons, including:
5.2 Why Use Insurance? 101

• risk aversion • better use of capital

• risk pooling • smoothing of cashflows
• economies of scale • social benefits.
• protection from unacceptable
risks

We look at each of these in some detail below.

5.2.1 Risk Aversion

Risk aversion is a common (but not universal) characteristic of hu-

man beings, and describes a general preference for certainty over
uncertainty. Let’s illustrate this by way of an example.
Suppose that you are offered a choice between a guaranteed R100
or an uncertain outcome dependent on the toss of a coin, if the coin
lands on heads, you get R200, but if it comes up tails, you have to pay
R100. The expected value of the outcome based on the coin toss is

EV(coin toss) = 0.5 × R200 + 0.5 × (−R100) = R50.

Remember, we can think of this expected value as the long term

average, or in other words, if we toss the coin an infinite number of
times, we would expect it to come up heads as often as it comes up
tails, which means that the average gain to you would be R50. So you
are faced with the choice between a guaranteed gain of R100 or an
uncertain gamble with an expected gain of R50. You would be very
likely to prefer the guaranteed outcome, and this is consistent with
risk averse behaviour. If you do prefer the uncertain outcome, you
show risk-seeking behaviour, which is the opposite of risk aversion,
and you are the type of person much loved by casinos!
Now let’s suppose that we change the terms of the gamble to ensure
? Can you show that the ex-
pected value of this game is
that it has the same expected value as the certain gain, so we will R100?
increase the gain on heads to R300 while keeping the loss for tails at
R100.
Now which would you prefer? If you are like most people and risk-
averse, you will still prefer the certainty. Now what if we increased the
gain on heads to R350, such that the expected value of the gamble is
R125, i.e. 25% higher than the certain gain?
Depending on your degree of risk aversion, you may still prefer the
certain gain because you may be uncomfortable with the possibility
of having to pay out R100 if the coin comes up tails.
102 Chapter 5 Insurance Principles

How does this translate to insurance?

Let’s suppose that you have a family who are dependent on the in-
come you earn, and you know that there is a 0.05 probability of you
dying in the next 20 years, which would leave them destitute. You can
choose to pay a regular monthly premium to a life insurer. If you do
this and then you pass away, the insurer will pay enough money to
your family to replace the income you would have earned. Alterna-
tively, you can choose to take the gamble that you will survive (after
all, there is a 0.95 probability of this). Most people would prefer the
certainty. This creates the opportunity for life insurance.

Exercises for 5.2.1 Risk Aversion

Ex.5.2 Think of how risk aversion gives rise to general insurance

(e.g. insurance cover for your car and house) and medical
schemes.

5.2.2 Risk Pooling

You are already familiar with the concept of risk pooling. Insurers
have tens or even hundreds of thousands of policyholders. This al-
lows the insurers to take full advantage of the law of large numbers;
the number of policyholders is so large that insurers can expect a
stable and relatively predictable claim pattern. This means that the
aggregate outgoing claims cashflows paid by the insurer do not fluc-
tuate wildly over time, reducing the risk for the insurer.
For policyholders, the large size of the risk pool they are joining is also
a positive—they can have fewer doubts about the financial security
of their insurer. Given how important the insurance claim can be for
people—in many cases, it is literally a life saver—the last thing they
want is to find out that their insurer is having financial trouble due to
unpredictable claim fluctuations.

5.2.3 Economies of Scale

The concept of economies of scale is related to the ability of larger

institutions to be more cost effective than small institutions. Clearly
it is possible for an institution such as a life insurer to develop spe-
cialist expertise in mortality, reserving, capital management and the
administration of premiums and claims in a way which would not
be possible for individuals or small groups. This expertise allows the
institution to keep the average costs down to a level only achievable
by operating on a large scale.
5.2 Why Use Insurance? 103

Example 5.1
Say that an insurer with fewer than 5 000 policyholders needs 1 actuary.
An insurer with 5 001–20 000 policyholders needs 2 actuaries, and
an insurer with 20 001–100 000 policyholders needs 3 actuaries. One
actuary is paid R1 million per annum.
Calculate the per-policy-cost of actuarial services for an insurer with
(a) 2 500 policyholders,
(b) 10 000 policyholders,
(c) 19 000 policyholders, and
(d) 80 000 policyholders.

Solution:
(a) This insurer needs 1 actuary, so the expense is R1 000 000×1/2 500 =
R400 per policy per year.
(b) This insurer needs 2 actuaries, so the expense is R1 000 000 ×
2/10 000 = R200 per policy per year.
(c) This insurer needs 2 actuaries, so the expense is R1 000 000 ×
2/19 000 = R105.26 per policy per year.
(d) This insurer needs 3 actuaries, so the expense is R1 000 000 ×
3/80 000 = R37.50 per policy per year.

Exercises for 5.2.3 Economies of Scale

Ex.5.3 1. For Example 5.1, calculate the per-policy-cost of actu-

arial services for an insurer with 5 000 policies and for
another insurer with 6 000 policies.
2. The trend in Example 5.1 was for the per-policy-cost
to decrease with increasing numbers of policies—but
what has happened in this case?
3. Use Excel to plot the per-policy cost of actuarial ser-
vices in the above example against the number of pol-
icyholders an insurer has . What do you notice?

5.2.4 Protection from unacceptable risks

Some individuals can manage some of their risks without insurance.

A key factor is whether the individual has some spare financial re-
sources, and whether the risk is relatively limited in size. For example,
someone with a high level of savings may not choose to take out
insurance for household contents as they feel they can fund the re-
placement of these items themselves. But for people with a low level
of financial resources, and for risks which have a very high value,
insurance may be the only protection from financial disaster.
104 Chapter 5 Insurance Principles

5.2.5 Better use for capital

Even individuals who may have spare capital may not want to keep it
aside “just in case” a risky event happens. They may want to invest
their capital, and this may make it difficult to access for some time—
in that case, they could not use it as a source of funds to replace stolen
goods, for example.

Exercises for 5.2.5 Better use for capital

Ex.5.4 Consider someone who has saved R60 000 as protection

against becoming unemployed. Suppose he is earning 3%
p.a. on his money, which he is likely to be keeping in a low
interest account because it needs to be easily accessible.
(You will learn in Module 4 that higher interest investments
are often less liquid, i.e. they are less easily accessible to
investors.) If he did not have to keep his savings for un-
employment in an easily accessible account, he would be
able to invest it in a higher interest account or another
investment portfolio.
Let’s say that he finds an unemployment insurance prod-
uct which costs R100 per month, in advance, and provides
him with R10 000 at the end of any month that he is unem-
ployed. Now he no longer needs to keep his savings in a
low interest account and can invest the R60 000 which he
keeps in his unemployment savings account at the much
higher rate of 18% p.a.
1. Calculate the accumulated value of R60 000 after 1
year at 3% p.a. and at 18% p.a.
2. Calculate the accumulated value after one year of
R100 per month, paid in arrears and for one year. Use
an interest rate of 18% p.a.
3. If he took out the unemployment insurance policy,
what would his financial position be at the end of the
year given he was employed at all times during the
year? Compare this to his financial position at the end
of the year if he did not take the unemployment insur-
ance product and was employed at all times during
the year?
4. Is the unemployment insurance product worthwhile
for him?
5.2 Why Use Insurance? 105

5.2.6 Smoothing of cashflows

Insurance offers a smooth and predictable capital outgo, compared

to the possible fluctuations caused by uninsured risks. Companies
(and individuals) often prefer that smooth outgo, which makes their
operations more predictable and their accounting practices easier.
Example 4.6 illustrated this smoothing of cashflows.

5.2.7 Social perspective

We can see how insurance adds value to individuals, and there is

obvious value to the shareholders of insurance companies in that
they expect their companies to generate profits for the risks that they
take. But does insurance also have a value to society as a whole?
After all, we might in economic terms deem insurance a merit good,
meaning that society judges that all citizens should have some access
to it on the basis of need rather than ability to pay.
The answer, broadly, is yes. It enables economic progress in that it
allows individuals to take economic risks knowing that their depen-
dants, their incomes and their property are protected. Through this
economic progress, society as a whole is able to progress: economic
growth means roads, power, running water and other social goods, as
well as more jobs in general. Insurance also plays a “safety net” role
which helps to alleviate poverty and dependence on the state (as does
private retirement provision through pension funds). This is even
more the case when the insurance is provided on a compulsory basis
for all citizens; for example, through a social security system. Under
these circumstances, it can play a redistributive role, which is broadly
acknowledged as being in the overall social interest and in a very un-
equal society like South Africa, characterised by past discrimination
and inequity, is essential to future social stability.
? Why do you think that in a
There are some counterbalancing forces, however. There are eco-
market with only one or two
nomic costs to running an insurance company, in the form of redi- insurers the cost of insur-
recting resources away from other productive purposes. If a few large ance may be high?
insurers dominate the market, competition may fail to force prices
down to a socially optimal level, which implies a loss to consumers
and overall economic inefficiency.
If insurers are not properly managed, they may become unable to
meet their liabilities, and the consequences of such large-scale failure
can be devastating; this is one of the primary reasons that actuaries
must be experts in risk and capital management.
106 Chapter 5 Insurance Principles

And finally, the advent of insurance can change people’s behaviour

in unintended ways. For example, the knowledge that your car is
insured against theft may make you less careful about where you
T This change in behaviour is park it and whether you lock your doors. Such behavioural changes
referred to as morale hazard.
can in fact increase the overall losses.
It is the risk that a person
will not act to prevent risky
events because they have
insurance. 5.3 Insurable Risks
Another term which can be
confused with this one is In Chapter 4, we talked about the characteristics of risks (fundamen-
moral hazard. Moral hazard tal, particular, speculative, pure, dynamic and static). The concept
is the risk that people who of insurance is to pool risks and transfer them to a third party. We
have insurance will try to need to ask ourselves whether every kind of risk would be suitable
get a payout by doing some- for insurance.
thing unethical, e.g. lying
or causing damage to their
possessions on purpose. Example 5.2 Consider the following risks. If you were an in-
surer, would you sell insurance for each of these risks? If not,
can you explain why? If yes, can you think of risks that you may
be facing as an insurer?
(a) Insure a gambler in a casino against loss;
(b) Insure people against the effects of a meteor hitting the
earth and wiping out most of humanity;
(c) Insure people against bad moods;
(d) Insure people against taxation;
(e) Insure people against unemployment;
(f) Insure people against a new type of flu that has just ap-
peared;
(g) Allow people to buy insurance contract on the life of the
Pope.

Clearly, some risks are not suitable for an insurer. Here are some
guidelines for what risks are insurable:

• It has to be a risk;
• Ideally, the risk should be a pure risk and not a speculative risk;
• Ideally, the risk should be a static risk and not a dynamic risk;
• Ideally, the risk should be a particular (not fundamental) risk,
i.e. an independent risk;
• It should be financial, quantifiable, and be limited;
• It should have a small probability of happening;
• There should be a large pool of risks;
• There should be low moral hazard;
• Past data should be available.
5.3 Insurable Risks 107

5.3.1 It has to be a risk

Risk pooling and risk transfer obviously only work if the event in
question is a risk, i.e. it has a component of uncertainty and finan-
cial consequence. So, for example, paying taxes is not a risk—it’s a
certainty (SARS will always find you!).
? Which of the situations in Ex-
ample 5.2 are not risks? Can
Example 5.3 you think of other examples?
So what would happen if someone wanted to take out insurance
against tax?

Solution: If an insurer tried to calculate the expected value of the tax

payment, they would have to use a probability of 1 that a tax payment
would occur. If there are 10 000 people wanting this insurance, and
each of them is expecting to pay R20 000 in tax, the total claims from
the insurer will be R20 000 × 10 000 = R200 million. This claim would
need to be split between 10 000 policyholders, meaning each of them
pays a contribution of R20 000—the same amount they would have
paid if they were not insured.

5.3.2 Ideally, the risk should be a pure risk and not a specu-
lative risk

Speculative risks involve at least one outcome that is favourable.

These risks include gambling, playing the stock market, applying
for new jobs, and such. Why would an insurer not want to insure
people against such risks? And conversely, would people want to
be insured against such risks?
Because speculative risks can have positive outcomes, people may
seek out those risks. Nobody forces a gambler to go to a casino, she
goes because she wants to win. From an insurer’s perspective, people
seeking out and exposing themselves to a risk is not good business—
it means that the insurer will be exposed to unnecessary risk that
could be avoided. People would become more risk-seeking if they
knew they would not suffer from the adverse consequences of their
gambling, and gamble more—taking on bigger risks. Insuring such a
risk would definitely not be in the interest of the insurer.
? Which of the situations in
In any case, people may not take out such insurance, because it will
Example 5.2 are speculative
necessarily reduce their wins. Insurance has a cost attached to it, risks? Can any of these be
and this cost will be payable even if the policyholder has a positive insured?
outcome. Given the nature of gambling, it can be shown that any
gambling insurance in particular would cost more than the potential
108 Chapter 5 Insurance Principles

wins in any game of chance—this would mean that gamblers are not
in favour of gambling insurance.
Pure risks, such as death or theft or fire, are much more suited to
insurance because the interests of the insurer and the insured are
aligned—the insured should try and avoid the “bad” event and the
insurer would also prefer that there were as few claims as possible.

5.3.3 Ideally, the risk should be a static risk and not a dy-
namic risk

Static risks have uncertainty and value that does not fluctuate over
time. Dynamic risks change with the economy, technology and other
factors. Why would an insurer struggle to insure a very dynamic risk?
Consider the risk of unemployment. During times of economic up-
swing, unemployment levels reduce as jobs are created; but during a
depression, unemployment levels rise. Changes in technology also af-
Insuring against the odds fect unemployment levels, as manual jobs become more automated,
In this section, we are outlining for example. An insurer who wants to offer unemployment insurance
the ideal criteria for insurance. would need to be able to predict the probability of a policyholder
You will find that insurers overrule becoming unemployed. How would you do that?
these criteria on a regular basis;
for example, it is possible to buy Normally, an insurer uses data about the past and adjusts it for the
insurance policies that pay off future. Let’s say that over the last year, the probability of losing a job
your bond when you become un- and becoming unemployed was 0.05. But during a depression, this
employed, even though unemploy-
ment is a dynamic risk. Actuaries
chance may increase to 0.25, while during an economic upswing it
can be very creative in designing may be as low as 0.01. How can the insurer find the right probability
products that overcome the prob- to value this risk of unemployment? Should it be 0.25, 0.05, 0.01 or
lems listed here. Insurers may also another number?
be willing to sell a product that
may generate volatility and loss Recall that dynamic and static risks are not very clearly distinct from
sometimes; for example, if it is each other—most risks classified as static actually do change over
bundled together with a product
time. A better rule may be: the more dynamic a risk, the more difficult
that is making good profits.
it is to insure!

Table 5.1 Insuring against the

odds. 5.3.4 Ideally, the risk should be a particular risk not a fun-
damental risk

Fundamental risks are risks which tend to affect many people at

the same time. Earthquakes, tsunamis and wars are examples of
fundamental risks.
? Which of the situations in Ex-
The problem with fundamental risks is that they violate the law of
ample 5.2 are fundamental
risks? Can any of these fun- large numbers, because the claims from those risks will not be inde-
damental risks be reinsured? pendent from one another. If there is a war, many people will die,
be injured, and have their possessions damaged because of the war.
5.3 Insurable Risks 109

The insurer will be inundated with claims, and the cost of claims will
move far away from the average no matter how large the insurer is.
The concept of risk pooling will no longer work. This is why events
like war and flood are often excluded from policies. They are called Insurance for insurers

catastrophes and they result in a clustering of claims. Reinsurance is insurance for insur-
ers. A reinsurance company has as
A way for an insurer to deal with claims resulting from catastrophes its policyholders other insurance
is to use reinsurance. companies. This means that the
total size of the risk pool for rein-
surers is much bigger and more
5.3.5 Financial, quantifiable and limited diverse than for any one insurer. A
war may be a catastrophe for an
An insurer must be able to pay out a claim—therefore, the conse- insurer in one country, but not for
a reinsurer who operates across
quence of the risk must be quantifiable in financial terms. The in-
many countries. For them, the war
surer also needs to be careful when accepting risks that could have will only affect a small percentage
claims that are unpredictably high, as one claim could have a huge of their business, and may not be
effect on the financial stability of the insurer. considered fundamental. You will
learn more about reinsurance in
The insurer can ensure that the claim size is limited by putting an Module 3.
upper limit on the claim size in the policy. An example of this is
insurance against litigation—these kinds of policies will often have Table 5.2 Insurance for insurers.
an upper limit of cover, say R5 million, to make the risk of litigation
? Which of the risks in Exam-
insurable. This is because legal fees and the costs of a law suit could ple 5.2 do not have quan-
quickly become very large. For example, with such a limit on the tifiable financial conse-
cover, if the court decides that you are liable to pay R6 million in quences?
damages, the policy will only pay you R5 million—the rest you will
have to pay yourself.

5.3.6 Small probability

In order for insurance to work, the risks should have a small probabil-
ity, a few percent or less, of occurring.

Example 5.4
Why can’t we insure a high probability risk?

Solution: You would think that high probability risks are more in need
of insurance than low probability risks. Let’s take the probability of
not getting 70% for Maths and Statistics in first year actuarial studies
at UCT as an example. We know that the probability of not getting this
result is quite high. Surely, someone should be selling you insurance
against this risk?
Let’s say that the cost associated with the risk is R30 000 (the cost of
adding one year of study to your degree), and the probability of not
getting the result is 0.50. That means that if there is no insurance, we
expect 125 out of 250 students will be paying R30 000 and 125 will
110 Chapter 5 Insurance Principles

be paying nothing. If there is insurance, everyone will be paying a

premium of at least R15 000.
Would you buy this policy? Most people would think that the premium
is just too high compared to the risk and prefer to take their chances!

5.3.7 Large pool

? In general, the smaller the By now this should be an obvious requirement—the concept of risk
probability of the risky event,
pooling and the law of large numbers only work if there is a large
the larger the pool needs to
be for risk pooling to work.
number of risks.
Can you think of why this is?
5.3.8 Limited opportunity for anti-selection and moral haz-
ard

What is anti-selection?
The concept of “anti-selection” is key to insurance. It relates to infor-
mation: when someone has more information about the risk than the
insurer, they can take advantage of the insurer. For example, say some-
one finds out that they have terminal cancer. They could approach an
insurer and buy a life insurance policy without disclosing their cancer
status (that’s the information imbalance: they know something that is
relevant to the policy, and the insurer does not). This would give them
a guaranteed payout after only a few premiums have been paid—good
for them, but what about the insurer?
If the insurer sells a product where such behaviour is possible, what
will happen is that people with terminal diseases will be attracted
to buy the policy. As a result the insurer will experience a higher
number of claims compared to what they expected. The insurer will
compensate by increasing premiums. This will make healthy individ-
uals go to other, cheaper insurers, leaving the sick individuals in the
pool, increasing claims further, and so forth until the pool becomes
unsustainable. This is sometimes called “the insurance spiral”.

Exercises for 5.3.8 Limited opportunity for anti-selection and moral

hazard

Ex.5.5 Consider example 5.4 again - how would anti-selection

make the premiums more expensive for this scenario?

What is moral hazard?

5.3 Insurable Risks 111

Moral hazard refers to unethical behaviour on the part of a poli-

cyholder. So a policyholder may for example try to claim a higher
amount than they are due by pretending that more was stolen from
them than actually was. Or a policyholder may damage their own
property (e.g. burn down their own business) in order to get the in-
surance money. Other examples of moral hazard are people feigning
disability in order to claim on a disability policy.

What can insurers do to protect themselves from moral hazard and

anti-selection?
U The “small print” of a pol-
Much of the “small print” in policies has to do with making sure that icy is also referred to as the
policyholders cannot take advantage of the insurer. For example, terms and conditions of the
certain causes of death may be excluded from a life policy, or there policy.
may be a period during which benefits as a result of certain risks
are not payable. The policyholder may be required to disclose any
pre-existing conditions, and the insurer will not pay if it finds out that
the policyholder failed to disclose a pre-existing condition.
An insurer can also reduce the risk of anti-selection by underwrit-
ing the risks. Underwriting usually involves the insurer asking the
prospective policyholder questions to try and establish the size and
nature of the risk. For example, a life insurer may ask a prospective
policyholder about their health to try and establish when the prospec- ? If a life insurer asks about a
tive policyholder is likely to die. Sometimes for large and abnormal prospective policyholder’s
health to establish the risk,
risks the underwriting process can be quite invasive; for example, a
what do you think a general
life insurer may require that that the prospective policyholder, who insurer selling motor insur-
wants a very large sum assured, have a medical check-up or even ance asks about to establish
visit a medical specialist. But some types of business are still subject the risk they face?
to anti-selection. This is not a problem if the premium charged can
reflect this, i.e. sometimes the only solution is to assume a certain
amount of anti-selection will take place and then set the premium
accordingly.

Exercises for 5.3.8 Limited opportunity for anti-selection and moral

hazard

Ex.5.6 Figure 5.1 is a screen capture from the on-line insurance

quotation system [Link]. The particular screen
capture is the form for a life insurance product.
1. Why does the form ask about the prospective policy-
holder’s
2. date of birth,
3. marital status,
4. smoking status,
112 Chapter 5 Insurance Principles

5. occupation,
6. income, and
7. education level?
8. What do you think would happen if you put 1902 as
your year of birth?

Figure 5.1 Screen capture from [Link].

5.3.9 Availability of past data

? Which of the risks in Exam- Even if a risk is otherwise suitable for insurance, if it is a new type of
ple 5.2 are likely to have lim-
risk, or a new type of policy, data may not be available to work out
ited past data?
the uncertainty and value of the risk. In those cases, it is difficult to
launch such a product. Insurers may wait for a few years and gather
information, or they may launch with very inaccurate “guesses” of
what the premium should be, and adjust after a few years.
? Do you think that this is an
Insurers need to be able to absorb the potential losses in such cases
ethical practice? Can you
think of arguments in favour where information is not available. One way to do this is to make the
and against such pricing? initial premium high.

Exercises for 5.3 Insurable Risks

Ex.5.7 Consider the list below, and comment if each risk is insur-
able, and if not, why not?
1. Risk of a house becoming haunted;
2. Risk of damage due to rain;
3. Risk of a business earning lower than expected profits;

4. Risk of damage from wear and tear to a car.

Chapter 6

Insurance Provision

6.1 Types of Insurance

Insurance has developed in response to the most basic financial risks:

• the risk of losing a family member and his/her contribution to

family income,
• the risk of loss or damage to property and possessions,
• the risk of being unable to work and as a result having no money
to support oneself,
• the risk of illness and the costs associated with medical care.

Different types of insurance have been developed to help people

mitigate these risks.

6.1.1 Life insurance

" Life insurance is also called
Life insurance policies all concern the event of death. Some policies life assurance. Assurance is
pay out a sum on death—supporting surviving dependents of the a specific term that relates
insured person financially. But life insurance companies also sell only to the life sector of the
other types of policies which are based around the event of death; for insurance market.
example, policies which pay out a certain amount if you survive to
the end of a term.

Exercises for 6.1.1 Life insurance

Ex.6.1 Why would anyone want a policy that only pays out on
survival for a given term?
Ex.6.2 For a twenty something year old like you, which policy do
you think would have a higher premium: one that pays

113
114 Chapter 6 Insurance Provision

R1 000 000 if you die in the next 10 years, or one that pays
R1 000 000 if you survive to the end of the next 10 years?
Think about the probability of each event, and the resulting
expected value.
Ex.6.3 Life assurance policies may also differ because they only
cover specific causes of death, like death as a result of an
accident only. Why would someone want accidental death
insurance?

6.1.2 Disability insurance

This type of insurance protects policyholders from the risk of becom-

ing disabled.

Example 6.1
Which of the four basic financial risks above does disability fall under?

Solution: The main risk of disability is the inability to work and pro-
vide for yourself. If you thought of illness, that is usually covered under
a separate special policy with a medical scheme. So someone who
gets into an accident, has medical treatment to help him recover, but
ends up unable to work, would claim twice: once from the medical
aid for medical expenses, and a second time from his disability policy
for loss of income as a result of the disability.

Disability insurance often pays out a benefit which is a monthly

annuity, like a salary. But in South Africa it is also possible to buy a
lump sum disability policy.
In most countries, there are no “disability insurance” companies—
these products are sold by life insurers.

6.1.3 General insurance

? Why would a company incur
As the name suggests, general insurance, which is also called “short-
a financial loss from workers
striking? term insurance”, insures a wide range of things, from houses to cars,
to pets, to ships and aeroplanes, to liability cover for companies
selling faulty products, and to companies for the risk workers going
on strike.
General insurance is short term. The policy usually has a one year
term and then has to be renewed. This gives general insurers an
advantage in their pricing as they can revise premiums at the end of
policy term.
6.1 Types of Insurance 115

Exercises for 6.1.3 General insurance

Ex.6.4 Comparing life assurance and general insurance, which do

you think has risks that are more dynamic? Do you think
that the general insurer’s ability to change premiums at
the end of the policy term helps them in this respect?

6.1.4 Pension funds

? Which of the 4 basic finan-
Pension funds are not insurers, but they fulfil a similar role. Their
cial risks is covered by pen-
function is to help people save for retirement. This involves putting sion funds?
aside a regular monthly amount over the course of a person’s working
life, and then paying them a pension for life after their retirement
date.
Funds are generally put in place by employers to help their employees
save, but the government and organisations like trade unions also
provide pensions.
" In South Africa, such pen-
Individuals can also provide pensions for themselves by buying a sion policies sold by life
pension policy from a life insurer. insurers are referred to as
retirement annuities.

6.1.5 Medical aid schemes

? Do you think not being able
Medical aids are insurers who specialise in covering people against
to decline cover and having
the risk of illness. The medical insurance business is very complex to charge everyone the same
because of the number of conditions covered, the varying cost of premium makes medical
medical treatments, and the huge risk of anti-selection. In South aids in South Africa more or
Africa, medical schemes are legally not permitted to refuse cover to less prone to anti-selection?
anyone or to charge one person more than another for a certain level
of cover.

6.1.6 Reinsurance
? Why don’t insurers just trans-
Reinsurers are generally big international organisations that function
fer all their claims and all
like an insurer for insurers. You can find reinsurers operating in any their risk to the reinsurer?
business where there is a risk of catastrophe which will result in very
high claims payable by any one insurer. Insurers protect themselves
from this risk of high claims by transferring that risk to a reinsurer,
who pools it with other insurance companies.
A typical reinsurance policy may state that the insurer may claim
from the reinsurer when their total claims in a year exceeded a certain
amount.
116 Chapter 6 Insurance Provision

Exercises for 6.1.6 Reinsurance

Ex.6.5 For general insurers, life insurers and medical aids, can
you think of situations where the claims in a particular
year would be very high and a reinsurance claim may be
triggered?

6.2 Insurance Providers

Insurance can be facilitated by different providers. The main require-

ments for a provider of insurance are:

• Sufficient capital to be able to meet claims as they arise;

• Sufficient expertise to be able to design and administer insur-
ance policies;
• Distribution channels for insurance sales;
• Willingness and ability to comply with specific regulations per-
taining to insurers. In most countries, insurers would need to
obtain a license from the government to practice.
• Large numbers of clients—insurers cannot operate on a small
scale, because their business is dependent on the law of large
numbers.

The main providers of insurance are insurance companies and the

state.

6.2.1 Insurance companies

These are organisations that sell insurance products. Their main

business is the collection of premiums and the payment of claims.
A feature of insurance companies is that generally their clients buy
policies on a voluntary basis. That is every individual or company
can decide whether to buy a policy, and who to buy it from. So
? Why would a life insurer, the risk of anti-selection can be prevalent in such private insurance
who charges the same pre- companies, as people can choose to participate in a risk pool or not
mium to a policyholder with
based on their circumstances. If an insurer sets their premiums too
a family history of cancer
and to a policyholder with-
high, they can expect people to go somewhere else for their policies.
out such a history, be ex- Alternatively, if a life insurer does not offer different rates to people
pected to attract more peo- with different health statuses, they can expect to attract clients who
ple with a history of cancer? are in poor health, and therefore will claim sooner.
An insurance company can be either a proprietary company or a
mutual company. These terms have to do with the ownership of the
company.
6.2 Insurance Providers 117

A mutual company is owned by its policyholders—it is basically just

a sophisticated version of a risk pool.
When a mutual company makes a profit, the profit is distributed to
the policyholders. The mechanisms for doing so are called “bonuses”. Sanlam and Old Mutual
In mutual companies, one of the roles of actuaries is to calculate Up until 1998–99, Sanlam and
and recommend to the board of directors the amount of bonus that Old Mutual were mutual insurers,
should be paid each year. owned by their policyholders.
Both companies de-mutualised
A proprietary insurance company is just a normal company, and it almost at the same time. This was
is owned by shareholders. Proprietary companies therefore also allow done by issuing shares for each
their policyholders to transfer the risk (and not just pool it). However, person who had a policy at the
time of de-mutualisation. Most
policyholders do not automatically participate in the profits of the of these policyholders sold their
company in this case. Generally, the profits are paid to shareholders shares immediately, while some
as dividends. kept them. Both of the companies
became proprietary companies
from that point forward, and their
Exercises for 6.2.1 Insurance companies shares were listed on the stock
exchange, and can now be bought
Ex.6.6 LifeAssure is a proprietary insurance company specialising and sold by anyone.
in life assurance. Their financial year runs from 1 February
to 31 January. In the year ending on 31 January 2013, they
had the following results: Table 6.1 Sanlam and Old Mutual.

Total premium income R550 million

Total claims paid out R200 million
Total expenses R50 million
The total money needed to be put aside for future claims
is R150 million.
The directors decide to distribute 80% of the profits to
shareholders as dividends. There are 6 million shares in
issue, priced at R285.40 per share.
Using an assumption of a 5% p.a. dividend growth and
a rate of interest of 12.5% p.a., calculate the share price
immediately after this dividend payment is made. Ignore
tax.

6.2.2 The state

The state can also provide benefits which compensate people against
risky events. For example, government can offer:

• Social grants; for example, grants for children, or orphans, dis-

ability grants, widow(er)’s pensions, old age grants, amongst
others;
118 Chapter 6 Insurance Provision

• Contributory pensions in retirement; many countries have na-

tional pension schemes which are mandatory for all workers
and ensure that everyone saves for retirement;
• Public healthcare; this can be offered as either a fund which
pays for medical treatments, or as public healthcare facilities
funded by the state;
• Unemployment benefits; paid to people who lose their jobs to
help them support themselves.

The state has a very different approach to insurance from insurance

companies. This is partly because the objectives of the state are
different: it seldom seeks to indemnify every person against their
losses. It is much more likely that the state is more interested in
helping the poor and providing a minimum benefit level.
The objectives of the state generally include solidarity: this refers to
“supporting each other”. Basically, the state recognises that different
people have different means and needs, and it aims to take contribu-
tions which are calculated according to means, but pay benefits that
are distributed according to needs. The mechanism of solidarity is
cross-subsidisation—those who have resources help those who do
not.

Example 6.2 The state may collect premiums for a benefit, such
as medical care, as a percentage of income, but distribute the
benefits according to actual medical treatment needed. This
will mean that a person earning R1 million p.a. will contribute
10 times as much as a person earning R100 000 p.a., but both
qualify for the same benefits. Another example is a benefit that
is funded from general taxation (meaning that the contributors
are only those individuals that earn enough to pay tax) but paid
only to the very poor (for example, an old age pension payable
only to people without any other source of income).

The state is also different from private companies in that it has the
power to enforce compulsory participation in its schemes. This alle-
viates the risk of anti-selection, and reduces the cost of insurance in
general.

Example 6.3
Why does mandatory participation make insurance provided by the
state cheaper than private insurance?
6.3 The role of actuaries and quants 119

Solution: There are three reasons: lack of anti-selection, economies

of scale, and lack of profit motive.
Consider an example: Say that a dangerous disease that causes death
is prevalent in a country. It is known that the disease is much more
likely to affect people living near the coast, but it has been known to
turn up elsewhere as well. In the country, which has 10 million people,
100 000 cases of deaths from this disease are reported each year. Of
this, 90 000 deaths are in the coastal regions, where only 3.5 million
people live.
The government is considering putting in place a benefit payable
to the dependants of people who die from this disease. The benefit
will be R20 000 per person. It will be mandatory for everyone in the
country to pay for this scheme. Ignoring interest, how much will
everyone have to contribute?
The premium will be R20 000 × 100 000/10 000 000 = R200 per person
per year.
The state would need to also collect a bit more from everyone to cover
the expenses of the scheme—say the cost of running this is R1 000 000
per year. The additional cost per person would be R1 000 000/10 000 000 =
R0.10 per person per year.
The state has no profit motive, so the total premium collected by the
state will be R200.10 per person per year.
What if an insurer tried to launch a policy to provide the same benefit?
The policy would be voluntary—anyone can either buy the policy
or not. We would however expect that the vast majority of policies
would be sold in the coastal regions, where this disease is much more
common and people are more worried about it. In the coastal regions,
the chance of dying from the disease is 90 000/3 500 000 = 2.57%. If
the insurer wanted to charge the same premium to everyone, it would
need to assume that nearly all policyholders will be from the coast,
and so it would need to charge 2.57% × R20 000 = R514 per policy.
The insurer would also need to cover the cost of administration. Let’s
say this is the same as for the government, i.e. R1 000 000 per year. But
not everyone will buy this policy—maybe only 2 million people will
buy it (since it will be voluntary). The cost of administration is then
R1m/2m = R0.50 per policy.
Finally, the insurer is an organisation that needs to make a profit. The ? Why isn’t all insurance pro-
profit loading for an insurer could be 15% of the premium. The total vided by the state if it is so
premium is then (R514 + R0.5) × 1.15 = R591.68 per policy—almost 3 much more cost effective?
times the cost of providing this benefit through the state.

6.3 The role of actuaries and quants

Insurance companies are the main employers of actuaries, and many

quants end up working there as well. Now that you understand the
120 Chapter 6 Insurance Provision

main principles of insurance, we can think about what actuaries and

quants could do for an insurer. These activities include:

• Pricing;
• Underwriting;
• Contract design;
• Reinsurance;
• Prudent reserving/capital management;
• Asset-liability management;
• Monitoring; and
• Expense budgeting

Each of these is considered briefly below.

6.3.1 Pricing

One of the key actuarial functions is to calculate the amount that

clients should pay for the insurance products they buy. This amount
is called the premium. Setting premiums requires actuarial meth-
ods because of the uncertain nature of the insurance business: the
total premiums received from policyholders must cover the claims
policyholders make, as well as the insurer’s expenses and the profits
required by the shareholders. Claims only happen after the policy
is issued, so they must be carefully estimated in order to price the
policies correctly.

Example 6.4 In any business, setting the price for the product
or service you are selling is crucial. If you are selling apples,
the pricing process is quite clear: the price needs to cover the
cost of the apples, plus the expenses of operating the business,
plus your profits. So for example, if you can buy apples at R3 a
kilogram, and it costs you R100 000 per month to operate the
business which sells 50 000 kilograms of apples per month, then
you need to sell apples for at least R5 (R3 + R100 000/50 000)
per kg. If your profit margin is 20%, your final sales price will be
R6 (R5 × 1.2) per kg.

When selling insurance policies, the hard part is to determine what

the policy will cost the insurer to provide — because, unlike the apple
business, these costs only arise after the policy is sold. So the costs
are the future claims the insurer will experience, which actuaries
must estimate by using estimates of claim frequency and value per
6.3 The role of actuaries and quants 121

policy. These estimates will never be exactly correct, and so there

is often a risk margin built into the premium. You can think of this
risk margin as a margin for error. In addition, actuaries monitor what
really happens and compare this to the estimates they made, and
then improve on the estimates in future years to try make sure the
estimates are more accurate.
Setting premiums is a balancing act: too high, and in a competitive
market no-one will buy your policies; too low, and your premiums
will not cover claims and expenses and you will make losses.
We will explore some simple pricing strategies in Chapter 7.

6.3.2 Underwriting

Premium rates will account for a certain level of risk in the insured
population. If the policies that come onto the books are much worse
risks than expected, this will lead to worse-than-anticipated claims
experience, and hence to losses.
In fact, at a given premium rate there is an obvious incentive for
individuals with worse risk than that underlying the premium rate to
sign up for a policy (anti-selection). The process of underwriting is
designed to limit the possibility for anti-selection.
Underwriting involves the assessment of medical, financial and other
relevant information in order to make a judgement as to whether the
applicant should be accepted for cover, and if so, on what terms.

Example 6.5 Life insurers will call for different types of medical
information depending on the applicant’s age, type and amount
of cover requested. For a young applicant with a low sum as-
sured, all that may be required is an HIV test and a cotinine test
(the latter to check whether applicants who declare themselves
as non-smokers are in fact tobacco-free!). An applicant in his
50s requesting R10 million cover would be asked to provide
much more medical evidence, including e.g. electrocardiograph
(ECG) results and liver function tests.

After all the underwriting information has been collected, the insurer
then has the option to:

• accept the proposal on standard terms,

• accept the proposal on special terms (e.g. with a specific exclu-
sion or subject to a loaded premium),
T A "loaded" premium just
• delay the decision, or means an increased pre-
mium!
122 Chapter 6 Insurance Provision

• reject the applicant for cover.

Actuaries are involved in the decisions about which groups to exclude

and what loadings to apply for sub-standard risks.

6.3.3 Contract design

Risk management features may be built into policy contracts directly.

Some examples include:

• excluding a risk which cannot be properly priced (e.g. death

due to invasion by a foreign army),
• transfer of investment risk, e.g. through a unit-linked benefit
design, and
• ensuring that minimal guarantees are offered.

Actuaries are actively involved in product development at insurance

“Long tail” insurance business companies, which includes setting all features of the contracts.
General insurance business which
can result in claims years after
the insured event is referred to as
6.3.4 Reinsurance
“long tail” business. This refers to
the long time after which claims Insurers are in the business of taking risk from individuals, but may
may still arise. themselves wish to pass on some of these risks elsewhere. Reinsurers
An example is employers’ liability are the entities which provide insurance to the insurers. You will learn
insurance where occupational more about reinsurance in a later section of the course, so no further
diseases or injuries can take a long detail will be provided here, safe to say that actuaries are involved in
time to manifest. A famous case
the management of reinsurers as well as in the setting of reinsurance
is exposure to asbestos, which
can lead to a number of ailments strategies at insurers.
and may take decades to mani-
fest. When someone is diagnosed
with an asbestos-related problem 6.3.5 Prudent reserving and capital management
by their doctor, they will need to
figure out whether they were ex- The nature of life insurance contracts is that they are typically long-
posed to asbestos in a previous term. Even general insurers, with one year policies, can still be liable
job. If they were, they may sue
for claims which are only reported years later. Such policies are
the employer. Employers can take
out employers’ liability insurance referred to as "long tail" business.
against having to pay out in such
When insurers evaluate their financial position, they need to put a
law suits, and such insurance busi-
ness is long tailed—the employer
value on these future liabilities. This is where actuaries get involved in
needs to have had the policy in calculating the probability and value of future claims. This process is
place when the exposure to as- known as reserving and is a major area of work for actuaries. Reserv-
bestos happened, not when the ing requires making assumptions about future experience, and on the
lawsuit happened!
basis of these assumptions, discounting future expected cashflows
back to the present (in the way you’ve begun to learn about in this
Table 6.2 “Long tail” insurance
business. course already). Obviously these assumptions should not be more op-
timistic than what is realistically expected, and indeed, it is a general
6.3 The role of actuaries and quants 123

actuarial principle that they should be prudent, i.e. they should err
on the conservative side. Prudence leads to delayed recognition of
profits, and effectively lessens the risk of unpleasant surprises, which ? Can you see why prudent
is a key objective of risk management. reserving would delay the
emergence of profits?
In addition to these prudent reserves, regulators require life insurers
to hold additional capital to protect against experience being sig-
nificantly worse than expected. Insurers will wish to hold sufficient
capital to provide this assurance and to satisfy regulators and poli-
cyholders about their financial strength and solvency, but since this
capital is effectively tied up and cannot be put to work in generating
profits through new business, there is an opportunity cost to holding
too much. Actuaries have a key role to play in the balancing act of
capital management.
Chapter 7 deals with reserving and emergence of profits in more
detail.

6.3.6 Asset-liability management

Asset-liability management is the process of managing the asset port-

folio in such a way that it moves as closely as possible with the liabili-
ties, in order to minimise the risk of assets having a lower value than
liabilities. This is a growing practice area for actuaries and quants.
A simple example of asset-liability management would be an appro-
priate portfolio of bonds backing an annuity book: the reduction of
interest rates in the market will lead to an increase in the value of the ? What would a bond portfolio
liabilities (because future benefit payments must be discounted to matching a book of 10 year
annuities consist of?
the present at a lower rate), but this will be offset by the correspond-
ing increase in the value of the assets.

6.3.7 Monitoring

You will recall that the final stage of the actuarial control cycle consists
of monitoring the experience, and feeding this back into the cycle’s
earlier stages. This is critical to the ongoing survival and success of a
life insurance company, which must conduct regular reviews of its
mortality, morbidity, new business, expense, investment and other
experience. The results of these investigations must feed into pricing
bases, reserving bases, investment strategy and many other areas of
the company’s operations.
For example, by monitoring the experience of a book of policies, the
insurer can decide whether and when some of the assets backing
that book can be released as profits. The emergence of profits is
discussed further in Chapter 7.
124 Chapter 6 Insurance Provision

6.3.8 Expense budgeting

Controlling expenses is an important part of the successful operation

of an insurer, and expense budgeting is an important part of this
(as it is for any company). Actuaries have a key role to play in the
expense budgeting process, through the analysis of past expense ex-
perience and the projection of future experience, taking into account
the company’s business strategies.

6.3.9 Asset Management

You will find that the majority of life insurance companies also have
an asset management division. This develops quite naturally as in-
surance companies are in the business of having a lot of money on
hand—all those reserves that they are holding to meet future claims.
This money must be invested in the market in order to make returns.
So life insurers tend to develop internal expertise in asset manage-
ment:

• Because internal asset managers will be best placed to under-

stand and match the liabilities of the insurer;
• Because using an external asset manager would mean paying a
fee to an external provider when it could be paid to an internal
provider;
? Why would an internal asset
manager still require a fee? • Because the insurer is likely to be of sufficient size to warrant
an internal asset management team (i.e. it is cost effective to
develop internal expertise).

Many quants and actuaries who are interested in investments work

for investment houses which are part of a large insurance company.
Chapter 7

Pricing, Reserving and

Emergence of Profits

In Chapter 6, we discussed the role of actuaries in insurance com-

panies. Here, we will explore the pricing and reserving functions
in more detail and equip you with some simple techniques which
illustrate how pricing and reserving work.
We then see how an insurer monitors the real experience of a product
and compares it with the assumptions that were made about the
product when pricing and reserving. This enables the insurer to
release profit from the product over time.
We focus mainly on general insurance type products in this chapter;
life insurance products are more complex and will be covered in
Module 3.

7.1 Pricing

Pricing is the process of calculating the premiums the policyholder

should pay to the insurer so that the insurer can:

• Pay out all future claims;

• Pay out all future expenses;
• Make some profit.

Sometimes a margin for risk is also built into the premium.

Such a premium, covering all future cashflows of the policy, is called
the “office premium” (OP).
U The office premium is also
often called the “gross pre-
mium”.

125
126 Chapter 7 Pricing, Reserving and Emergence of Profits

Generally speaking,

EPV(Office premiums) = EPV(Claims) + EPV(Expenses)+

EPV(Profits) + EPV(Risk margin).

Or, if there is no risk margin:

EPV(Office Premiums) = EPV(Claims)+EPV(Expenses)+EPV(Profits).

In order to calculate the office premium the policyholders will have to

pay, we will have to calculate the expected values of claims, expenses,
profit, risk margin, and also account for the fact that the premiums
may be payable on a regular basis rather than just once in advance.
Since future claims and expenses are unknown, this pricing calcula-
tion is done using a set of assumptions.

7.1.1 Pricing Assumptions

T A set of assumptions is
The pricing basis is usually fairly realistic. This is because insurers
called a basis; so the pric-
ing assumptions are also compete with each other based on price, so using too much margin
referred to as the pricing for caution in premium calculations will result in high premiums,
basis. which will mean that policyholders will go elsewhere.
The assumptions that need to be made include:

• Claim frequency / probability / number of claims;

• Average claim amounts;
• Future expenses;
? Why does the insurer need to
make an assumption about • Investment returns.
investment returns when
pricing?
Example 7.1 A general insurer has issued a laptop policy which
pays for a new laptop when the policyholder’s existing one is
lost or stolen. The premiums for the policy were calculated
assuming that 4.5% of policyholders will make a claim each
year, and that the average claim amount will be R8 700.
The claim frequency of 4.5% and the claim value of R8 700 are
assumptions — the general insurer does not know that this will
be what actually happens over the year, they can only assume
what will happen.

Once the pricing assumptions are set, the insurer can use these as-
sumptions to calculate the various components of the office pre-
mium.
7.1 Pricing 127

7.1.2 Claims

The most important part of pricing is determining the amount that

will be needed to cover future claims. This is calculated as the ex-
pected present value of claims:

EPV(Claims) = Probability of a claim ×

Present value of the amount of the claim.

Or, for policies where more than one claim is possible,

? Which insurance policies
allow more than one claim?
EPV(Claims) = Expected number of claims×
Expected present value of the average claim.

Example 7.2
Calculate the total expected cost of claims for a pet insurance policy
with the following pricing basis:
• expected number of claims per year: 0.3,
• expected average claim: R840.
Ignore interest for now.

Solution:

EPV(Claims) = Expected number of claims × Expected claim amount

= 0.3 × R840
= R252.00.

Exercises for 7.1 Pricing

Ex.7.1 Calculate the total cost of claims for the laptop policy in
Example 7.1. Ignore interest.

We have been ignoring interest until now, but an accurate evalua-

tion of future cashflows must take into account the time value of
money. Therefore the cost of claims will generally be calculated by
discounting future claims at the assumed rate of interest.

Example 7.3
In Example 7.2 above, recalculate the cost of claims using an interest
rate of 12% per annum.
128 Chapter 7 Pricing, Reserving and Emergence of Profits

Solution: The cost of claims is the expected present value of claims:

EPV(Claims) = 0.3 × 840 × v t

where t is the time when the claim is expected to be paid. We do

not know when this will happen, so we need to make an assumption
about when claims occur. We will assume that they are uniformly
distributed over the year, and therefore they take place in the middle
of the year, i.e. 6 months after the policy is taken out, on average. So
t = 0.5 years, and

EPV(Claims) = 0.3 × 840 × v 0.5 = 238.12

T Note that interest makes very
little difference in this case,
because the policy term is In the example above, we have had to make an assumption about
very short - only one year. the timing of claims. Many, but not all, claims can be assumed to
Interest has a much greater be uniformly distributed over some period of time. In those cases,
impact for policies with a
we can say that we expect the claims to happen in the middle of that
long term, like life insurance
policies. period of time on average.

Exercises for 7.1.2 Claims

Ex.7.2 In each of the scenarios below, state when you would ex-
pect claims to happen, on average:
1. Claims are uniformly distributed over the next 6 months.

2. Claims are uniformly distributed over December and

January
3. Claims are uniformly distributed over the next 5 years.

Example 7.4
Calculate the expected present value of claims for the following poli-
cies, using an effective interest rate of 10% p.a.:
(a) A life insurance policy with a 1 year term for a 40 year old man
has a 0.0023% chance of paying out. The claim amount will be
R1 000 000. The timing of the claim is uniformly distributed over
the year.
(b) A travel insurance policy: the probability of a claim is 0.3% per
month, and the value of the claim is R5 000 on average. Claims
are always settled at the end of the month. The term of the policy
is 12 months.
(c) A car accident policy with an expected annual number of claims
of 0.12 and average expected value of each claim of R20 000.
Claims are expected to be spread evenly over the year, and it is
assumed that they will occur in the middle of each month on
7.1 Pricing 129

average and be settled immediately. The policy has a 1 year term.

(d) The same as for (c), but the claims are expected to be spread
evenly over the year and they are assumed to occur in the middle
of the year on average.

Solution:
(a)
EPV(Claims) = 0.000023 × 1 000 000 × 1.1−0.5 = R21.93.
(b)

EPV(Claims) = 0.003 × 5 000 × a 12 j where j = 1.11/12 − 1

= 15 × 11.4005
= R171.01.

(c) The expected number of claims per month is 0.12/12 = 0.01 as

the claims are uniformly distributed over the year.

EPV(Claims) = 0.01 × 20 000 × ä 12 j × (1 + j )−0.5 ,

where j is as for (b) and ä 12 j = 11.49140. Alternatively:

? Can you see why the result is
0.5 multiplied by half a month’s
EPV(Claims) = 0.01 × 20 000 × a 12 j × (1 + j )
worth of interest? Draw a
gives the same result. timeline if needed.
EPV(Claims) = R2 289.17.

(d)
EPV(Claims) = 0.12 × 20 000 × 1.1−0.5 = R2 288.31.
Notice how little impact the monthly spreading of claims makes
for these 1 year policies — the answers to (c) and (d) are very
close.

7.1.3 Expenses

Policies can incur all types of different expenses. The most common
include:

Upfront costs

These are incurred at the outset of the policy and are usually quite
high as a proportion of total cost. They can cover items such as
commission to brokers, upfront administration costs of setting up a
record or file, issuing documents, etc.
Upfront costs are usually expressed as a fixed Rand amount, but could
be expressed as a percentage of the office premium.
130 Chapter 7 Pricing, Reserving and Emergence of Profits

Example 7.5 Below are two examples of upfront costs for a

policy:
(a) The upfront costs for the policy are R400.

EPV(Upfront costs) = R400.

(b) The upfront costs for the policy are 200% of the monthly of-
fice premium. You don’t yet know what the office premium
? If the upfront costs for this
is, so you will have to value the upfront costs as:
product are 200% of the first
premium, what is the finan- EPV(Upfront costs) = 2 × OP,
cial position of the insurer
during the first month of this and solve for OP later when you have set up the full equa-
policy? How do you think the tion of value.
insurer handles this?

Regular costs

These are often paid at the same time as the premium, so for an
annual policy with monthly premiums, there could be a monthly
expense of some sort. Regular costs could be a fixed Rand amount,
T Unless otherwise stated, or they could again be a percentage of the office premium, or the
it is reasonable to assume
risk premium, or some other amount altogether — it depends on the
that regular costs are paid
in advance, since premiums policy design.
are always paid in advance.
Example 7.6 Again, using an interest rate of 10% p.a.:
(a) The regular costs for a policy with a 10 year term are R200
per year.

EPV(Regular costs) = 200 × ä 10 = R1 351.80.

(b) The regular costs for a car accident policy with a one year
term are R15 per month payable when the monthly premi-
ums are due.

EPV(Regular costs) = 15 × ä 12 j = R172.37,

where j = 1.11/12 − 1.
(c) The regular costs for a travel policy are 5% of the office pre-
mium. The policy has a 1 year term and monthly premiums.
Again, we don’t yet know what the OP will be, so all we can
do is quantify the costs in terms of the OP:

EPV(Regular costs) = 0.05 × OP × ä 12 j ,

where j = 1.11/12 − 1.
7.1 Pricing 131

Claim costs

Some policies will have a cost associated with each claim. This could
relate to validating the claim, or the administration of paying the
claim.

Example 7.7
Below we use the same policies as in example 7.4. Again, interest
is 10% p.a.:
(a) The probability of a claim from a life insurance policy over
a year for a 40 year old man is 0.000023, and the cost of
processing the claim is R1 000. The timing of the claim is
uniformly distributed over the year.

EPV(Claim costs) = 0.000023 × 1 000 × 1.1−0.5 = R0.02.

(b) The probability of a claim for a travel policy is 0.003 per

month, and the cost of processing the claim is R120. Claims
are always settled at the end of the month. The term of the
policy is 12 months.

EPV(Claim costs) = 0.003 × 120a 12 j where j = 1.11/12 − 1

= R4.10.

(c) The expected annual number of claims on a car accident

policy is 0.12 and the cost associated with each claim is
R500. Claims are expected to be spread evenly over the
year, and it is assumed that they will occur in the middle
of each month on average. The policy has a one year term.
The expected number of claims per month is 0.12/12 = 0.01
because the claims are uniformly distributed over the year.

EPV(Claim costs) = 0.01 × 500ä 12 j × (1 + j )−0.5 = R57.23,

where j = 1.11/12 − 1.
(d) The same as for (c), but the claims are expected to be spread
evenly over the year and they are assumed to occur in the
middle of the year on average:

EPV(Claims) = 0.12 × 500 × 1.1−0.5 = R57.21.

Again, there is very little difference between (c) and (d) —

that’s why assuming claims take place in the middle of the
year is typically reasonable.
132 Chapter 7 Pricing, Reserving and Emergence of Profits

7.1.4 Premiums

Premiums are generally payable monthly, but some policies have a

single premium, or annual premiums, or premiums payable quarterly
or at some other frequency. Premiums are always paid in advance —
the insurer will not offer cover for claims unless it receives some
money first.

Example 7.8
(a) Single premium payable at the start of the policy:

EPV(Premiums) = OP.

This case has no discounting since the premium is paid just

once at outset.
(b) Monthly premium for a policy with a 1 year term, where
the annual rate of interest is i:

EPV(Premiums) = OPä 12 j , where j = (1 + i )1/12 − 1.

(c) Monthly premium for a 10 year policy:

EPV(Premiums) = OPä 120 j .

In (b) and (c) above, we are assuming that premiums are payable for
the full term of the policy. This is the case in most general insurance
policies, but unlikely in a life insurance policy — once someone dies,
they generally stop paying premiums! We will cover calculations of
payments that stop on death in the section on Life contingencies in
Module 3 of this course.

7.1.5 Risk margin

The risk margin will generally be expressed as a loading onto the

EPV(Claims), since it is basically an implied increase in claim num-
bers and / or value. So it can be valued as a simple increase of k to the
EPV(Claims), where k is the risk margin. However, the risk margin
could be expressed in some other way; for example, as a proportion
of the office premium.

Example 7.9 If the risk margin is 15% of claims, then:

EPV(Risk Margin) = k × EPV(Claims)

7.1 Pricing 133

7.1.6 Profit

Profit is generally expressed as a percentage of the office premium.

The insurer will require that a certain percentage of each premium
paid contributes towards profits. So the profit is expressed as a per-
centage multiplied by the EPV of the premiums.

Example 7.10 If the profit requirement is 15% of premiums,

then the EPV(Profits) for the policies below will be:
(a) Single premium payable at the start of the policy.

EPV(Profits) = 0.15 × OP.

(b) Monthly premium for a policy with a 1 year term and inter-
est of j per month:

EPV(Profits) = 0.15 × OPä 12 j

(c) Monthly premium for a 10 year policy:

EPV(Profits) = 0.15 × OPä 120 j .

7.1.7 Calculating the risk premium

U The risk premium is also
The risk premium is the premium that is required from policyholders
known as the “net pre-
to cover the expected cost of claims only. It makes no allowance for mium”, i.e. net of charges
expenses or profits. The risk premium is a useful measure because it and profit.
allows the insurer to know what proportion of its office premiums is
covering claims.
" A quick warning about termi-
The risk premium is calculated by equating the EPV(Risk Premium) nology: The risk premium
to the EPV(Claims). in insurance is different
from the risk premium we
used when we calculated the
Example 7.11 risk adjusted interest rate for
(a) A general insurance policy has a probability of a claim oc- valuing risky assets. The risk
premium concept in insur-
curing of 0.25 and an average claim value of R10 000. As-
ance refers to the amount a
sume claims are uniformly distributed over the year. The policyholder would have to
policy term is one year and a single annual premium is to contribute to cover the ex-
be paid at the start of the year. Using an interest rate of 15% pected claims on his policy.
p.a., the risk premium for this policy is: The risk premium in asset
valuation is the amount by
which the risk-free rate is in-
EPV(Risk Premium) = EPV(claims) = 0.25×10 000v 0.5 = 2 331.26 creased to allow for the risk
of an asset.
134 Chapter 7 Pricing, Reserving and Emergence of Profits

(b) If the premium on the above policy was payable monthly

in advance instead, and the same interest rate was used,
the risk premium (RP ) would be:

EPV(Risk Premium) = EPV(claims)

RP ä 12 @l = 0.25 × 10 000v i0.5
1
where l = 1.15 12 − 1 = 1.1715%. Therefore

RP = 206.96

Exercises for 7.1.7 Calculating the risk premium

Ex.7.3 Refer back to Example 7.3. Use an interest rate of 12% per
annum.
What is the risk premium if premiums are payable
annually in advance?
If premiums are payable monthly in advance for 12
months, what is the risk premium? (Assume that the
premium will be paid for 12 months, regardless of
whether there is a claim or not. )
Ex.7.4 Compare the annual premium and the sum of the 12 monthly
premiums that you calculated in Ex.7.3. What do you no-
tice?

7.1.8 Calculating the office premium

The office premium allows for expenses, profits and risk margins,
and a policy can have any permutation of these components. When
asked to calculate a premium, it is useful to bear in mind the basic
equation of value:
T Sometimes it can help to
calculate all the EPVs sepa-
EPV(Premiums) = EPV(Claims) + EPV(Expenses)+
rately, and then to put the
equation of value together. EPV(Profits) + EPV(Risk margin)

The above formula can be adjusted as needed to allow for the details
of the particular policy structure.

Example 7.12
A cellphone insurance policy has the following pricing basis and fea-
tures:
7.1 Pricing 135

• 0.09 claims are expected per year, uniformly distributed over the
year.
• Each claim is expected to be R2 000 on average.
• Policy expenses are R50 upfront, R5 per month and R20 per
claim.
• The insurer requires a profit margin of 20%.
• The pricing basis uses an interest rate of 2% per month, and
premiums are payable monthly over a term of 12 months.
Calculate the office premium.

Solution:

EPV(Claims) = 0.09 × 2 000 × 1.02−6 = R159.83

EPV(Expenses) = 50 + 5ä 12 + 0.09 × 20 × 1.02−6 .

Note that we have assumed that the monthly expense is paid in ad-
vance. This is a good time to work out ä 12 = (1−1.02−12 )/(1−1.02−1 ) =
10.7868 — we will need it throughout this example.

EPV(Expenses) = 50 + 53.93 + 1.60 = R105.53

EPV(Premiums) = OP × ä 12 = 10.7868 × OP
EPV(Profits) = 20% × OP × ä 12 = 2.1574 × OP.

There is no risk margin.

EPV(Premiums) = EPV(Claims) + EPV(Expenses) + EPV(Profits)

⇒ 10.7868 × OP = R159.83 + R105.53 + 2.1574 × OP
R265.36
OP = = R30.75 per month.
10.7868 − 2.1574
Note that some of the components of this equation of value have
very similar forms: claims and claim related expenses, premiums and
profits, claim cost and risk margin, etc. The equation above could
have more simply been written as follows:

OPä 12 × (1 − 0.2) = (2000 + 20) × 0.09 × 1.02−6 + 50 + 5ä 12 .

See if you can follow this expression. Use whichever you find most
comfortable — the second expression is easier to calculate but it is
also easy to leave out one of the items!

Exercises for 7.1 Pricing

For all of the exercises below, the interest rate is 12% p.a.
Ex.7.5 A car theft policy with a one year term and monthly pre-
miums is priced assuming that there will be 0.011 claims
per policyholder per year, and each claim will be R180 000
on average. Claims occur uniformly over the year. The
136 Chapter 7 Pricing, Reserving and Emergence of Profits

policy has expenses with a present value of R1 201, a profit

requirement of 18% and a risk margin of 5% of the EPV of
claims. Find the monthly office premium.
Ex.7.6 A personal accident policy has a term of one year, and a
single premium. The pricing assumption is that there is a
0.3% chance of a claim per year and that claims will average
at R400 000 and be uniformly distributed. The policy has
upfront costs of R200, claim costs of R1 000 and a profit
requirement of 13% of the office premium. Calculate the
office premium.
Ex.7.7 A five year business insurance policy has monthly premi-
ums of R194.26. The pricing basis allows for 0.025 claims a
year, occurring uniformly over the year. The average value
of the claim in year 1 is R50 000, but it is expected to grow
with inflation of 7% p.a. The policy has regular costs of
R500 per year, incurred in advance, and no other costs.
What is the profit margin used to price this policy?

7.2 Reserving

Reserving is a process where the insurer is asked to put aside some

money to back policies which have been sold. This serves as a buffer
against bad experience, making sure that the insurer can always cover
all the claims and expenses that it needs to pay. If things turn out
T The reserve can therefore as expected, the reserve is returned to the insurer at the end of the
be seen as a loan from the
policy.
insurer to the product.
This amount is calculated by assuming that the “realistic” assump-
tions used by the pricing actuaries have not resulted in premiums that
are enough to cover claims and expenses if things go badly for the
policy. The reserving calculation is based on the following equation
of value:
T This means that the pre-
miums and the reserves EPV(Premiums) + Reserve = EPV(Claims) + EPV(Expenses),
together must be enough
to cover future claims and so
expenses!
Reserve = EPV(Claims)+EPV(Expenses)−EPV(Future office premiums).

7.2.1 Reserving assumptions

The reserves are calculated using a set of assumptions that is pru-

dent.
7.2 Reserving 137

Exercises for 7.2.1 Reserving assumptions

Ex.7.8 What are the assumptions that we need to make when

calculating reserves?
Ex.7.9 For each of the assumptions you listed in the previous
question, how would you make that assumption more pru-
dent?

Prudent means that the assumptions lead to higher reserves.

U What would the reserves
equal to if the reserving basis
Example 7.13 For example, compared to the pricing basis, the was set to be equal to the
reserving basis could allow for: pricing basis?

• Higher claim numbers (in other words, higher probability

of a claim arising);
• Higher claim amounts;
U Effectively, we are saying:
• Lower investment returns. “What if investment returns
are poor and so we have less
The reserve is therefore calculated to be the amount needed at the money to pay claims?” For
short-term policies of a year
outset of the policy as a buffer against bad experience.
or less, the interest rate as-
sumption is not very impor-
7.2.2 Claims, Expenses and Premiums tant though.

Since the reserve is defined as Reserve = EPV(Claims)+EPV(Expenses)−

EPV(Future office premiums), the reserving calculation will involve
calculating the expected present value of claims, expenses and office
premiums. This is done in the same way as for pricing, except that
the reserving basis and not the pricing basis is used.

7.2.3 Profits

The reserving calculation does not account for profits. The reason
for this is that profit is not a liability: we are not obliged to pay
the insurer a profit. We only reserve for cashflows (like claims and
expenses) that we are obliged to pay. Profits are only paid when the
experience of the policies is favourable - when we are setting reserves,
we are concerned about future scenarios when things turn out badly
for the insurer, so there would be no profit at those times.

7.2.4 Risk Margin

The reserving basis is prudent, which means that it already has a built
in margin for risk through all the prudent assumptions it is using. A
risk margin could be added to this calculation for two reasons:
138 Chapter 7 Pricing, Reserving and Emergence of Profits

• To make the basis even more cautious by explicitly allowing for

an additional risk margin in addition to prudent assumptions.
• Alternatively, the actuary could use realistic assumptions plus
an explicit loading for risk to make the reserves more prudent.

The risk margin would most likely be calculated as percentage of the

EPV(Claims) just as it is in the pricing calculation.
Ultimately it does not matter whether the EPV(Claims) is increased by
a specific loading ((1 + k) × EPV(Claims) or whether the components
of EPV(Claims), i.e. the expected number and average value of the
claims, are increased — as long as the reserving basis does result in a
higher EPV(Claims) than the pricing basis.

7.2.5 Calculating the reserve

We can now calculate the reserve for a book of policies:

Example 7.14
Recall that the cellphone insurance policy we priced in example 7.12
has the following features:
• Policy expenses are R50 upfront, R5 per month and R20 per
claim.
• The insurer requires a profit margin of 20%.
• Premiums are payable monthly over a term of 12 months.
We calculated the office premium to be R30.75 pm.
Suppose that we have sold 10 000 such policies. Now we need to set
up a reserve for these 10 000 policies.
The pricing basis used an interest rate of 2% per month, and 0.09
claims are expected per year, uniformly distributed over the year.
Each claim is expected to be R2 000 on average. The reserving basis is
more conservative:
• 0.15 claims per year,
• Average claims of R2 500 on average,
• Interest of 1.5% per month.
We assume that expenses will be as expected.
Calculate the reserve.

Solution:
The reserve is then:

Reserve = EPV(Claims) + EPV(Expenses) − EPV(Future Premiums).

7.2 Reserving 139

EPV(Claims) = 0.15 × 2 500 × 1.015−6 ,

EPV(Expenses) = 50 + 5ä 12 + 0.15 × 20 × 1.015−6 ,
EPV(Premiums) = 30.75ä 12 ,
ä 12 = 11.07112.

Putting it all together, and grouping similar items together, we get:

Reserve = 0.15 × (2 500 + 20) × 1.015−6 + 50 + (5 − 30.75)ä 12

= 345.6969 + 50 − 285.0813
= R110.6156658 per policy.

Hence,

Total reserves = 10 000 × R110.6001585 = R1 106 156.66.

Exercises for 7.2.5 Calculating the reserve

Ex.7.10 The car theft policy from Ex.7.5 has a one year term and
an office premium of R338.67 per month. Each policy has
expenses with an expected present value of R1 242 on the
reserving basis. Calculate the reserve required for 30 000
such polices, assuming that there will be 0.015 claims per
policyholder and each claim will be R210 000 on average.
Claims occur uniformly over the year. The reserving inter-
est rate is 8% p.a.
Ex.7.11 The personal accident policy from Ex.7.6 has a term of
one year, and charges a single premium of R1 536.47. The
policy has upfront costs of R200, and claim costs of R1 000.
The reserving assumption is that there is a 0.5% chance of
a claim per year and that claim amount will be R480 000
on average, and be uniformly distributed over the year.
Calculate the reserve required for 15 000 such policies. The
reserving interest rate is 8% p.a.
Ex.7.12 The business policy in Ex.7.7 charges monthly premiums
of R194.26 and has a term of 5 years. The policy has regular
costs of R500, incurred annually in advance, and no other
costs. The reserving basis allows for 0.04 claims a year,
occurring uniformly over the year. The average value of
the claim in year 1 is R60 000, but it is expected to grow
with inflation of 9% p.a. The reserving interest rate is 10%
p.a. How much needs to be set aside as a reserve to back
10 000 such policies?
140 Chapter 7 Pricing, Reserving and Emergence of Profits

7.3 Emergence of profit

Emergence of profit refers to how the profit that insurers earn on the
policies they sell can be released from policies. The profits that are
generated by a book of policies will depend on:

1. The profit margin included in the office premium; and

2. the actual experience of the policy. i.e. the actual claims, ex-
penses and investment returns earned by the policy.

Example 7.15
If a book of policies experiences claims and expenses that are exactly
as assumed in the pricing basis, what profit will the insurer make on
this policy?

Solution: The insurer will make exactly the profit margin that was
priced into the policy. So if the profit margin was 10% of all office
premiums, the profit at the end of the policy term will be equal to the
accumulated value of 10% of the office premiums.

Once we know what amount is being set aside for the reserve, we
can observe what happens in the actual financial position of the
product over time. Think of a product having a “bank account” or an
“investment account”. Over time, the cashflows moving in and out of
this account will be:

+ Initial reserve amount

+ Premiums
- Claims
- Expenses

Exercises for 7.3 Emergence of profit

Ex.7.13 Consider a marine insurance product which launched to-

day. 100 policies with a term of 1 year were sold, each of
which pays premiums of R10 000 per month. Upfront ex-
penses setting up the policies are R1 000 per policy, and a
commission of R15 000 for the year is paid for each policy
at the start of the contract term. The insurer put aside a re-
serve of R1 500 000 for these policies. What is the financial
position of this book of policies at the end of today?
7.3 Emergence of profit 141

T We are for now assuming

If we are interested in the financial position of the policy at some time
that investment returns
after the start of the policy term, we need to account for the timing of are earned evenly over the
all the cashflows by accumulating them with the actual interest that period to make the calcula-
was earned. Positive cashflows will earn investment returns in the tions easier.
account and the negative cashflows will “miss out” on returns and
also have to be adjusted with interest to allow for this.
This development of the financial position is done as actual time
passes, and it allows for the actual experience of the policies — the
actual claims paid out, actual expenses, actual premiums and actual
investment returns earned.

7.3.1 Actual claims experience

Real claims experience can be reported in different ways. It could

be the total amount paid out in claims over the period; the amount
paid out per month; actual amounts and dates of claims; the actual
frequency and average size of the claims. However they are reported,
what we are looking to calculate is the total value of all the claims,
rolled up with interest to the end of the period.

Example 7.16 All of the examples below relate to the same

car insurance policy, and are different ways to report the same
experience over a 3 month period. There are 1 000 policies in
issue. Actual returns achieved were 4% over the 3 month period:
(a) The total claims were R9 945 740.
• Here, we have very little information about the timing
of the claims. We might assume that the claims hap-
pened on average half way through the 3 month period.
The value of the claims at the end of month 3 is then:

R9 945 740 × 1.040.5 = R10 142 704.

(b) The claims over the 3 months were: R3 490 100 in month 1,
R3 355 300 in month 2, and R3 100 340 in month 3.
• Accumulating those at the actual interest rate, and as-
suming each cashflow took place in the middle of the
respective month, we get:

R3 490 100×1.045/6 +R3 355 300×1.043/6 +R3 100 340×1.041/6

= R10 148 475.

T Can you see why the accumula-

tions are 5/6, 3/6 and 1/6? Draw
a timeline!
142 Chapter 7 Pricing, Reserving and Emergence of Profits

(c) The number of claims per policy was 0.1 per month, and
the average claim size was R33 152.47.
• We can convert this to 1000*0.1*R33,152.47 = R3,315,247
in claims per month, and value them assuming they
took place in the middle of each month:

R3 315 247×1.045/6 +R3 315 247×1.043/6 +R3 315 247×1.041/6

= R10 143 282.

• Or, we can say that if number of claims per policy was
0.1 per month, it was 0.3 for 3 months. That gives us
an estimated total number of claims of 1000 × 0.3 ×
33 152.47 = R9 945 740 for the 3 months, and then we
can proceed as for (a) above.
Note that the final value of the claims is slightly different de-
pending on which estimation method we use. This is because
we are assuming slightly different timing. The most correct an-
swer would be to get a list of all the amounts paid out and all
the dates on which they were paid. This would be best done in
a spreadsheet or computer program because there will be many
entries. As it is, all we can do is estimate, and you can note again
how little impact the timing has on the final value by comparing
the results under (a), (b) and (c).
T Remember that interest is
much more powerful over a
longer period.

7.3.2 Real investment returns

We already used the real investment returns to calculate the accumu-

lated claim amounts in the previous section. The main issue to note
is that the returns could be quoted as an annualised figure, a monthly
figure, or a figure per period — regardless of the length of the period
of experience.

Example 7.17 All of the examples below relate to a 4 month

period:
(a) The investment return earned was 4.311% over the period.
(b) The investment return over the period was equivalent to
13.5% p.a.
(c) The investment return over the period was 1.061% per
month.
Note that all of the above are equivalent to each other.
7.3 Emergence of profit 143

7.3.3 Real expense experience

At this stage, we are likely to assume that expenses turn out exactly as
expected in the pricing basis. The only difference is that they must
be accumulated at the real rate of return achieved over the period,
not at the assumed pricing or reserving rate.
Note that in reality, expenses are often estimated for pricing and
reserving, and the real expense experience may be different.

Example 7.18 For example, there may have been a R50 upfront
expense priced and reserved for, but in reality the cost ends up
being R48.70 because the department was more efficient than
assumed.

7.3.4 Real premium experience

Again, at this stage we assume that policyholders pay the correct

amount of premium exactly on the first day of each month. In reality,
premiums could be late, or some policies may stop paying premiums;
this is called “lapsing”. But we ignore this for now, and assume that
the office premium charged was in fact paid, and on time. We also
assume that premiums carry on regardless of whether a claim is made, U This would not be the case
in a life insurance policy —
as is usually the case for general insurance policies.
why?
So the only challenge is to accumulate the premiums with the real
investment returns achieved.

7.3.5 Building up actual assets to time t

We can now build up the financial position of a book of policies to

time t after the launch of the policy. We call this built up amount “ac-
tual assets” to indicate that this should represent the actual amount
of money that should be in our investment account with respect of a
book of policies at time t .

Actual assetst = FV(Reserves0 ) + FV(Premiums)−

FV(Claims) − FV(Expenses),

where FV indicates an accumulation with actual interest to time t .

If i is a monthly interest rate, t is the number of months in the period,
premiums are paid monthly, regular costs are incurred at the start of
144 Chapter 7 Pricing, Reserving and Emergence of Profits

each month, and claims are assumed to have occurred in the middle
of the period, then this can be expressed as follows:

Actual assetst = Reserves0 ×(1+i )t +Premiumss̈ t −Claims×(1+i )t /2 −

Upfront Costs × (1 + i )t − Regular Costss̈ t .

Example 7.19
Take the cellphone insurance policy that we have already priced and
reserved for in example 7.12 and 7.14. Recall that the premium we
calculated was R30.75 per month, and that there were 10 000 such
policies. We set up an initial reserve of R1 106 015.18 for these policies.
The experience over the next 2 months has been as follows:
A total of R280 000 has been paid out in 120 claims, uniformly over
the 2 months. Expenses have been as expected, i.e. R50 upfront, R5
in advance per month and R20 per claim for each of the policies.
Premiums have come in as expected. Investments have returned 15%
p.a. Calculate the actual assets just before the 3rd premium is due.

Solution:
We will need j = 1.151/12 − 1 = 1.171%, and s̈ 2 j = 2.03582.

2
Reserve2 = Reserve0 × 1.15 12 + 10 000 × (R30.75 − R5) × s̈ 2 j −
1 2
(280 000 + 20 × 120) × 1.15 12 − 50 × 10 000 × 1.15 12
= 1 132 081.40 + 524 111.00 − 285 708.29 − 511 783.54
= 858 700.57.

Exercises for 7.3 Emergence of profit

Ex.7.14 The car theft policy from Ex.7.5 and 7.10 has an office pre-
mium of R338.67 per month. A reserve of R10 469 930 has
been set up for a book of 30 000 such policies. It is now 1
month later, and the performance of the policy so far has
been as follows: 1.1% interest has been achieved over the
period and 35 claims were made amounting to a total of
R7 000 000. Expenses amounted to R250 per policy and
were paid at the start of the month. Premiums were as ex-
pected. What are the actual assets of this book of policies
now?
Ex.7.15 Recall that the personal accident policy from Ex.7.6 and
7.11 has a term of one year and charges a single premium
of R1 536.47. A reserve of R14 666 148 has been set up for
15 000 such policies.
7.3 Emergence of profit 145

Since then, 8 months have passed and the experience has

been as follows: 0.4% of the policies have claimed, and
the average amount of the claim has been R450 000. The
policy has upfront costs of R200, and claim costs of R1 000,
and those have been as expected. Premiums have all been
paid as expected too. Investments have delivered a return
of 1.2% per month. Calculate the actual reserves after 8
months if claims were spread uniformly over the year.
Ex.7.16 Recall from Ex.7.12 that a five year business insurance pol-
icy has monthly premiums of R194.26. The policy has
regular costs of R500, incurred annually in advance, and
no other costs. A reserve of R40 119 148 was set up for
10 000 policies of this kind 4 years ago. Since then, the
claims experience has been as follows:
Year 1: R23 980 000
Year 2: R24 200 000
Year 3: R26 500 000
Year 4: R23 100 000
Claims have been spread evenly over each year. Invest-
ment returns have averaged at 13% per annum over the
period. Expenses have been as expected, and so have pre-
miums. Calculate the actual reserves at the end of year
4.

7.3.6 Calculating the emerging profit

Now that we can calculate the financial position (i.e. the “actual
assets”) of a book of policies at any point in time, we should be able
to compare it with the required reserve at the same point in time.

Example 7.20 An insurer has issued 25 000 car insurance poli-

cies at the start of the year. It is now 1 July, and the insurer has
calculated that the financial position of this book, taking into
account the initial reserves, the premiums received, the claims
and the expenses paid out, is R315 million.
U Actual assets are calculated
The insurer’s reserving actuary has calculated that the required retrospectively, i.e. taking
reserve for this book of policies (taking into account expected into account what happened
future claims, expenses and premiums) is R285 million. in the past.
That means that R30 million (R315 million - R285 million) can U Reserves are calculated
be released from the product. That R30 million is partially re- prospectively, i.e. taking into
turning the insurer’s initial reserve that was “loaned” to these account what we expected
policies, and in part profit that has emerged on the policy. will happen in the future.
146 Chapter 7 Pricing, Reserving and Emergence of Profits

If the actual assets are higher than the required reserve, the insurer
can take the difference out as profit and repayment of capital invested
in the product when reserves were first set up.
If the actual assets are lower than the required reserve, the insurer will
have to invest more capital in the product, i.e. increase the reserves
to the required level.

Example 7.21
Returning to the cellphone policy, it is 2 months into the year, just
before the 3rd premium is due, and in Ex.7.19 we calculated that the
actual assets of the policy are now R858 700.57.
We can now calculate the reserve that is required for this policy for the
remaining 10 month term of the policy. We will use the same reserving
basis as at the beginning of the year, i.e. a claim frequency 0.15 claims
per year, average claim amount of R2 500 and interest of 1.5% per
month. Policy expenses are R50 upfront, R5 per month and R20 per
claim as before.
Once we calculate the reserve, we can work out whether the insurer
can take any profit out of the policy, and how much.

Solution: The reserve required for one policy is:

Reserve = EPV(Claims) + EPV(Expenses) − EPV(Future Premiums),

where:

10
EPV(Claims) = 0.15 × × 2 500 × 1.015−5
12
10
Note that we are using a claim frequency of 0.15 × 12 since we only
want claims for 10 months the year, and that we are assuming the
remaining claims will take place 5 months from now on average.

10
EPV(Expenses) = 5ä 10 + 0.15 × × 20 × 1.015−5
12
There are no more upfront expenses due, so we are ignoring those.
Regular expenses are payable for 10 more months. The claim expenses
are following the same pattern as claims.

EPV(Premiums) = 30.75ä 10

Again, only 10 more premium payments are expected now.

We need to calculate ä 10 1.5% = 9.3605, and then we can put it all
together, grouping similar items together:
10
Reserve2 = 0.15 × × (2 500 + 20) × 1.015−5 + (5 − 30.75)ä 10
12
7.3 Emergence of profit 147

Reserve2 = 292.40 − 241.04 = R51.356776 per policy.

Hence,

Total reserves2 = 10 000 × R51.356776 = R513 567.76

Comparing the required reserve of R513 567.76 to the actual assets of

R858 700.57, we can see that the insurer can release some of the actual
assets and still have sufficient reserves for the remaining 10 months.

Profit released = R858 700.57 − R513 567.76 = R345, 132.81

This means that actual assets can be reduced to R513 568, and the
insurer takes R345 133 as profit and return of the reserves it supplied
at the start of the policy.

Exercises for 7.3.6 Calculating the emerging profit

Ex.7.17 In Ex.7.14 we calculated that the actual assets of the car

theft policies one month after the start of the policies
would be R6 235 922.
The policy has an outstanding term of 11 months and an
office premium of R338.67 per month. The policy has re-
maining expenses with an expected present value of R1 000
on the reserving basis. Calculate profit that can be released
(or the additional reserves required) at time 1 (just before
the second premium is due) for 30 000 such policies , as-
suming that the reserving assumptions are:
• annual claim frequency of 0.015 claims per policy-
holder;
• average claim size of R210 000;
• interest rate of 8% p.a.;
• claims are uniformly distributed.
Ex.7.18 In Ex.7.15 we calculated the actual assets of a book of per-
sonal accident policies as R9 806 577 after the policies have
been in force for 8 months. Each policy has a term of one
year and charges a single premium of R1 536.47. The poli-
cies have upfront costs of R200, and claim costs of R1 000.
The reserving assumption is that there is a 0.5% chance of
a claim per year and that claim amount will be R480 000
on average, and be uniformly distributed over the year.
Calculate the profit that may be released, or the increase
in reserves needed for this book of policies at the end of
month 8. The reserving interest rate is 8% p.a.
148 Chapter 7 Pricing, Reserving and Emergence of Profits

Ex.7.19 The five year business insurance policies from Ex.7.16 have
accumulated actual assets of R32 861 355 at the end of 4
years. Recall that there are 10 000 policies and they have
monthly premiums of R194.26, regular costs of R500 per
year in advance, and no other costs. The reserving basis
allows for 0.04 claims a year, occurring uniformly over the
year. The average expected value of the claim in year 5 is
R85 000. The reserving interest rate is 10% p.a. Calculate
the release of profits or injection of reserves that is required
from the insurer at the start of year 5.
Solutions 149

Selected Numerical Solutions for Part 2

Chapter 3 Solutions
1
Ex. 3.1 1. 13
2. 18
37
1
3. 12
1
4. Either 365 or more accurately 41 × 366
1
+ 34 × 365
1
1
5. 36
Ex. 3.4 Ωk = {1,2,3,4,...} i.e. all positive integers
Ex. 3.5 1. 0.85
2. 0.15
Ex. 3.6 1. 0.000768
2. 0.917123
3. 0.001098
4. 0.297928
Ex. 3.7 1. Yes, 3.026%
2. No, need population of smokers
3. Yes, 4.186%
Ex. 3.9 3. • Game 1: 6.54
• Game 2: 1.54
• Game 3: 4.5
Ex. 3.10 2. -R2.7027
Ex. 3.11 4.25
Ex. 3.12 R652.75
Ex. 3.13 1. In 6 months time
2. 15 March
3. Mid January
Ex. 3.14 1. R2078.14
2. R3579.98
3. R56.30

Chapter 5 Solutions
Ex. 5.3 1. Insurer 1: R200 Insurer 2: R333.33
Ex. 5.4 1. AV@3% = 61 800, AV@18% = 70 800
2. 1 296
3. 69 503.96 vs 61 800
4. Yes
150 Chapter 7 Pricing, Reserving and Emergence of Profits

Chapter 6 Solutions
Ex. 6.6 R280.00

Chapter 7 Solutions
Ex. 7.1 R391.50

Ex. 7.2 1. 3 months from now

2. 31 December
3. 2.5 years from now

Ex. 7.3 1. 238.12

2. 20.89

Ex. 7.5 R338.67

Ex. 7.6 R1536.47

Ex. 7.7 17%

Ex. 7.10 R10 468 280

Ex. 7.11 R14 666 148

Ex. 7.12 R40 119 148

Ex. 7.13 R900 000

Ex. 7.14 R6 235 922

Ex. 7.15 R9 806 577

Ex. 7.16 R32 861 355

Ex. 7.17 R868 094 Profit

Ex. 7.18 R2 065 165 Additional reserve needed

Ex. 7.19 R17 766 815 Profit

 Module 3

Institutions

151
Chapter 8

General Insurance

General insurance covers policyholders against losses arising from damage to U General insurance is called
“general” because the defini-
their property, or damage that they or their property inflict on others. Policies are
tion of it is very wide — it is
generally designed on the basis of indemnity, i.e. restoring the policyholder to basically any policy that is
where they would be had the event not occurred, although it is possible to have not life or health insurance.
a specified sum assured for some event, e.g. R1 000 000 on the loss of an index But even then the distinc-
finger. tion can be vague; for ex-
General insurance policies cover a policyholder against a variety of perils, ample “gap cover”, which
including fires, floods, accidents, thefts and fraud. Recall the four basic financial is a South African insur-
ance product that covers
risks we identified in Module 2: general insurance protects policyholders against
the difference between what
the risk of loss or damage to property and possessions.
a healthcare procedure actu-
ally cost, and how much the
Perils and hazards person’s medical aid policy
covers, is a type of general
A peril is what we have been calling a “risky event”: fire, flood, accident, strike insurance policy — but it
action, etc. The peril is what causes the loss to happen. relates to health! Legally,
Here is a list of the perils that a car is exposed to: you can tell if something is
• fire, a “general insurance” prod-
uct by checking if it is being
• theft (both theft of the car and theft of the contents of the car),
sold by a company that is
• accident,
authorised to sell general
• flood, insurance products. Go to
• hail, [Link]
• hurricane, search_fsp.htm and type
• lightning, and in the name of any insurer,
then click the “Details” but-
• vandalism.
ton, and then click “Products
• Can you think of more?
Approved”, and you will see
A hazard is a condition that makes a peril worse. So if the peril is fire, the hazard the types of insurance this
could be using a wood stove. company is authorised to
For the perils above, here are some hazards: sell.
• Fire: electrical problems, old car.
• Theft: lack of security system, parking on the street and living in an area with
high crime rates.

153
154 Chapter 8 General Insurance

• Accident: inexperienced driver, rainy weather, poor visibility.

• Flood: parking/living on floodplain.
• Hail: parking outdoors, living/driving in an area where hail storms occur.
• Hurricane: parking outdoors, living in an area prone to hurricanes.
• Lightning: parking outdoors, living in an area prone to storms.
• Vandalism: parking on the street, living in an area prone to crime, having a car
without an alarm system.
• Suggest some more hazards for each of the above!

Exercises for 8 General Insurance

Ex.8.1 Consider the perils that affect a small business. What are the hazards
that can be associated with each peril?

Case study
Let’s say that an insurer wants to develop a new type of policy aimed at students.
This policy will protect the student from the loss of their electronic equipment, e.g.
cellphones, laptops, iPads, etc., due to damage or theft. What are the perils that a
student’s personal electronic devices are exposed to?
• Theft,
• accident, e.g. being dropped,
• fire, and
• being lost.
For each peril above, think of the hazards that may make the peril worse. To identify
particular hazards, think of how the electronic equipment is transported, where the
owner is likely to be, etc. And remember to think outside of your own experiences.
The policy will be marketed and sold to students nationwide, and they may face
very different risks from you.

8.1 Who needs general insurance, and what types of prod-

ucts are there?
Individuals and companies both need general insurance. There are four main
types of general insurance, each protecting the policyholders from a different
type of risk. We discuss each now.

8.1.1 Property damage insurance

8.1 Who needs general insurance, and what types of products are there? 155

Exercises for 8.1.1 Property damage insurance

Ex.8.2 List the different types of things that you own that could be insured.

Individuals are at risk of their property being damaged, destroyed or stolen. U Property damage policies
This can include houses, household goods, vehicles, livestock, luggage, and any- do not cover damage due to
normal wear and tear — can
thing else they own. Companies also need to protect what they own, from build-
you think why not? Think
ings and property, to vehicles and stock. Property damage insurance indemnifies about the uncertainty part of
the policyholder against the loss of or damage to the policyholder’s own material this risk in particular.
property.
The main types of property that are subject to such damage are:
• residential buildings, e.g. houses, flats,
• moveable property, e.g. contents of house, personal electronics, stock and
equipment,
• commercial buildings, e.g. offices, factories, shops,
• land vehicles, e.g. cars, trucks, trains,
• marine crafts, e.g. boats, ships,
• aircrafts, and
• crops and livestock.

Exercises for 8.1.1 Property damage insurance

Ex.8.3 Consider a wine farm in Stellenbosch. What types of property insur-

ance cover are they likely to need?

8.1.2 Financial loss

This type of cover indemnifies the policyholder against events that have directly
caused them a financial loss. Examples include:
• A company may suffer because a debtor does not pay the amount owed; this
would be covered by a “pecuniary loss” policy.
• Money is stolen or embezzled by employees; this would be covered under
“fidelity guarantee” cover.
• A company may suffer losses as result of having to suspend work for a period
of time, e.g. due to a strike or damage to the factory; this is covered under
“business interruption” cover.
These types of cover are most often required by companies rather than individu-
als.
156 Chapter 8 General Insurance

Exercises for 8.1.2 Financial loss

Ex.8.4 If “strike action” is a peril, what hazards may be associated with it?
T Note that hazards are asso-
ciated with the policyholder Think about what factors would make it more likely that one employer
rather than the general envi- experiences business interruption as a result of strike action more
ronment. So for example, a frequently than another.
recession is not a hazard, be-
cause it affects the economic
8.1.3 Fixed benefits
environment — it is not a fea-
ture of a particular business. These are policies that simply offer an agreed sum assured when an agreed future
event occurs. For example, many of them insure specific body parts like a hand or
? Recall the guidelines for an eye. These types of policies are often taken out by celebrities or other persons
what suitable for insurance,
whose livelihood depends on a particular skill or look. It tends to be difficult to
from Module 2. Which of
assess such risks and the principle of indemnity does not really fit these policies
these guidelines do not fit
with fixed benefit insurance? — they are often priced with high margins to make sure that the experience is not
adverse.

8.1.4 Liability insurance

Liability insurance provides indemnity where the insured is legally liable to pay
compensation to a third party. Any legal expenses relating to such liability are
usually also covered. These policies will not cover illegal acts or negligence.
? Can you think of why illegal The basic benefit provided by any liability insurance product is an amount to
acts and negligence are not
indemnify the policyholder fully against a financial loss.
covered?
The main types of liability insurance are:
• Employer’s liability: this protects an employer from employees demanding
compensation for work-related injuries, accidents and illnesses.
• Motor third party liability: this protects vehicle owners from damage caused
by them to other drivers, pedestrians or property.
• Public liability: this protects companies from paying compensation to
members of the public as a result of something that is seen as the com-
pany’s fault, e.g. accidents on their premises or damage caused by their
employees.
• Product liability: this protects companies from damage their products may
have caused to the public.
• Professional indemnity: this protects professionals such as actuaries, doc-
tors, lawyers, financial advisors etc., from any claims against them as a result
of advice or services rendered to clients in their professional capacity.

Exercises for 8.1.4 Liability insurance

Ex.8.5 For each event below, state which type(s) of liability insurance claim it
may result in:
1. Someone slips on a banana peel at Fruit and Veg City.
8.2 Accessing general insurance products 157

2. A child swallows all the magnets in a magnetic toy kit.

3. A doctor misdiagnoses cancer as the common cold.
4. A car has faulty breaking system that results in a collision with
another car.
Think of who may claim in each case, and against whom!

8.1.5 Policies as combinations of cover

Many events naturally lead to several different types of claims arising, for example:
• A car accident may result in damage to your own property (property dam-
age), and damage to someone else’s car (motor third party liability).
• A fire in a factory may result in property damage, injury to employees (em-
ployer’s liability), and business interruption.
Many policies will carry a combination of different forms of cover sold as a pack-
age to better meet the needs of a policyholder. For example, “comprehensive car
insurance” covers the policyholder against accident, theft, and third party claims.

8.2 Accessing general insurance products

General insurance policies can be sold to clients through different channels. You
could use a broker (also called intermediary), who is an independent advisor
who can compare the products of different insurers and advise the client on the
merits of each product. Or you could get a quote directly from the insurer, either
through a call centre or via the internet. The capacity for advice is more limited
in these cases, but there are usually no fees or commissions payable with those
sales.
The internet has also led to the development of “aggregator” websites where
a client can obtain several quotes from different insurers via a single request.
This enables clients to compare the pricing of different offerings without using a
broker. These types of websites are generally free of charge for the consumer. T [Link] is an example
of such a site. How do you
think Hippo may be generat-
8.3 Underwriting ing revenues?

Underwriting is the process at the start of a policy where an insurer decides

whether to accept a policyholder, and on what terms.

Example 8.1
An insurer may require that an applicant for a motor vehicle insurance
policy has sufficient security features in their car. For example, an insurer
may refuse cover to an applicant if their car does not have an immobiliser
installed.
158 Chapter 8 General Insurance

The underwriting process also allows the insurer to collect information about
a policyholder so that the premium can be set to accurately reflect the probability
and size of claims from the policy. This is called “rating”, and we will discuss it in
detail in the section on pricing.
? Which do you think needs
more underwriting: A ma-
rine property insurance pol- Case study
icy or a car insurance policy? In the case of the policy for student’s personal electronics, how would you assess
the risk for each student? Would you decline cover to some people? Think about
your friends — are there some people that you think should be charged more than
others? Why? What information could the insurer request to identify these people?

8.4 Cashflows of a general insurance policy

General insurance policies have the following characteristics:
? In what type of insurance
do you think premiums stop • Premiums are payable either as a single premium in advance, or as regular
when a claim is made? premiums over the term of the policy. Premiums continue until the end of
term and do not cease if a claim is made.
• Policies generally have a short term, e.g. one year. After the term is finished,
T General insurance is also
a policyholder can choose to renew his policy, but a new premium will be
called “short term insurance”
because of this. calculated.
• Usually, there can be any number of claims during the term of one policy.
For example, a policyholder could have several accidents in a year and lodge
several claims on a motor insurance policy.
• Usually, both the claim frequency and the claim value are very unpredictable.
This is because most policies operate on the basis of indemnity, and the
? Which of the types of insur-
exact amount that will be required to put the policyholder in the same
ance detailed in Section 8.1
has predictable claim values,
position will differ from event to event.
however? General insurance policies generally cover the policyholder against any claims
U Claims which have occurred which occur within the policy term. Some types of cover have a very short period
but have not been reported between a claim occurring and being reported, e.g. burglaries are usually reported
yet are called “Incurred But very quickly. But there are policies in which it can take months or even years
Not Reported” claims: IBNR for a claim to be reported. These types of policies are said to have a “long tail”.
claims. Actuaries need to An insurer selling this type of policy must be prepared to honour claims a long
estimate the amount that is
time after the original policy has expired. When reserving for general insurance
likely to be due under IBNR
claims and set aside reserves business, those potential claims must be properly allowed for.
for these claims.
8.5 Pricing 159

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Premiums
Claims

Figure 8.1 The cashflows of a hypothetical general insurance policy.

Asbestos:
Exercises for 8.4 Cashflows of a general insurance policy Asbestos is a name given to a set
of naturally occurring minerals.
Ex.8.6 An engineering firm experiences the following events: It used to be popular as it has
1. 30 October 1999: Engineer Marcus falls off a ladder while doing several desirable features that
make it useful in manufacturing
routine maintenance and breaks a leg. and construction. However, it
2. 17 November 1999: Marcus submits a bill of R5 000 for his medical was discovered that prolonged
treatment to the accounting department. inhalation of asbestos fibres can
cause serious health problems,
3. December 1999–February 2000: A commission investigates if Mar-
including lung cancer. Its use is
cus had been following all the safety procedures. It is eventually now banned in many countries
found that he did, and therefore the accident was a genuine work- around the world, including South
place accident. Marcus is reimbursed. Africa.
The diseases arising from asbestos
4. 15 January 2000: A secretary is diagnosed with severe carpal tunnel
can take years to manifest them-
syndrome. An independent medical practitioner determines this selves. As a result, many employee
was caused by typing in an uncomfortable position over the course liability claims related to asbestos
of the previous year. The cost of treating the condition is R200 000. exposure have arisen many years
after the employee liability cover
The engineering company had an employer liability policy which ran ended. But as the exposure to the
from 1 January 1999 until 31 December 1999. Do you think the two asbestos occurred during the term
claims would be covered under this policy? Explain your reasoning. of the policy, the insurer is liable to
pay. These claims are examples of
long tail claims.
8.5 Pricing For more information see:
[Link]
archive/old-articles/part-3/
We have already covered the general principles of calculating Risk Premiums
asbestos-claims-in-europe/.
and Office Premiums in Module 2. Now that we are dealing with general insur-
ance in particular, we need to consider how an insurer would allow for different Table 8.1 Asbestos: long tail
policyholders’ actual risk in its premium calculations. claims.
160 Chapter 8 General Insurance

Example 8.2 Consider the following pairs of potential policyholders. Can

you determine who would be charged more for the policy they require?
Remember to consider both the likelihood and the value components of the
risk.
(a) A person living in Constantia and a person living in Nyanga applying
for car theft insurance?
(b) A person driving a new BMW and a person driving an old VW Beetle
applying for car accident insurance?
(c) A person whose house has burglar bars and an alarm system and a
person who doesn’t applying for burglary cover?

In the above example, your answers were dependent on both the hazards
that each person was exposed to, and the relative values of their possessions.
The concepts of risk factors, rating factors and exposure will help you to assess
premiums for different policyholders.

8.5.1 Risk and rating factors

As we have seen, different hazards will have an effect on the expected value of the
claims. Hazards can affect the risk faced by increasing the probability of claims
and the value of the claims.
In order to calculate the premium, we must be able to estimate the claim
frequency as well as the value of the loss. This means that we somehow have to be
able to measure the effect of all the hazards. This is done by using risk and rating
factors.
Risk factors are characteristics of the policyholder that precisely define how
much risk he or she is exposed to. Take a motor vehicle policy that covers you
against accidents. The risk factors will include:
• the ability of the driver,
• the ability of other drivers on the roads that you use,
• the condition of the roads that you drive on,
• the condition of the car,
• how much time the car spends on the roads,
• how busy the roads are, and
• the cost of fixing the car after an accident.
• See if you can think of some more?
If we were able to correctly quantify this information, it would be easy to
compare different policyholders and estimate how likely each one is to have an
accident. For example, a person who is a terrible driver who drives every day for
six hours a day in city traffic is a much worse risk than a person who is a driving
instructor, who only drives once a week in Franschhoek.
8.5 Pricing 161

So the ideal, for a general insurer, is to collect this information — the risk
factors — accurately for each policy it issues.
? If you were asked the ques-
tion: “do you mostly drive
Example 8.3 in good or poor conditions?”
What may be the problem with obtaining this information? — what answer do you think
would result in the lowest
Solution: One problem is that not all risk factors are easily measurable. This could premium? Is lying on ap-
be because the information is subjective. For example, driving ability is a very plication forms an exam-
subjective measure. Most of us believe we are better than the average driver! ple of moral hazard or anti-
Qualitative risk factors like “the conditions on the road” are also difficult for people selection?
to estimate, and moreover people may be tempted to twist the truth a bit if they
can figure out what will result in a lower premium.
Also, some information relates to the future, and cannot be exactly determined: the
true risk factor is “how much will the car be driven in the next year”, but we don’t
know this at the beginning of the year. We could ask how much it was driven last
year, but that is no longer the true risk factor: it is a rating factor.

Rating factors are approximations of risk factors that are:

• easily measurable,
• roughly proportional to the risk (but not necessarily exact), and
• difficult to misrepresent.
So rating factors for the car accident insurance policy might be:
• gender of driver,
• age of driver,
• how long the person has been driving for,
• whether the car is used for business or personal use,
• address at home,
• whether the car is parked on the street or locked away at night,
• address at work, and
• the estimated current value of car.
These are much more concrete and measurable. They are not a perfect measure T Some risk factors are mea-
surable and in that case, they
of risk, but can be used to approximate the risk and they are less susceptible to
should be used as rating fac-
manipulation by the policyholder. tors. For example, on a car
theft policy, the value of the
Exercises for 8.5.1 Risk and rating factors car is both a risk and a rating
factor.
Ex.8.7 For each of the rating factors above, which risk factors do you think
they are proxies for? Consider that sometimes more than one rating
factor will be required to proxy a risk factor.
162 Chapter 8 General Insurance

General insurance policy extract

If you look at an insurance policy, or get a quote for insurance, you can see the
rating factors used by the insurer.

Figure 8.2 This is an extract from a sample policy for a house in Cape
Town insured by Outsurance. The insurance is for fire, theft, damage
and liability to other parties.
Consider the various items noted. What peril does the pitch of the roof relate
to? What about the construction materials? Why does it matter that the house is
occupied at night, or whether it’s a commune or not? What perils is each rating
factor relevant to? What effect may each of the items listed in the policy have on
risk?

Case study
For the policy protecting student electronics, list the risk factors that exactly affect
the amount of risk that a policyholder may be exposed to, and then think of some
rating factors that could be used to approximate the risk factors.

Once the insurer has a set of rating factors, they can estimate the probability
of a claim for each policy. They do this by collecting data about past claims,
and dividing them into groups based on the rating factors. Then it is possible to
estimate the risk for each sub group.

Case study
The insurer has conducted such a study with respect to phones, iPads and laptops.
The rating factors that have been found to be significant are which university a
student goes to and whether the student owns a car or not. The results, based on
? Can you think of why these information collected in past years, are as follows:
two factors might have an
impact on loss/damage to
electronic goods?
8.5 Pricing 163

University A B
Car Yes No Yes No

Students with phones 8458 5444 7801 2340

Number of phones lost/stolen 754 1054 897 504
Number of phones damaged 56 50 65 20

Students with iPads 1054 221 511 106

Number of iPads lost/stolen 58 20 49 30
Number of iPads damaged 12 2 8 1

Students with laptops 2455 1003 1366 555

Number of laptops lost/stolen 78 112 140 97
Number of laptops damaged 101 50 87 34

1. Draw up a similar table and for each sub-group, calculate the odds of having
each item lost/stolen, or having it damaged. By each sub-group, we mean
“Students at university A who have cars and have a laptop”, or “Students at
university B who have no car and have a phone”.
2. From your table, consider whether having a car has a positive or negative
effect on having an item stolen/lost. Can you think why this may be?
3. Which item is generally most at risk of damage? What about loss?

8.5.2 Exposure
Another useful term for premium calculations is the “measure of exposure” for
the risk. This refers to the value component of the policy, and it is a way to
standardise the premium. T Note that the exposure mea-
sure is one of the risk/rating
factors — one that is pro-
Example 8.4 For a car policy the measure of exposure is often the car-year, portional to the value of the
because some policies will have more than one car on them, or could be claim!
for a period other than 1 year. It is difficult to compare a policy for 3 cars
T Remember Natasha and
insured for 6 months at a cost of R500 to a policy where 5 cars are insured
her house risk pool in Sec-
for 1 year at a cost of R2400. In order to compare them, it is easiest to find a tion 4.5.8? We determined
“common denominator”, and convert each policy. In this case, the common that the premium should be
denominator is “How much would this policy be if it was for 1 car and 1 different depending on the
year?”. value of the property, so that
The first policy costs R500/3 × 2 = R333 per car per year. the premium could be calcu-
lated as R833 per R1 million
The second policy is R2400/5 = R480 per car per year. Now we can see that of property per annum. The
the second policy charges more per car per year. So “car-year” served as a “R1 million of property per
unit of exposure. annum” is the measure of
exposure in this case.
164 Chapter 8 General Insurance

Exercises for 8.5.2 Exposure

Ex.8.8 For each of the policies below, use the given measure of exposure to
calculate the premium per unit of exposure:
1. A house insurance policy costs R500 per month for a house valued
at R2.8 million. The unit of exposure is R1 million of property per
year.
2. Medical gap cover costs R150 per month for a family of four. The
unit of exposure is a person-year.
3. A travel insurance policy taken out by a couple going on a three
week vacation costs R850. The unit of exposure is a person-day.

Case study
What could be the measure of exposure for our proposed policy to protect students
from the loss of their electronic equipment? Some possible measures are R1 000 of
equipment per annum, or 1 gadget per annum. The last measure would mean that
a student who has a cellphone and an iPad and an MP3 player and takes out the
policy for 6 months would be expected to pay more or less 1.5 times as much as a
student who only has a laptop and takes the policy out for 1 year. Can you work out
how we got to this?

8.5.3 Pricing general insurance policies

When pricing a policy, the following need to be determined:
• What is the insured property?
• What are the perils which the policy will cover?
• What is the probability and present value of claims for this policy?
• What is the measure of exposure that we will price for? (Once we have the
premium per unit of exposure, we can multiply it by the number of units of
exposure associated with a particular policyholder).
• What are the risk factors that determine how likely the various insured perils
are to take place?
• What are the rating factors that we can use as a proxy for each risk factor?
• What adjustment to the probability of the event do we make for each rating
factor?
This would give us enough information to calculate the risk premium for each
policyholder.
T Remember that we defined
the risk premium as:
Example 8.5
Risk premium = Say that a car insurance policy has the following characteristics: for a female driver
expected number of claims× who has had her license for more than 5 years, the expected number of claims as a
present value of each claim.
8.5 Pricing 165

result of an accident is 0.3 per car per annum. The expected value of each accident
related claim is R8 950. The rating factors used by the company are gender and
whether a driver has had his license for 5 or more years, or less than 5 years. The
insurer has analysed its data and found that:
• Men are 5% more likely to have accidents than women, and
• Drivers who have had their license for less than 5 years are 17% more likely to
have an accident than more experienced drivers.
(a) Calculate the annual risk premium for the female driver who has one car and
has had her license for 5 years. Ignore interest.
(b) Calculate the annual risk premium for a male driver who just got his license
and has 2 cars. Ignore interest.

Solution:
(a) Risk premium = 0.3 × R8 950 = R2 685 p.a.
(b) Risk premium = 0.3 × 1.05 × 1.17 × R8 950 × 2 = R6 597 p.a.

Exercises for 8.5.3 Pricing general insurance policies

Ex.8.9 A business building fire insurance policy has the following features:
• For an office block constructed out of brick and with fire alarms
installed, the risk premium is R10.20 per square meter per month.
• If the building is a shop, the premium increases by 5%. If it is a
factory, the premium increases by 12%.
• If the building is constructed out of wood, the premium increases
by 10%.
• If the building has no fire alarms, the premium increases by 20%.
Assume that the increases are multiplicative.
1. What is the unit of exposure used in the calculation of the premi-
ums?
2. What are the risk premiums payable by the following:
3. A 40m2 shop constructed out of wood, with a fire alarm?
4. A 5 000m2 factory constructed out of bricks, without fire alarm?

5. An office block constructed out of bricks, with no fire alarms

and a 200m2 area?

Case study
It has been decided that the policy for student electronics will be designed as
follows:
• The measure of exposure is R1 000 of electronic goods per year.
166 Chapter 8 General Insurance

• The risk factors are: the area where the student lives and goes to school,
whether the student drives or walks or takes public transport, and how careful
the student is.
• The rating factors are: the university the student attends, and whether the
student owns car.
The company has put together the following table, based on previous research on
students:

University A B
Car Yes No Yes No
Chance of claim p.a. 0.08849 0.19316 0.12875 0.22859

The company has also estimated that the average claim is for 50% of the cover, i.e. if
the student has electronic goods worth R10 000, any particular claim lodged by the
student is likely to be R5 000 on average.
Calculate the annual risk premium per R1 000 of cover for a student from university
A who has a car (ignore interest).
Answer: Risk premium = Pr(claim) × Value(claim) = 0.088493 × R1 000 × 50% =
R44.25.
What about these students:
• Student from university B who has a car?
• Student from university A who has no car?

Once the risk premium is known, the office premium for each policyholder
can be calculated using the techniques covered in Module 2.

8.6 Claims
The claim characteristics of general insurance policies differ depending on the
type of cover. In this course, we are using very simplified assumption about the
occurrence of claims, that allow us to make simple calculations of the form

expected number of claims × present value of each claim.

In reality, more complex statistical methods are used to value expected general
insurance claims.
Some features of general insurance policies which affect claims include:

8.6.1 Excess
Excess is a policy feature where for each claim, the policyholder pays an agreed
amount, and the insurer covers the remainder of the claim. This is a way of
? What effect would excess sharing risk between policyholder and insurer.
have on claim value? Would
Including an excess in the policy design has various advantages for the insurer
there also be an effect on
and the policyholder:
claim frequency?
8.6 Claims 167

• policyholders pay lower premiums;

• insurers need to process fewer claims, and therefore administration and the
associated cost is reduced. This is because all claims with values less than
the excess will no longer need to be processed.
T Some policies allow policy-
holders to select the level of
Example 8.6 excess they would like to pay.
A household policy has in the past not had any excess attached to it. The experience This is a way of allowing poli-
in the past has been that claims are R11 000 on average and claim frequency is 0.1 cyholders to choose the level
claims per policy per annum. The insurer is now changing the policy to introduce of risk they can self-fund
an excess of R5 000 per claim. How could this affect: and insure the rest.
(a) Claim frequency?
(b) Claim value?

Solution:
Claim frequency: If some claims are expected to be lower than the excess, then
those claims will no longer be reported after the excess is introduced. That means
that claim frequency is likely to be lower with the excess.
T The effect of excess on claims
Claim value: The immediate effect is that those claims that are reported, are re-
is only predictable if exact
duced by R 5000 since the excess will be paid by the policyholder. However, this
claim distribution is known,
does not mean that the average claim value reduces by R5 000! The actual change
or if the excess is relatively
in the average claim amount will depend on the distribution of the claims. For
small so that the vast majority
example, if half the claims were R2 000 and half were R13 000, then the R2 000
of claims exceed the excess.
claims would no longer occur, and the R13 000 claims would be reduced to R8 000
— so introducing excess would reduce the average claim to R8 000. On the other
hand, if 75% of the claims were R5 000 and 25% were R29 000 (still averaging out
to R11 000!), the average claim after the excess was introduced would increase to
R24 000 (R29 000 − R5 000). The actual claim distribution will be more complex
U A secondary effect on claim
than this, but this example illustrates that the effect of excess on average claims is
value and frequency may be
difficult to predict!
to do with moral hazard —
can you see how this would
happen?
Exercises for 8.6.1 Excess

Ex.8.10 The following table shows the distribution of claims for a personal
effects insurance policy:

Claim value Number of claims p.a. over 10 000 policies Average claim
R0–R1 000 202 R350
R1 000–R2 000 450 R1 400
R2 000–R5 000 700 R3 120
> R5000 230 R7 040

The policy currently has no excess.

168 Chapter 8 General Insurance

1. Use the above data on past experience to calculate the appropriate

annual risk premium for this policy, ignoring interest.
2. Calculate the risk premium if different levels of excess are intro-
duced:
3. R1 000 excess;
4. R2 000 excess;
5. R5 000 excess.
? What effect do you think
does this feature has on
claim frequency? And what 8.6.2 No claims bonus
about claim values? Another common feature of general insurance policies is a no claims bonus.
U A no claims bonus often car- This pays the policyholder a ‘bonus’, effectively a refund of some premiums,
ries on and grows from year at the end of the policy term if the policyholder did not have any claims during
to year. This has the effect of the period.
increasing policyholder loy- When a no claims bonus is in place, policyholders often decide to forego
alty as people prefer to stay smaller claims in order to retain their bonus. This reduces the number of claims
with their current insurer
for the insurer, making administration less onerous. Average claim values, how-
because they have built up a
ever, would be expected to increase. Can you see why?
high no claims bonus.

? A no claims bonus may also Example 8.7

reduce morale hazard — can
Say a policy has in the past had claims of R5 000 on average and a claim frequency
you explain how?
of 0.35 claims per annum.
(a) Calculate the annual risk premium based on these past statistics.
The insurer has now decided to introduce a no claims bonus. The actuary estimates
that claim frequency will reduce to 0.25 claims per annum. The average claim is
now however expected to be higher, at R5 500.
(b) Calculate the revised risk premium using these assumptions:
(c) If the insurer charges the full premium of R1 750 p.a., what level of no claims
bonus can they afford?

Solution:
(a) RP = 5 000 × 0.35 = R1 750 p.a.
(b) RP = 5 500 × 0.25 = R1 375 p.a.
(c) Let’s assume there are 1 000 policies. The expected number of policies without
claims under the new system is (1 − 0.25) × 1 000 = 750. The no claims bonus
must be funded from the excess premium the insurer receives: it charges
R1 750, but under the no claims bonus the actual risk premium is R1 375. So
the excess premiums are (R1 750 − R1 375) × 1 000 = R375 000. If we distribute
the R375 000 between the 750 policyholders who qualify for a bonus, each
person will get R375 000/750 = R500 in no claims bonus.
8.7 Reserving 169

8.7 Reserving
Reserving refers to setting aside a provision for future claims. The examples of
reserving in Module 2 were simple examples of general insurance reserving for
claims which had not occurred yet. Other reserves may be set up for:

• IBNR claims reserve: claims which the insurer expects have already hap-
pened but have not been reported to the insurer yet.
• Outstanding reported claims reserve: claims which the insurer knows
about but has not paid yet.
• Catastrophe reserve: a reserve to assist the insurer in the case of a catas-
trophic event that causes a surge in claims.
U Another way for an insurer
to protect themselves from
catastrophes is to take out
Exercises for 8.7 Reserving
reinsurance.
Ex.8.11 Which type of insurance policy do you think has more outstanding re-
ported claims: a comprehensive car insurance policy, or a professional
liability insurance policy? Why?

8.8 Monitoring
The general insurance business has fairly dynamic risks, which are not very well
suited to insurance. However, insurance is still possible due to the short term
nature of the business. Unlike life assurers, whose premiums need to hold for ? Why are general insurance
risks dynamic?
terms up to 10 years or longer, general insurers have the luxury of recalculating
premiums annually. This means that they can quickly take advantage of trends in
claim experience and to a certain extent can adjust for past mis-pricing by loading
future premiums. General insurers will thus undergo annual pricing exercises
where the experience over the past year is examined and incorporated in expected
claim experience for the next year.

Exercises for 8.8 Monitoring

Ex.8.12 An insurer is reviewing the claims experience on a book of 100 000

household insurance policies. The experience of the policies over the
last year has been as follows:
• Premiums received: R524 million
• Expenses: R42 million
• Claims which occurred during the year and were settled: R458
million
• Claims which were reported during the year but are to be settled
after year end: R23 million
• Estimated IBNR claims which occurred in the year: R30 million
170 Chapter 8 General Insurance

The insurer requires a profit margin of 15% of office premiums. Ignor-

ing interest and inflation:
1. Calculate the actual profit on the book of policies
2. If the office premium was calculated assuming expenses would
amount to R42 million, calculate what the insurer was expecting
to pay out in claims over the year.
3. If the insurer wishes to price the policies for next year based on last
year’s experience, calculate the percentage change in premiums
for next year.
Chapter 9

Life Insurance

A life insurance contract provides for the payment of a specified sum on the
occurrence of an event or events which are dependent on a human lifetime or
lifetimes.
Life insurance policies developed in response to the risks which arise due U Note that the event does not
need to be death — it could
to death and survival. If you recall the four basic financial risks we identified in
also be survival!
Module 2, life assurance assists with two of them:
• the risk of losing a family member and his/her contribution to family income,
and
• the risk of being unable to work and as a result having no money to support
oneself — particularly in the context of being too old to work. (There is an
overlap here between life insurance and pensions products.)

9.1 What are people’s needs and what types of products do

life insurers offer?
People’s financial risks relating to life and death can broadly be categorised ac-
cording to the event they concern, the form the benefits take, and the term over
which cover or payments are needed.
The events can relate to life or death:
• The risk of dying and not leaving sufficient funds to support dependants or
pay off debts;
• The risk of surviving beyond a specified time point, and as a result not
having funds to meet expenses that arise at that time (and that would not
arise if the person was dead).
In both cases, the financial need might be for different forms of benefit: a once-off
payment or regular payments such as those from an annuity.

171
172 Chapter 9 Life Insurance

Exercises for 9.1 What are people’s needs and what types of products do life in-
surers offer?

Ex.9.1 Consider the following scenarios. Are the people in the scenarios con-
cerned about dying or surviving? Would a lump sum or an annuity be
more suitable to meet their needs? If it is an annuity, how long should
it be payable for?
1. Kurt is the sole breadwinner in his family. His salary pays for the
mortgage. He is worried that if he dies, his wife will not be able to
meet the repayments.
2. Thandi has two small children. She is worried that if she dies, they
will need money for food, clothing, and school fees until they turn
18.
3. Ebrahim is worried about how he will support himself after he
stops working at the age of 65.
4. Ruth knows that if she were to pass away, her funeral would be
very expensive and a real burden on her family.
Another major distinguishing feature of life insurance policies is the term.
Life insurance policies can give cover for the whole remaining life of a person or
for a predefined term.
Based on these features, insurers have developed a wide range of products.
We discuss some of the more common products in the following subsections.

9.1.1 Products related to death

These policies pay out when the insured life dies.

Whole-life policy

This type of policy pays a benefit, the sum assured, when the insured life dies,
whenever that may be. For such a policy, the policyholder usually pays regular
monthly premiums to the insurer while they are alive.
? How many claims does the Obviously the policy ends when the policyholder dies, but the policyholder
insurer expect to pay for a
can also ‘lapse’ the policy. If a policyholder lapses a policy the policy comes to an
whole-life policy?
end and their estate will not receive the benefit on their death. The policyholder
lapses a policy by stopping the payment of premiums, but if they keep paying pre-
miums, the policy remains in force until they die and their estate or beneficiaries
will receive the benefit.
This type of policy is useful as a means of providing for funeral expenses or
tax liabilities that arise on death of the insured life, e.g. inheritance tax or estate
duties. It is a general purpose contract which provides long term protection to
the dependants of the insured life. It may also be a means of transferring wealth
on the death of the insured.
Benefits from a whole-life policy are usually in the form of a lump sum but in
some cases may be in the form of an annuity.
9.1 What are people’s needs and what types of products do life insurers offer? 173

Term assurance policy

A term assurance policy pays a benefit, the sum assured, on the death of the
insured life provided the death occurs within a specified term chosen at the outset.
If the life survives to the end of the term of the policy they will receive nothing.
For such a policy, the policyholder usually pays regular monthly premiums to the
insurer while they are alive, but at most for the length of the term of the policy.

Exercises for 9.1.1 Term assurance policy

Ex.9.2 Do you think a term assurance policy would be cheaper or more expen-
sive than a whole-life policy with the same sum assured and underlying
life? Why?

Example 9.1 A person may take out a term assurance policy such that
the policy provides protection to their children while they are young and
would require support on the death of a parent. Once the children are self
sufficient, the need for protection is lower and there may be no further need
for insurance.
? A ‘decreasing term assur-
Another example of a term assurance policy is ‘key person insurance’. A key ance’ has a sum assured
person insurance policy is a term assurance policy taken out by a company which reduces over the term
to protect them from the financial loss that might arise on the death of a of the policy. What financial
specified key person within the organisation. need do you think this could
meet?

Exercises for 9.1.1 Term assurance policy

Ex.9.3 How may the death of an employee lead to a financial loss for an or-
ganisation? What type of employees do you think companies buy key
person insurance for?

9.1.2 Products related to survival

These products pay out an amount if a person or persons are still alive on a
specified future date, or dates.

Pure Endowment

A pure endowment policy has a specified term and pays out a lump sum if the ? Could a pure endowment be
“whole-life” in term?
insured life is still alive at the end of the term of the policy. In a way, this is a form
of saving for an expense that will not be relevant if the person is dead.

Example 9.2 A pure endowment could be used to save for a holiday of a

lifetime. By using a pure endowment policy to save for such an expense you
174 Chapter 9 Life Insurance

would be acknowledging that should you die over the savings period, you
won’t really need the holiday anymore!

Example 9.3
Why would someone use a pure endowment rather than just saving for the expense
through normal investments or bank accounts?

Solution: Technically, a pure endowment will offer a higher “return” than a compa-
rable investment, because some policyholders, those who die before the term, do
not receive the benefit. You could think of their ‘savings’, i.e. their premiums with
interest, being distributed amongst the survivors.

Pure endowment insurance policies are not very common, however — most
people do not see the benefit of not receiving a payment on death as well.

Life annuities

A life annuity is a policy which provides the insured (the annuitant) a regular
payment while they are still alive. The annuity may be for the remainder of the
insured life’s life time, in which case it is referred to as a whole-life annuity. The
annuity payments may be made while the insured life is alive, but may cease after
a specified term, in which case the annuity is called a temporary annuity.
These life annuities meet a financial need for an income for the remainder of
the life of the insured. For example, a whole-life annuity may be used to provide
an income during retirement, and a temporary life annuity may used to provide
income during a limited period, for example to pay the school fees of the insured’s
children.
Life annuities may start immediately or may start at some future date. If they
start at some future date they are referred to as a deferred annuity.
A life annuity may be purchased with a single premium, or if it is a deferred
annuity, regular premiums may be used to purchase the annuity.

Example 9.4 A ‘retirement annuity’ in South Africa is a type of deferred life

annuity. The insured life pays regular contributions during their working
life and the annuity commences on the day of their retirement.

Exercises for 9.1.2 Life annuities

Ex.9.4 For each of the following pairs, assuming all other characteristics are
the same, state who is likely to be able to buy a larger annuity with the
same single premium:
1. A 60 year old or a 65 year old?
9.1 What are people’s needs and what types of products do life insurers offer? 175

2. A man or a woman?
3. A 50 year old buying a deferred annuity starting at age 60, or a 60
year old buying an immediate annuity?
4. A person buying an increasing annuity or one buying a flat annu-
ity?
U An annuity with a ‘guaran-
5. A person buying an annuity with a 5 year guarantee or a person teed term’ is guaranteed to
buying an annuity without a guarantee? make payments for the length
of the guaranteed term, re-
9.1.3 Products which pay out on both death and survival gardless of whether the life
survives or not. The annuity
Endowment assurance payments continue after the
guaranteed term if the life sur-
An endowment assurance policy is a combination of a pure endowment and a vives the term. For example,
term assurance. the estate of a 65 year old who
buys an annuity with a 5 year
Exercises for 9.1.3 Endowment assurance guaranteed term and dies at
age 66 will still receive annuity
Ex.9.5 1. When does a pure endowment pay out? payments for another 4 years
after the insured life’s death.
2. What about the term assurance?
3. If you combined a term assurance and a pure endowment with the
same terms and sum assured, when do you expect this combined
product to pay out?

Endowment assurances have a predefined term. If the insured life dies before
the end of the term, a benefit will be paid; and if the insured is still alive at the
end of the term, they will also receive a benefit. Endowment assurances provide
a combination of protection and savings. The insured life usually pays regular
premiums while they are alive during the term of the policy.

Example 9.5
How does an endowment differ from a savings plan?

Solution: If a person dies early in the term of the endowment assurance policy,
the policy will pay out the full sum assured, which is likely to be more than the
accumulated premiums paid for the policy. The savings plan will only pay out the
accumulated premiums. Similarly, if a person survives to the end of the policy,
the endowment assurance payout is likely to be lower than the accumulation of
premiums in a savings account. This is because some of the premiums were used
to pay out benefits to those who died early, i.e. there is a cost associated with
protection provided during the term of the policy.

Exercises for 9.1.3 Endowment assurance

Ex.9.6 Would the annual premium for a 10 year endowment assurance policy
be smaller or greater than that of a whole-life policy? Why? Would the
176 Chapter 9 Life Insurance

difference be greater for a person aged 40 or one aged 70? Assume the
sum assured is the same on both policies.

9.1.4 Products associated with disability

Life insurers also deal with products that pay out on disability. We will not cover
such products in detail. Examples include lump sum disability insurance, disabil-
ity income cover, and critical illness cover.

Example 9.6
What do you think may be the main differences between life assurance and disabil-
ity insurance business?

Exercises for 9.1 What are people’s needs and what types of products do life in-
surers offer?

Ex.9.7 Consider the scenarios below and recommend an insurance product

(sometimes more than one product may be appropriate). Make sure
you state the form of benefits and the term if appropriate.
1. Jim wishes to save for the tertiary education of his five year old
daughter.
2. Themba is taking a 3 year contract on an oil rig, and wants to make
sure that his wife is looked after if he dies during the contract.
3. Frances is paying off a mortgage on her house. Her husband will
not be able to afford the instalments by himself if she were to pass
away. The current mortgage balance is R800 000 and it still has 12
years to go before it is paid off.
4. Katlego wants to make sure that when he dies, his family does not
need to finance his funeral.

9.2 Accessing life insurance products

9.2.1 Distribution channels
For any company, a key business question relates to how it will get its products
to its customers. Life insurance companies have the additional challenge of
providing complex financial products that are poorly understood by potential
clients. Distribution channels for insurers are not simply sales intermediaries, but
must also act as interpreters and educators for clients. We consider four broad
mechanisms for delivering life insurance to end consumers.
9.2 Accessing life insurance products 177

Intermediaries (financial advisers or brokers)

Historically, most intermediaries
Financial intermediaries, also called advisers or brokers, aim to find the best deal have been remunerated by way of
for their clients, the end customers. To this end, they are not tied to any one life commission paid by life insurers
insurance company, but will obtain quotes from a range of insurers. on the sale of their policies. This
resulted in unscrupulous interme-
The clients serviced by intermediaries tend to be those who are the wealthiest diaries selling policies based on
and most financially sophisticated. This makes for a competitive environment maximising commission rather
where insurers can sell products with fairly complex structures. It also creates than on acting in their clients’
the potential for anti-selection, and hence underwriting tends to be fairly strict. best interests. This has led to a
raft of regulation, including leg-
Taken together with the socio-economic class selection, business written via
islated maximum commission
intermediaries tends to show lower mortality and morbidity experience than that rates for different types of policies,
written via other channels. disclosure requirements and the
licensing of financial advisers by
the regulatory body. It has also
Example 9.7 Why does the fact that intermediaries focus on the wealthiest seen the emergence of a growing
clients mean that there may be anti-selection? And what about the level of class of financial advisers who
financial sophistication — how can that lead to anti-selection? charge their clients explicit fees
for their advice and do not accept
commissions from insurers.

Agency force (tied agents) Table 9.1 Remunerating intermedi-

aries
Tied agents sell the products of only one insurance company, or in some cases,
a handful of insurers. An example may be the employees of a bank who sell
their clients the products of an insurance company with which the bank has a
relationship.
Agents are remunerated by commission paid by the insurer. T Do you know which South
African bank is affiliated
The profile of the customers will vary, and with it the complexity of the prod-
with which insurer? The
ucts that can be sold. Because the agents do not compare the prices and benefits major ones are:
of different insurers’ products, the products sold by tied agents generally do not
(a) First National Bank
have to be as competitive as those sold through intermediaries. Underwriting will and MMI (the name
also tend to be less severe. for the combined Mo-
mentum/Metropolitan
Own salesforce insurance company);
(b) Nedbank and Old Mu-
The insurance company may employ salespeople for the specific purpose of tual; and
selling their products. Their remuneration is likely to be based on a mixture (c) Standard Bank and Lib-
of salary and commission. The customer profile will vary, and once again the erty.
products do not need to be as competitive as the salesforce will not provide any
direct comparisons to competitors’ offerings. This assumes, of course, that end
consumers do not carry out their own comparisons.

Direct sales

This is a growing area of distribution driven in particular by technological devel-

opment. There are now insurers who operate entirely by selling directly to the
public, e.g. 1Life Direct and [Link]. But even traditional large insurers now
offer some of their products direct to the public. Direct sales may include:
178 Chapter 9 Life Insurance

1. Direct mail sent to prospective clients;

2. Telephone sales where call centres call prospective clients;
3. Internet sales: this is obviously the fastest-growing area of the market. Tech-
nology allows prospective clients to provide their information online and
receive quotes immediately.

9.2.2 Individual vs. group business

Usually when we discuss life insurance contracts, we are talking about contracts
sold to individuals. It is however important to realise that some insurance is
provided on a group basis. For example, an employer may wish to provide his
? The sum assured for group employees with life insurance cover, and may buy a group life assurance (GLA)
policies is often specified as
contract to do so. In this case, all employees of the company will get cover —
a multiple of salary rather
a major requirement of group cover is that the whole group must get cover, no
than a Rand amount — can
you think why? exceptions.

Example 9.8
Do you think the premiums payable under GLA are generally lower or higher than
those offered to individuals?
U A common feature of GLA is
that only people whose cover
Solution: The key to thinking about this question is the potential for anti-selection.
is greater than the “free cover
Individual cover is more susceptible to anti-selection, because each individual
limit” have to go for medical
makes their own decision about whether to buy insurance or not. Individuals
screenings. The free cover
who know that they are more at risk for some reason will be more likely to buy
limit will be a maximum sum
policies than those who think they are unlikely to die in the near future, resulting
assured acceptable without
in anti-selection. This will cause the insurer to set higher premiums for individual
medical tests, and it will be
contracts.
specified by the insurer when
Group life assurance makes cover compulsory for all individuals in the group — no
quoting for the cover.
individual anti-selection will take place. As a result, there is also far less underwrit-
ing for a group life assurance!

Group life assurance can be used in a number of ways, for example:

? There is no group version of
whole-life insurance — can • by an employer to provide a benefit to dependants on the death of an em-
you think why? ployee (this is the most common use);
• by a credit card company to provide a benefit on death equal to the balance
outstanding on a credit card; or
• by any supplier of goods with payment in instalments to cover the risk that
recovered goods are less valuable than the outstanding loan balances due
on death.

9.3 Underwriting
Underwriting for life insurance products involves the insurer gathering medical
and financial information about the prospective policyholder.
? How do you think an insurer
could ensure that informa-
tion supplied on forms is
complete and honest?
9.4 Cashflows of life insurance products 179

Medical information can be collected in the form of a questionnaire, or by

requiring a medical examination to be performed by a doctor. The exact nature of
the medical underwriting will depend on the type of policy and the characteristics
of each applicant — such as age, general health status, and the amount insured.

Example 9.9 Do you think annuity contracts have more or less stringent
underwriting requirements than whole-life insurance contracts?

While it is rare for applicants to be rejected for cover, it does happen. For T A contentious issue in life
example, if a person has been diagnosed with a terminal illness and they are trying assurance is the “right to un-
derwrite”: is it appropriate
to purchase a term assurance policy, an insurer may reject their application.
that some people are denied
A common way of managing risk is to use exclusions. The details of a life cover? For example, should
insurance contract may stipulate that the benefit is not to be paid if the cause HIV positive applicants be
of death belongs to a certain list of causes, i.e. there will be causes of death for denied life insurance cover?
which the policy does not pay the sum assured. A common example is suicide,
which is typically not covered in the first two years of life cover. ? Why do you think suicide is
initially excluded and then
Another common exclusion is deaths resulting from risky pursuits such as
covered after 24 months?
cave-diving, rock climbing, and bungee-jumping. Insurers also typically exclude
deaths if they result from a pre-existing condition, i.e. the insurer will not pay a
death benefit if the cause of the death is a condition that the insured life had and
knew about prior to taking out the insurance.

Exercises for 9.3 Underwriting

Ex.9.8 For each scenario below, think about what type of underwriting infor-
mation an insurer may require:
1. A 25 year old earning R20 000 per month buying a whole-life policy;

2. A reporter buying a 3 year term assurance policy; and

3. An unemployed man buying a term assurance policy with a R10
million sum assured.
Ex.9.9 For each of the above, do you think that the person may get refused
cover or be subject to exclusions?

9.4 Cashflows of life insurance products

Term: Life assurance products range in term, from a few years to the whole life of
an insured, although they are typically long term.
Premiums: Products can be single premium or have regular, usually monthly,
premiums. With the exception of life annuities, which are usually purchased with
a single premium, the regular monthly premium model is most common.
180 Chapter 9 Life Insurance

Premium increases
It is common for insurers to guarantee the premiums payable by a policyholder
? Which will have a lower initial only for a limited period, even where the contracts have very long terms. For
premium, a policy with no example, for a whole-life assurance policy taken out by a 25 year old, the insurer
premium escalation or one may guarantee the premiums for a period of, say, 10 years. After this period, the
where premiums are set to insurer has the right to recalculate premiums, which will usually increase reflecting
increase with inflation? the increased risk posed by the policyholder. In addition to these occasional reviews,
policyholders will often have the choice about the rate at which premiums escalate
each year, ranging from no escalation through to fixed rates or rates linked to the
inflation rate.

U Microinsurance is a new Claims: Most life insurance policies only pay out one claim, which makes them
development in life insur-
different to general insurance. Even if the benefit is paid in the form of an annuity,
ance which aims to protect
the very poor. The “micro” there is still only one claim event.
refers to the very small sum Premium term: Another unusual feature of life insurance is that the premiums
assured. There are different cease on claim event. This makes sense since in most policies, the claim event is
types of micro insurance, death — insurance companies do not require their clients to pay after they are
and these include funeral dead! But even for life annuities, the single or regular premiums are all received
and accident cover — both before any payment is made to the policyholder.
of which are types of life in-
Surrender values: Another cashflow between the insurer and the policyholder
surance. Premiums must be
very low cost to appeal to the
is a surrender value. This is a feature of some policies which is paid out if the
low income market. Some insured life decides to surrender the policy, foregoing future benefit payments but
microinsurance policies also being relieved of the obligation to pay premiums. Such features are typically
are also extremely “micro” only associated with contracts where a benefit would inevitably have been paid,
in their term: An insurer e.g. endowment assurances or whole-life assurances.
selling accident insurance
policies in Kenya has found
that there is much demand Paid up values
for accident insurance cover Sometimes the policyholder may wish to discontinue premiums but would like to
over terms as short as 24 continue to be covered. In such cases life insurers may offer the conversion of the
hours! Find out more at contract into a paid-up policy. In this case, the sum assured is adjusted downward
[Link] to reflect that the insured life will not be paying any further premiums.

? Why do you think does a

term assurance not pay a The following figures present examples of cashflow diagrams for several dif-
surrender value? ferent types of policies. In most cases, the diagram is only for one particular
outcome. See if you can think of alternative outcomes that would give rise to
different diagrams.
9.4 Cashflows of life insurance products 181

DEATH

Premium
Claim

Figure 9.1 Example of the cashflows of a whole-life policy with regu-

lar level premiums.

DEATH WITHIN TERM

TERM

Premium
Claim

Figure 9.2 Example of the cashflows of a term assurance policy with

regular premiums. In this example the insured life dies within the
term.
182 Chapter 9 Life Insurance

SURVIVAL
TO END OF
TERM

TERM

Premium
Claim

Figure 9.3 Example of the cashflows of a pure endowment policy with

regular premiums. In this example the life survives to the end of the
term and receives the benefit.

Exercises for 9.4 Cashflows of life insurance products

Ex.9.10 Draw cashflow diagrams representing the possible cashflows patterns

that could occur under an endowment assurance policy.
Ex.9.11 Draw cashflow diagrams representing the possible cashflows patterns
that could occur under a term assurance policy.
9.5 Pricing 183

Premium
Benefit

DEATH

Figure 9.4 Cashflows of a life annuity.

Exercises for 9.4 Cashflows of life insurance products

Ex.9.12 Suggest what type of policy each of the individuals below may have
purchased by considering the cashflows they experience.
1. Tatenda pays premiums of R200 per month, increasing annually
with inflation. When he dies after 4 years, his wife receives R1 mil-
lion.
2. Kim pays premiums of R100 per month for 6 years. She dies at
time 6, and she receives nothing from the insurer.
3. Dave pays a single premium of R100 000. Ten years later, Dave is
still alive, and the insurer pays him R300 000.
4. Rose pays a R1 000 per month, increasing at 10% per annum, for
30 years. At this point she stops paying premiums, and the insurer
pays her R30 000 per month until she dies at age 80.

9.5 Pricing
Pricing in life assurance is a major area of work for actuaries. In order to price a
life assurance contract, an actuary must be able to estimate the expected present
value of the future benefits, expenses, and premiums.
For a contract that pays out on death, it is important to know the probability
of a person dying in any future year. This will inform the premium calculation in
two ways — the time until death will affect the time value of the sum assured, and
it will also affect the number and present value of premiums that will be received.
184 Chapter 9 Life Insurance

For a product that pays out based on survival, like a life annuity, the expected
time until death will determine how many annuity payments to expect.
? Why is gender a risk factor? Premium rates should differentiate by risk and rating factors, e.g. gender, age,
smoker status. This is in order to ensure that the premiums charged are in line
with the risk that a policyholder represents to the insurer.

Exercises for 9.5 Pricing

Ex.9.13 What other risk/rating factors may a life insurer use when setting pre-
miums? Think of the types of questions that you may be asked when
applying for a life insurance policy. Alternatively, visit the website of
a life insurer who sells policies via the internet and see what types of
question they ask. Think of why they would need the information they
ask for.

The study of how to calculate the expected present value of payments that are
contingent on human lifetimes is called “Life Contingencies”. Chapter 11 is an
introduction to some basic techniques used to value life contingent payments
and will enable you to calculate premiums for various life insurance contracts.

9.6 Reserving
Reserving is a very important function in life insurance companies. Liabilities
are very long term, and proper reserving calculations will ensure the financial
security of the insurer, the speed with which profit emerges, and indirectly also
the pricing of the policies.

Example 9.10
Why do you think that the reserving process would influence pricing?

Solution: Consider a product with very conservative reserving assumptions. The

insurer would be asked for a substantial injection into the reserves at the start of this
policy. The more the insurer is asked to invest at the start, the more it will need to
earn in profits to justify that amount of capital being tied up. So a strong reserving
basis is likely to lead to higher profit margins being required by the insurer, which
increases the office premiums.

As before, the approach to reserving is similar to that used for pricing. How-
ever, recall that:
• The reserving basis is always stronger (more conservative) than the pricing
basis.
• There may be government requirements as to the strength of the reserving
basis, and the exact method of calculation of the reserve.
? Why is it not too important to
allow for the exact risk factors • The reserving calculations are usually done using fewer risk and rating fac-
of each policyholder when tors — the calculations are generally less accurate for individual members.
calculating a reserve?
9.7 Claims 185

Chapter 11 will give you the skills needed to calculate simple reserves for a
life insurance product.

9.7 Claims
The sum assured is defined at the outset of the policy.
However, given the long term of the policy, many policies increase the sum T Does life insurance follow
the principle of indemnity?
assured over time according to different mechanisms:
? Why is it useful to increase
Standard (non-participating) contracts the sum assured over time?

Standard life insurance contracts specify fixed benefits at the outset. These bene-
fits do not change throughout the course of the policy.

With-profits (participating) contracts

A life insurance contract is with-profits if the policyholder is entitled to receive

part of the surplus of the company or of a sub-fund within the company. The
extent of the entitlement is usually at the discretion of the company. The company
will declare bonuses, which usually involve increasing the benefits a policyholder
will receive, based on its performance.

Index-linked contracts

An index-linked contract enables the consumer to obtain a benefit that is guaran-

teed to move in line with the performance of an index specified in the contract.
Normally the index will be an investment or economic index — a commonly used
index is the consumer price index, which is a measure of inflation. Premiums
may move in line with the same index, or may be fixed in monetary terms.

Unit-linked contracts

A unit-linked benefit is determined by the performance of units whose value

is linked to the performance of specific assets. Premiums are used to buy the
units. An insurance policy which is unit-linked enables consumers either to
obtain a higher expected level of benefit for a given premium or to pay a lower
expected level of premium for a given level of benefit than under a comparable
non-linked version of the contract. This is because the consumer accepts a
significant element of the investment risk. By accepting greater risk, the consumer
gains a higher expected return.

Exercises for 9.7 Claims

Ex.9.14 Mr Moodley purchased an index-linked whole-life policy at the start

of 2007. The initial sum assured was R2 million. The sum assured is
increased on the first day of each month in accordance with the CPI
186 Chapter 9 Life Insurance

index. Mr Moodley dies during March 2013. Calculate the amount that
will be paid out to his beneficiaries using the table below, which shows
the value of the CPI index at the end of each month.

Table 9.2 CPI values. Source: Statistics South Africa, URL: [Link]
CPI/[Link].

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Average
2006 64.6 64.7 65.0 65.4 65.7 66.2 66.9 67.4 67.6 67.7 67.6 68.0 66.4
2007 68.5 68.4 69.0 69.9 70.3 70.9 71.6 71.9 72.5 73.1 73.4 74.0 71.1
2008 74.8 75.1 76.3 77.7 78.5 79.6 81.2 81.8 81.9 81.9 82.0 81.1 79.3
2009 81.4 82.3 83.4 83.8 84.1 84.5 85.4 85.6 86.0 86.0 86.0 86.2 84.6
2010 86.4 87.0 87.7 87.8 88.0 88.0 88.6 88.6 88.7 88.9 89.0 89.2 88.2
2011 89.6 90.2 91.3 91.6 92.0 92.4 93.2 93.4 93.8 94.2 94.5 94.6 92.6
2012 95.2 95.7 96.8 97.2 97.2 97.5 97.8 98.0 98.9 99.5 99.8 100.0 97.8
2013 100.3 101.3 102.5 102.9

9.8 Monitoring
Monitoring the experience in life assurance contract includes monitoring:
1. Mortality rates experienced in each product vary over time;
? Why may the mortality rates
vary over time? Why may the 2. Claims experience, in particular any concentration of claims or the payment
mortality rates of the insured of very high sum assures;
population vary from the
3. The costs actually incurred by the products compared to the budgeted costs;
overall mortality rate in a
country? 4. The actual level of new business sold and of policies lapsed compared to the
expected levels.

New business levels

The risk of selling too little business is clearly understood: new business is required
for a sustainable insurance company, and short-term pressure will be felt in that
fixed costs must be spread over a smaller number of policies than projected. But
there is also a risk of selling too much business. A feature of life insurance business
is that it involves upfront costs which take some time to recover from premiums;
this is known as “new business strain”. These costs relate to acquiring the business
(marketing and sales costs such as commission) and the administration costs of
loading a new policy.
Selling significantly more business than expected can put huge pressure on operat-
ing costs because of incurring more of these upfront costs than anticipated. Note:
these high volumes may still be desirable in the long run, provided the short-term
costs can be financed.
Likewise, if the mix of new business by type of policy differs dramatically from what
is expected, this can put strain on the insurer. Different classes of business have
9.8 Monitoring 187

different levels of new business strain, so a higher proportion of policies with higher
new business strain will increase operating costs beyond what was expected. Many
expenses may also be catered for by loading premium rates with an expense charge,
which may (in part at least) be a percentage loading on the premium rate, which is
in turn a function of the sum assured. If the average size (sum assured) of the new
policies turns out to be lower than expected, then this loading may be insufficient.

The risk of lapses

Insurers will try to ensure that surrender terms are such that there is no strain as a
result of policies lapsing (i.e. policyholders stopping the payment of premiums, in
which case cover ceases). However, in some cases this may be impossible, especially
at very early durations when initial expenses have not been recouped. If lapses in
these areas, where the lapse implies the crystallisation of a loss for the insurer, are
higher than expected, then this puts pressure on profits.
Insurers must also beware of “selective withdrawals”, where healthier lower risk
policyholders are more likely to lapse their policies, leaving behind an unhealthier
higher risk pool of policies. It’s obvious that, assuming the premiums can be
afforded, policyholders in very poor health will not allow their policies to lapse,
given the high probability of getting a benefit paid from them. This leads to a group
with potentially worse overall mortality than anticipated, with the obvious negative
effects on profitability.
Chapter 10

Reinsurance

Insurance is a mechanism for individuals and companies to transfer risks to an ? Can you think of situations
where an insurer might want
insurer. Insurers make a business out of accepting risks and pooling them. But
to take out reinsurance for
sometimes insurers want to transfer some of the risk that they have accepted.
some of its policies?
Reinsurance has developed to meet this need; it is insurance for insurers. Rein-
surance is a contractual arrangement under which an insurer secures cover from a U An insurer is also called the
reinsurer for potential losses to which it is exposed under insurance policies it has direct writer because they
issued. The insurance company pays reinsurance premiums to the reinsurance write policies directly to the
company, and in exchange the reinsurer covers a portion of the claims that the public — unlike reinsurers
who only deal with insur-
insurer has to pay out. The insurer has ceded some risk to the reinsurer.
ance companies.

10.1 Why do insurers need reinsurance? U Cede means “pass on” or

“give away”, as in “cede some
risk to a reinsurer”. Cession
The main reasons why insurance companies take out reinsurance are:
refers to a risk that has been
• to smooth claims experience; ceded.
• to limit large losses;
? What does the reinsurer get
• to access the reinsurer’s expertise. out of a reinsurance arrange-
Each of the above helps the insurer run its business more safely, and allows the ment?
insurer to expand its business. These reasons are discussed further below.

10.1.1 Smoothing of claims experience

Insurers use risk pooling to reduce the volatility of claims. In an ideal world, U Remember the volatility of
the total claims an insurer would experience over time would have very little claims is very high for an
individual, but may be much
variability. However, in reality insurers can still have claims that are volatile from
lower for large groups of
year to year.
people due to risk pooling.

Example 10.1
List the reasons why an insurer’s claim experience may be volatile.

Solution: Volatility could be caused by:

189
190 Chapter 10 Reinsurance

• a single very large claim;

• a catastrophe leading to a large number of connected claims, e.g. a large
earthquake in an urban area;
• cumulative events leading to a number of claims across different policies, e.g.
a fire claim for building damage being followed by an interruption of business
claim from the same policyholder;
• a concentration of risks in one geographic area leading to a large number of
claims arising from that area, e.g. if a motor insurer sells policies in only one
area — a large hail storm in that area could result in a large number of claims;
• the insurer having too few policies to pool the risk sufficiently;
• the insurer having too little expertise to price the risk effectively, resulting in
higher than expected claims.

Reinsurance can reduce the volatility of claims. Reducing the claims volatility
means that the insurer can have a more predictable claims experience, which is
important for the stability of the company. The insurer will also be able to use
lower margins for risk when calculating its reserves, and as a result the reserve
requirements will be lower. Lower reserve requirements will allow the insurer to
write more business as it will have more capital available.

10.1.2 Limiting large losses

In 1994, Ladysmith experienced Insurers can be financially crippled by a very large single claim, or by a large
severe floods which resulted in number of smaller claims. In extreme circumstances, this could cause an insurer
damage of R60 million, and led to to go bankrupt as they cannot afford the payments.
thousands of domestic property
insurance claims.
Reinsurance helps protect an insurer from becoming insolvent as a result of
The Titanic, which sank in 1912, large losses. It can also allow the insurer to write policies for risks that are too
was insured for £1 million — large for it to handle alone, knowing that the reinsurer will accept part of the risk.
about R1.5 billion in today’s terms. This helps insurers grow their business.
Imagine what effect one such
claim could have on an insurer!
10.1.3 Accessing the reinsurer’s expertise
Table 10.1 Examples of large Reinsurers provide another service to insurers: their experience and skills in
losses. the pricing and assessing of large risks. A reinsurer will offer its clients a wide
range of services, ranging from international data on various risks, to its own
pricing and underwriting expertise. For example, an insurer launching a new
type of product may be able to get assistance from its reinsurer who has worked
with such products. Reinsurers develop their expertise by working with many
insurers in many different countries, and as a result they experience a wide range
of scenarios which a single insurer may not have to deal with.

Retrocession
Even reinsurers sometimes want to cede some of their risk, i.e. reinsurers will pass
on some of the risk they face to other reinsurers. This is called “retrocession”. It is
reinsurance for reinsurers, and is a very specialist area of the reinsurance market.
10.2 Ways of writing reinsurance business 191

The ceding reinsurer in a retrocession agreement is called the retrocedant, and

the reinsurer who takes on the risk is called the retrocessionaire.
Some well known reinsurers include Swiss Re, General Re, Munich Re, NRG, and
SCOR. Most of these companies will engage in some retrocession.

10.2 Ways of writing reinsurance business

Reinsurance may be written on either a treaty or facultative basis.

10.2.1 Treaty reinsurance

A treaty contract is entered into at the beginning of a specified period, and it
applies to a given class or given classes of business over that period. Once the
treaty is in place, both the insurer and the reinsurer are bound by it: every single
claim that is made on the specified book of business is automatically divided
between the insurer and the reinsurer in an agreed way. The terms are agreed
upfront: the amount of premiums paid by the insurer to the reinsurer, the type
of underwriting that must be done, how the claims are shared, etc. The treaty
document is usually complex and very specific as to the terms of the arrangement,
but after the treaty is in place, the reinsurance process is effectively automatic
and there is no further discretion needed.

Example 10.2 Cashflows in a reinsurance treaty

Suppose that a general insurer enters into a reinsurance treaty with a rein-
surer. The treaty is for the year 2014, and it covers all of the insurer’s house-
hold insurance policies, both existing and new policies that will be taken
out during the year. The treaty sets out how much the insurer will pay to the
reinsurer for each policy it has on its books, and exactly how much of every
claim the reinsurer has to cover.
In 2014, the claims start being reported. The insurer pays the claims to the
policyholders, but then presents the reinsurer with reinsurance claims of
its own under the terms of the treaty. The reinsurer settles these claims,
and effectively reimburses the insurer a portion of the amounts it paid out
to its policyholders. Figure 10.1 illustrates the cashflows involved in the
reinsurance treaty.

Figure 10.1 Illustration of the cashflows involved in a reinsurance

treaty agreement.
192 Chapter 10 Reinsurance

Advantages of treaty reinsurance

• Risks are reinsured automatically. This is administratively simpler, quicker

and cheaper.
• The insurer knows that reinsurance will be available if the risk falls within the
limits of the treaty, and they know the terms of the reinsurance agreement
for such risks.
• As a result the insurer can issue new policies instantly without having to find
a reinsurer who is willing to accept the risk. This can improve their standing
in the insurance market.

Disadvantages of treaty reinsurance

• Once the treaty is set up, then both parties must operate within the terms of
the treaty, i.e. the reinsurance agreement may constrain the way the insurer
and reinsurer operate.
• The insurer is bound to pay premiums, and only the specific types of risks
outlined in the treaty will be reinsured automatically. For other risks not
covered by the treaty, the insurer will have to approach a reinsurer and
negotiate the terms of a reinsurance arrangement.
• The terms of the reinsurance treaty are agreed at the outset and the insurer
may find that they are paying too much for the reinsurance cover, and
similarly the reinsurer may find they are charging too little for the cover they
provide. Neither party can change the terms of the treaty without the other
party agreeing.

Treaty reinsurance is a good way to insure large books of similar policies, where
the individual policyholders are relatively similar to each other.

Exercises for 10.2.1 Treaty reinsurance

Ex.10.1 For each of the classes of business below, state whether it is suitable or
not for treaty reinsurance and why:
1. motor vehicle insurance,
2. small business insurance,
3. marine insurance, and
4. celebrity body part insurance.

10.2.2 Facultative reinsurance

A facultative agreement is a much less specific contract. It states that for a specific
type of risk, the reinsurer is prepared to consider providing cover, but it does not
obligate either party to provide or request cover, and it does not set out the terms
under which cover will be provided. It is up to the reinsurer to propose a price
U A facultative agreement is and specific terms to the insurer for each risk it wants reinsured. The insurer
essentially an agreement to
do business, but only if both
parties want to at the time.
10.3 Types of reinsurance 193

does not have to accept this proposal. However, it can choose not to reinsure, or
approach another reinsurer to look for a better price.
This form of reinsurance is mostly used for large risks that cannot be accepted
automatically, but that have to be individually evaluated. Insurers often have
a combination of treaty and facultative arrangements: the treaty to cover all
“standard” risks, and a facultative arrangement to cover everything that does not
fall under the treaty.

Advantages of facultative reinsurance to the insurer

• A facultative arrangement provides the insurer with flexibility and choice.

• The insurer is under no obligation to use a particular reinsurer, and so
they can choose who to reinsure with. The insurer can approach several
reinsurers in search of the best terms for each risk individually

Disadvantages of facultative reinsurance to the insurer

• It can be a time-consuming and costly exercise to negotiate the terms a risk

will be reinsured on.
• The insurer has no certainty that the required cover will be available when
they want to reinsure a risk.
• Even if reinsurance cover is available, the price and terms may be unaccept-
able.
• The insurer may be unable to accept a large risk until it has been able to
find the required reinsurance cover. This means the insurer cannot accept
large risks automatically, and consequently its standing in the market may
be reduced.

10.3 Types of reinsurance

Reinsurance is generally classified as either proportional reinsurance or non-
proportional reinsurance. Under proportional reinsurance, the reinsurer covers
a proportion of each claim. Under non-proportional reinsurance, the reinsurer
covers the part of the claim (or claims) exceeding some agreed amount.

10.3.1 Proportional reinsurance

Under proportional reinsurance, the reinsurer covers an agreed proportion of T Proportional reinsurance is
each risk. This proportion may be constant for all risks covered, this is called also called original terms
reinsurance.
quota share, or the proportion may vary across risks and risk classes, e.g. indi-
vidual surplus. Both forms are administered automatically, i.e. via a treaty.
Proportional reinsurance reduces the proportion of the claim that a ceding
company pays, and so it is used mostly as a means of accepting larger risks than
would otherwise be possible. The premiums received from the policyholders and
194 Chapter 10 Reinsurance

the claims paid to policyholders are shared in the same proportion as specified in
the contract, and therefore the reinsurer and the direct writer will have the same
underwriting experience.

Exercises for 10.3.1 Proportional reinsurance

Ex.10.2 Proportional reinsurance will cover a defined proportion of every claim.

Does this type of reinsurance protect the direct writer from very large
claims?

Quota share reinsurance

Under quota share reinsurance, the reinsurer assumes an agreed percentage of

each risk. The reinsurer shares in every claim under the agreement, no matter how
small. For example, if there is a R1 000 loss under a 40% quota share reinsurance
contract (sometimes referred to as a 60/40 quota share), the insurer (also called a
cedent) would bear 60% of that loss, i.e. R600, and the reinsurer would bear 40%
of that loss, i.e. R400. The percentage stated in the quota share agreement always
refers to the percentage of loss borne by the reinsurer.
U The portion of the risk that
the reinsurer assumes is
called the “ceded risk,” and Example 10.3
the portion that the cedent Under a 30% quota share agreement, determine the amount paid by the insurer
keeps is referred to as the and reinsurer for the following claims
“reinsurance retention.”
(a) R50 000
(b) R100 000
(c) R1 500 000

Solution:
(a) Reinsurer’s portion = 30% × R50 000 = R15 000
Insurer’s portion = 70% × R50 000 = R35 000
(b) Reinsurer’s portion = 30% × R100 000 = R30 000
Insurer’s portion = 70% × R100 000 = R70 000
(c) Reinsurer’s portion = 30% × R1 500 000 = R450 000
Insurer’s portion = 70% × R1 500 000 = R1 050 000
10.3 Types of reinsurance 195

R 20 000 000

R 17 500 000
Reinsurer's Portion
Insurer's Portion
R 15 000 000

R 12 500 000
Sum Assurred

R 10 000 000

R 7 500 000

R 5 000 000

R 2 500 000

R0
Risks

Figure 10.2 Split between insurer and reinsurer payments for claims
on a 30% quota share reinsurance agreement.

A quota share contract can sometimes be subject to a quota share limit. A ? Why would a reinsurer want
quota share limit is the maximum rand amount that the reinsurer is willing to to include a quota share
pay for any claim. So for a 25% quota share with a limit of R600 000, the reinsurer limit as part of the reinsur-
will pay R125 000 of a R500 000 claim, but will only pay R600 000 of a R3 000 000 ance agreement?
claim.

Example 10.4
Under a 50% quota share agreement with an upper limit of R500 000, determine
the amount paid by the insurer and reinsurer for the following claim amounts
(a) R750 000,
(b) R1 500 000, and
(c) R3 000 000.

Solution:
(a) Reinsurer’s portion = min(50% × R750 000, R500 000) = R375 000
Insurer’s portion = R750 000 − R375 000 = R375 000
(b) Reinsurer’s portion = min(50% × R1 500 000, R500 000) = R500 000
Insurer’s portion = R1 500 000 − R500 000 = R1 000 000
(c) Reinsurer’s portion = min(50% × R3 000 000, R500 000) = R500 000
Insurer’s portion = R3 000 000 − R500 000 = R2 500 000

Exercises for 10.3.1 Proportional reinsurance

Ex.10.3 An insurer has a quota share treaty with a reinsurer. The following two
claims are settled as follows:
196 Chapter 10 Reinsurance

1. a R900 000 claim receives a payment of R300 000 from the reinsurer,
and
2. a R2 000 000 claim receives a R500 000 payment from the reinsurer.
What is the percent quota share under the contract, and what is the
quota share limit?

Individual surplus reinsurance

Individual surplus reinsurance is similar to quota share reinsurance in that premi-

ums and losses are shared on a proportional basis. However, the proportion paid
on each policy is not the same, but it depends on the original size of the policy.

Estimated maximum loss

The “size” of the policy is measured by using the estimated maximum loss (EML)
on the policy. The EML on a policy can be very obvious (for example, in life assur-
ance, the EML is always the full sum assured), or it can be more difficult to estimate.
For example, on a household fire policy, the EML may not be the whole property
value, because the property value includes land that is not damaged by fire. Instead,
the EML may be the estimated cost of rebuilding the house.
The EML is sometimes also called the policy limit.

Individual surplus reinsurance is best explained by using an example.

Example 10.5 Consider an insurer that has a book of marine insurance

policies. The insurer wishes to enter into an individual surplus reinsurance
agreement. The book includes the following policies:
• a small fishing boat with an EML of R350 000,
• a ferry with an EML of R1 000 000, and
• a passenger ship with an EML of R4 000 000.
The insurer first has to choose a retention limit which is the maximum
amount it is willing to pay for any claim. In this case, the insurer chooses a
retention limit of R500 000.
Any policy where the EML is smaller than the retention limit is now excluded
from the treaty, i.e. the insurer retains 100% of the claims on these policies
because they are expected to be below the retention limit. For the other
policies, the retention limit is used to calculate the proportion each party
will pay in the event of a claim.
Let’s consider each policy in turn.
• The small fishing boat has an EML of less than R500 000. This policy
is not included in treaty; the insurer will cover 100% of the claims that
arise on it.
10.3 Types of reinsurance 197

• The ferry has an EML of R1 000 000. The retention limit is R500 000,
and that means that if the largest estimated claim occurs, the insurer is
willing to cover R500 000 out of the R1 000 000 claim. The retention limit
is used to determine the proportion the insurer and the reinsurer share
the claims. For this policy, the insurer retains R500 000/R1 000 000 =
50%, and reinsurer takes 50%. That means that for any claim that occurs,
the reinsurer will cover 50% of it.
• The passenger ship has an EML of R4 000 000. The insurer retains
R500 000/R4 000 000 = 12.5% of this policy. This means that whatever
the size of the claim, the insurer will pay 12.5% of it and the reinsurer
will cover the other 87.5%.
Now suppose that the policies experience the following claims:
• the fishing boat sinks and the claim is R350 000;
• there is a fire on the ferry which causes R50 000 of damage;
• the passenger ship sinks, and the replacement cost turns out to be
R4 500 000.
What are the insurer’s and reinsurer’s shares of each of the claims?
• The fishing boat is 100% covered by the insurer, so the R350 000 is paid
by the insurer.
• The insurer retains 50% of the claims from the ferry, and so the insurer
pays R25 000 and the reinsurer pays R25 000.
• The claim is higher than the EML, but that does not matter. The rein-
surer is still liable for 87.5% of the claim. Therefore, the insurer pays
R562 500, and the reinsurer pays R3 937 500 of the claim.
Keep in mind that the EML is just an estimated maximum loss at the start
of the policy — as time goes on, the claims may grow, but the proportions
set at the start of an individual surplus policy stay fixed. The retention limit
is the maximum that the insurer thought it would pay given a certain EML.
But if the real claim turns out to be higher than the EML, the insurer will end
up paying more than the retention limit.

So for each policy where the EML is lower than or equal to the retention limit,
the insurer retains 100% of the claims. For policies where the EML is greater than
the retention limit, the proportion retained by the insurer is calculated as
retention limit
retention % =
EML
and the proportion ceded to the reinsurer is
EML − retention limit
ceded risk % = .
EML
These proportions are fixed, and are used to determine how all claims on the
policy are shared.
198 Chapter 10 Reinsurance

Example 10.6
Consider a life insurance company that has four policies with the following sums
assured:
(a) R500 000,
(b) R1 000 000,
(c) R2 000 000, and
(d) R5 000 000.
The life insurance company takes out an individual surplus reinsurance contract
with a retention limit of R750 000. Determine the proportions of each claim paid by
the insurer and reinsurer on each of these policies if a claim were to occur.

Solution: Note that since the policies are life insurance policies, any claim will be
equal to the sum assured, and so the EML is equal to the sum assured.
(a) The sum assured is below the retention limit and would therefore not be
covered by the individual surplus reinsurance contract. The insurer will cover
the full claim amount.
750 000
(b) The retention % = = 75%, and the ceded risk % = 25%.
1 000 000
750 000
(c) The retention % = = 37.5%, and the ceded risk % = 62.5%.
2 000 000
750 000
(d) The retention % = = 15%, and the ceded risk % = 85%.
5 000 000

Example 10.7 Figure 10.3 illustrates the portions of a claim paid by the
insurer and the reinsurer. The example used in the figure is of a life insurance
company that has a individual surplus reinsurance treaty in place with a
retention limit of R2 000 000.
10.3 Types of reinsurance 199

R 20 000 000

R 18 000 000
Reinsurer's Portion
Insurer's Portion
R 16 000 000

R 14 000 000
Sum Assurred

R 12 000 000

R 10 000 000

R 8 000 000

R 6 000 000

R 4 000 000

R 2 000 000

R0
Risks

Figure 10.3 Individual surplus reinsurance example for life insurance

business with a retention limit of R2 000 000. Because the EML on
life assurance policies is the sum assured, the insurer will always pay
the full R2 million for all policies with sums assured greater than R2
million.

It is vital to remember that with individual surplus reinsurance contracts, the

proportion retained by the direct writer gets set on the day that the policy is taken
out and not when the claim is paid. The proportion is then applied to the claim,
even if the final claim is not the same as the EML and even if that proportion
means that the insurer ends up paying more than the retention limit.

Example 10.8
Consider a life insurance company that has an individual surplus reinsurance
agreement with a retention limit of R1 000 000. The company has just sold a whole-
life assurance contract to Mario with an initial sum assured of R2 000 000 that
increases by 10% p.a., and is paid immediately upon death. Determine the amounts
paid by the insurance company and the reinsurer if Mario dies:
(a) in 6 months;
(b) in 3.5 years;
(c) in 5 years.

Solution:
Firstly note that the
1 000 000
retention % = = 50%,
2 000 000
and ceded risk % = 50%, i.e. each claim is split 50/50.
(a) The sum assured in the first year is R2 000 000. Therefore both the insurer and
reinsurer will pay R1 000 000 in 6 months.
200 Chapter 10 Reinsurance

(b) The sum assured in the fourth year is R2 000 000×1.13 = R2 662 000. Therefore
both the insurer and reinsurer will pay R1 331 000 in 3.5 years.
U Notice that the amount paid
(c) The sum assured in the fifth year is R2 000 000×1.14 = R2 928 200. Therefore
by the insurer exceeds the
both the insurer and reinsurer will pay R1 464 100 in 5 years.
retention limit.

One disadvantage of individual surplus reinsurance for insurers is that the

? Is this more of a problem claim is not capped at the retention level. Therefore, theoretically the insurer is
for companies that issue still exposed to unlimited risk under this type of agreement.
life insurance or general
insurance?
Exercises for 10.3.1 Proportional reinsurance

Ex.10.4 An insurer has decided to reinsure its building by using an individual

surplus agreement. The EML is set at R200m. The maximum loss
the insurer can sustain is R55m, but they decide to set the surplus
level (retention) at R50m to allow for a margin of safety. Calculate the
proportion paid by the insurer and reinsurer under the following loss
scenarios:
1. a loss of R2 million from a lighting strike.
2. the roof of building is ripped off in a hurricane, and R30m in
damage is sustained.
3. an earthquake occurs and as the building was built prior to the
earthquake building standards being implemented, much of the
building collapses and the assessors have declared the building
as unfit for use, and therefore must be knocked down and rebuilt.
The estimated cost to do this will be R250m.

10.3.2 Non-proportional reinsurance

Proportional reinsurance can be used to spread risk and to reduce the proportion
of each risk retained. However, this does not cap the cost of very large individual
or aggregate claims that may occur.

Exercises for 10.3.2 Non-proportional reinsurance

Ex.10.5 Consider a general insurer who insures property. Under what circum-
stances could this insurer experience:
• a very large individual claim?
• a very large aggregation of claims?
Ex.10.6 What about a life insurer?

The reinsurance solution to these types of scenarios is called excess of loss

(XL) reinsurance. XL reinsurance can apply to the losses from individual claims
(risk XL), or from aggregations of claims (aggregate XL).
10.3 Types of reinsurance 201

Risk XL

Under risk XL reinsurance, the cost of any single claim to an insurer is capped at
a certain level called the excess point. The claim amount above the excess point
is paid by the reinsurer. That is, for a claim which is larger than the excess point,
the insurer pays the part of the claim below the excess point, and the rest of the
claim is paid by the reinsurer. Any claim amount under the excess point is paid
by the insurer in full. T Do you recall how an excess
in general insurance works?
This is the same idea.
Example 10.9
An insurer has a risk XL agreement with a per-claim excess point of R500 000. The
following claims are made:
(a) R100 000,
(b) R650 000, and
(c) R1 000 000.
Find the amounts of each claim paid be the insurer and the reinsurer.

Solution: The split between the insurer and the reinsurer, for each claim, will be as
follows:
• Insurer R100 000; Reinsurer R0;
• Insurer R500 000; Reinsurer R150 000;
• Insurer R500 000; Reinsurer R500 000.

An additional feature of XL policies is the upper limit. If a contract has an

upper limit, the portion of a claim above the upper limit is again payable by the
insurer. Figure 10.4 illustrates how an insurer and a reinsurer will share a claim
under a risk XL reinsurance contract with an upper limit. The insurer will pay the
parts of the claim below the excess point and above the upper limit. The reinsurer
will pay the remainder of the claim.

Figure 10.4 The portions of a claim to paid by an insurer and a rein-

surer under a risk XL reinsurance contract with an upper limit.
202 Chapter 10 Reinsurance

Example 10.10
Suppose that in Example 10.9, an upper limit of R800 000 is introduced. What are
the amounts payable by each party for each of the 3 claims now?

Solution:
• Insurer R100 000; reinsurer R0;
• Insurer R500 000; reinsurer R150 000;
• Insurer R500 000 + R200 000 = R700 000; reinsurer R300 000.

It is not uncommon for an insurer to use several layers of XL reinsurance

to spread the reinsurance between several reinsurers. The upper layer of XL
reinsurance usually has no upper limit or an unlimited upper limit.

Example 10.11
An insurer has a risk XL agreement with reinsurer A, with a per-claim excess point
of R250 000 and an upper limit of R500 000. The insurer also has another risk XL
agreement with Reinsurer B, with a per-claim excess point of R500 000 and an upper
limit of R1 000 000.
(a) Draw a diagram to illustrate the split of claims between the insurer and the 2
reinsurers.
(b) State the amount payable by each party for each of the following claims:
i. R175 000;
ii. R380 000;
iii. R510 000;
iv. R3 000 000.

Solution:

(a)

Figure 10.5 The split of claims between the insurer and the 2 reinsur-
ers.
(b) i. Insurer: R175 000.
10.3 Types of reinsurance 203

ii. Insurer: R250 000; Reinsurer A: R130 000.

iii. Insurer: R250 000; Reinsurer A: R250 000; Reinsurer B: R10 000.
iv. Insurer: R2 250 000; Reinsurer A: R250 000; Reinsurer B: R500 000.

Example 10.12 Consider an insurer with the following risk XL agreements:

• an XL agreement with reinsurer 1 that has an excess point of R1 250,000
and an upper limit of R4 000 000, and
• an XL agreement with reinsurer 2 that has an excess point of R4 000 000
and an upper limit of R9 000 000.
Figure 10.6 illustrates the break down of claim payments made by the insurer
and the two reinsurers. Note that the excess of claims over the upper limit
of R9 000 000 revert back to the insurer.
R 20 000 000

R 17 500 000
Additional amount covered by Insurer
Reinsurer 2's Portion
Reinsurer 1's Portion
R 15 000 000
Insurer's Portion

R 12 500 000
Sum Assurred

R 10 000 000

R 7 500 000

R 5 000 000

R 2 500 000

R0

Risks

Figure 10.6 Break down of the claim payments for an insurer with a
risk XL agreement with an excess point of R1 250 000 and an upper
limit of R4 000 000 with reinsurer 1, and a risk XL agreement with
an excess point of R4 000 000 and an upper limit of R9 000 000 with
reinsurer 2.

Aggregate XL

Aggregate XL is similar to risk XL, except that instead of calculating the reinsurer’s
portion of each claim, the agreement covers the total amount claimed under the
conditions of the agreement. The reinsurer is required to pay if the aggregated
claims over a specified period, usually one year, exceed the excess point.
The claims can be aggregated in several different ways:
204 Chapter 10 Reinsurance

• Aggregation by event refers to claims caused by a single event, such as a

hurricane or flood. An example of this type of XL agreement is catastrophe
XL. For example, if a general insurer has a catastrophe XL treaty with a
reinsurer with an excess point of R10 million. If one of the catastrophes
listed under the agreement occurs, the insurer will calculate the total of all
the claims caused by that catastrophe and if that total exceeds R10 million,
the reinsurer will pay for the amount exceeding R10 million.
• Aggregation by cause refers to claims that are caused by the same peril.
This could be fire, for example, or car accident. As for catastrophe XL, the
insurer calculates the total claims that are caused by certain perils and as
soon as this total exceeds the excess point, the reinsurer starts reimbursing
the insurer for any claims above the excess point.

Example 10.13
Suppose a general insurer has an aggregate XL reinsurance agreement in place. The
agreement relates to claims caused by fire. The agreement covers household and
business insurance policies. The excess point is R12 million and the upper limit
is R20 million. The policy covers all claims which occur during the 2013 calendar
year.
The claims that occur during the year are as follows:
For household insurance:
• claims due to theft: R5 million,
• claims due to flooding: R2 million, and
• claims due to fire: R8 million.
For business insurance:
• claims due to fraud: R15 million,
• claims due to theft: R8 million, and
• claims due to fire: R11 million.
Will the insurer be able to claim under its aggregate XL agreement, and if so, what is
the amount?

Solution: The total claims resulting from a fire are R8 million from households and
R11 million from businesses. The total is R19 million. This exceeds the excess point
by R7 million. It does not exceed the upper limit. So the total amount paid by the
reinsurer is R7 million.

Another way to aggregate claims is by class of business. Here, all claims that
occur on a certain type of policy, or even on all policies of an insurer, regardless
of the cause, are covered. This type of aggregate XL is called Stop Loss. It is a
powerful contract because a large number of policies are generally covered, and
the excess point will generally be high for this type of agreement. An example
would be for an insurer to take out stop loss reinsurance on all of its motor
insurance policies. All motor insurance claims would count towards the total,
10.3 Types of reinsurance 205

regardless of whether they were accidents, thefts, fire or anything else that was
covered under the motor policies.
In Stop Loss, it is common for a reinsurer to demand that the insurer covers
part of the claim made to the reinsurer as well. This is to prevent the insurer from
writing business that is a bad risk. Just like in other types of XL insurance, there
can also be an upper limit.
Figure 10.7 illustrates how a stop loss arrangement would work. In the ex-
ample used to construct the figure, the stop loss arrangement requires that the
insurer pay part of the claim made to the reinsurer, and the arrangement has an
upper limit.

Direct Writer Share of Claim

Upper
Limit

Direct
Writer’s
Reinsurer’s Share of Claim
Claim Share of
Claim

Excess
Point

Direct Writer Share of Claim

Figure 10.7 Break down of a Stop Loss Payment.

Due to the potential for morale hazard, this type of cover is usually only U For example, if an insurer
offered where losses are quite variable and poor experience is not the insurer’s had stop loss cover for dis-
fault, e.g. crop insurance. ability insurance, once
the excess point has been
breached there is no incen-
tive for the insurer to stop
admitting more claims be-
cause the reinsurer will have
to cover them. This is an
example of morale hazard.
206 Chapter 10 Reinsurance

Exercises

Exercises for 10 Reinsurance

Ex.10.7 Define the following terms:

• Cedent • EML
• Retention • XL
• Policy limit • Upper limit
• Excess point • Direct writer
• Original terms reinsurance • Retention limit
• Ceded risk • Excess point
• Retrocession

Ex.10.8 List the reasons why an insurance company would choose to reinsure
some of their business.
Ex.10.9 Discuss the disadvantages, if any, of reinsuring your business.
Ex.10.10 Compare the pros and cons of having a treaty as opposed to a faculta-
tive reinsurance arrangement.
Ex.10.11 Explain the difference between insurance, reinsurance, and retroces-
sion.
Ex.10.12 You are a medium sized life insurance company who writes only
mortality (death) business.
Your company has a reinsurance arrangement with ReinsuranceRUs.
Under the treaty you keep 35% of each risk, up to a maximum retention
of R800 000. How much would you pay under each of the following
death claims?
1. R500 000,
2. R800 000,
3. R2 500 000, and
4. R10 000 000.
Ex.10.13 Discuss the purpose of catastrophe XL reinsurance.
Ex.10.14 Explain the difference between proportional and non-proportional
reinsurance and give an example of each.
Ex.10.15 What type of insurance is best suited to help an insurer reduce the
volatility of claims?
Chapter 11

Life Contingencies

11.1 Introduction
When actuaries calculate premiums and reserves for life insurance policies, they U “Contingencies” comes from
“contingent”, which means
need to be able to estimate the probability that a person will be dead or alive at
“dependent”. Life contin-
some future date. This field of study is called “life contingencies”. gencies therefore refers to
The uncertainty about future lifetimes is something life assurance companies events which depend on the
and pension funds have to worry about all the time. They need to place values future life time/s of a life or a
on payments that depend on the future lifetime of a policyholder or member. group of lives.
This chapter introduces the basic ideas behind valuing payments which are
contingent upon, i.e. depend on, the future lifetime of a life. U In life insurance, a person is
often referred to as a “life”.

Example 11.1 A payment may be made on the survival or death of a life.

For example:
• A whole-life policy pays out the sum assured on death;
• A pure endowment policy pays out the sum assured on survival to the
end of the term;
• A life annuity pays out regular amounts as long as the life survives to In this chapter we only consider
each payment date; one decrement, death. In reality, a
life can leave a policy or pension
• Premiums for life assurance policies are payable on survival to the fund for several reasons and these
premium payment date. are called decrements.
For example, a life may lapse a
whole-life policy. And in a pen-
In life assurance, the “uncertainty” component of risk arises from the uncer-
sion fund, a member may retire
tainty of the length of a life’s future lifetime. Life tables are a convenient means early, retire on ill-health grounds,
of summarising the probability distribution associated with a life’s future lifetime. retire at the normal age, die, and
They allow us to calculate useful probabilities. withdraw.
You will consider multiple decre-
ments in much more detail in a
11.2 Life tables later course.

In order to be able to estimate the probability that a person will die or survive to a Table 11.1 Multiple decrements.
certain age, actuaries use information collected about people dying in the past.

207
208 Chapter 11 Life Contingencies

After some processing, this information is often summarised in tables called “life
tables”. Life tables summarise the probability distribution of future lifetimes for a
population by age and gender.
? How do you expect the prob-
ability of dying changes with
age? What about gender —
11.2.1 l x and d x
do you think men or women Below is an extract from the English Life Table No. 12 (Males). It is based on the
have higher mortality?
1961 census of the male population of England and Wales, and the male deaths
T Although several of the life which occurred in those countries in the years 1960–62.
tables we use for illustra- x lx dx
tive purposes are very old,
0 100000 2449
they have the advantage
1 97551 153
that they appear in the old
book of Formulae and tables 2 97398 96
for actuarial exams pub- 3 97302 67
lished by the Faculty and 4 97235 60
Institute of Actuaries and 5 97175 55
available free in electronic .. .. ..
form to [Link] rele- . . .
vant parts of that book are 40 93790 220
available to you at the Vula 41 93570 242
site for this course. See also 42 93328 268
Appendix 11.9. 43 93060 297
44 92763 330
T The starting value, l 0 , in
45 92433 369
this table is 100 000, but any
.. .. ..
starting value can be chosen. . . .
This starting value is called 95 573.54 185.74
the radix of the table; for
96 387.80 130.39
ELT 12 (Males) the radix is
l 0 = 100 000. The table above
97 257.41 89.59
was actually based on a 98 167.82 60.30
much larger number of peo- 99 107.52 39.771
ple, so that is why it shows 100 67.749 25.733
decimal places—obviously
this does not mean a half a
It is convenient to think of a life table as the record of a population of l 0 = 100 000
person is alive! babies born on the same day. An l x refers to the number of these babies that are
expected to survive to the age of x. So out of the 100 000, 93 328 are expected to
still be alive at age 42, l 42 = 93 328, and only 573.54 are expected to still be alive at
age 95.

More formally:
l x is the number of lives who are expected to survive (at least) to age x, out of the
specified starting number.

For example, l 4 = 97 235 is the number out of the starting 100 000 expected to
survive to age 4, and similarly l 5 = 97 175 is the number expected to survive to age
5.
11.2 Life tables 209

Exercises for 11.2 Life tables

Ex.11.1 Answer the following questions:

1. How many of the original 100 000 are expected to be still alive at:
1. Age 1?
2. Age 3?
3. Age 99?
2. What is l 2 ?
3. What is l 44 ?

The second column, d x , represents the expected number of people who die
each year in this population.

More formally:
d x is the number expected to die between age x and age x + 1 (out of the starting
number, or equivalently out of the l x lives expected to survive to age x).

So for example, the number of people who are expected to die between the
ages of 43 and 44 (d 43 ) is 297, and d 4 = 60 is the number of children expected to
die between ages 4 and 5 out of the original 100 000.
Can you see that each consecutive l x+1 is the result of subtracting the deaths
in the previous year, d x , from the people alive at the start of the previous year, l x ?

In other words,
the link between the quantities l x and d x is just:

d x = l x − l x+1 .

This link enables you to derive the l x column from the d x values given an
initial value l 0 , and similarly we could find the d x s from the l x s.

Exercises for 11.2 Life tables

Ex.11.2 Confirm for yourself that in the table above the number of people alive
at age 42 is 93 328. Use the number of people alive at age 41 and the
number of deaths aged 41.
Ex.11.3 Below is another extract from ELT 12 (Males), with some entries re-
moved. Calculate the missing entries.
210 Chapter 11 Life Contingencies

x lx dx
60 ... 1805
61 77119 1947
62 75172 ...
63 73084 ...
64 70856 ...
65 ... 2499
66 65991 2625
Check your answers by referring to Appendix 11.10.
Ex.11.4 Use ELT 12 (Males) in Appendix 11.10 to calculate:
1. The number of people expected to die between ages 40 and 43 (i.e.
at ages 40, 41, and 42)
T These types of calculations
can be done in two different 2. The number of people expected to die before age 2?
ways, one of which is much 3. The number of people expected to die before age 61?
quicker than the other! 4. The number of people expected to die between age 3 and age 40?
5. The number of people expected to die before age 101?
6. The total of all the d x s in the ELT 12 (Males) table up to age 100?

11.2.2 p x and q x
A life table does not just contain the number of lives at each age. Other values are
generally also already worked out for you to make your calculations easier.
Here is another extract from ELT 12 (Males), now with two new columns:

x lx dx px qx
70 54806 3051 0.94433 0.05567
71 51755 3130 0.93952 0.06048
72 48625 3195 0.93429 0.06571
73 45430 3243 0.92862 0.07138
74 42187 3273 0.92242 0.07758
75 38914 3282 0.91566 0.08434

How are p x and q x derived from l x and d x ? Based on that, what do you think
p x and q x represent?

More formally,
q x is the probability that someone now aged exactly x will die within one year:

dx
qx = .
lx

Similarly, p x is defined to be the probability that a person aged exactly x survives at

least one year:
l x+1
px = = 1 − qx .
lx
11.2 Life tables 211

The p x s and q x s are usually tabulated in the life table. If not, they can easily " The notation involved in life
be calculated from the d x s and l x s. contingencies can be quite
dense. It is important that
you are happy with it and
Exercises for 11.2.2 p x and q x
understand the definitions.
Ex.11.5 Use l x and d x in the ELT 12 (Males) extract above to show that p 73 =
0.92862 and that q 71 = 0.06048.
Ex.11.6 If d 69 = 2959 and l 69 = 57 765, calculate p 69 and q 69 .

The p x s and q x s are probabilities: the probability of survival to age x + 1 or

death between ages x and x +1, given the life is aged x now. We need probabilities
to help us with the “uncertainty” part of the risk calculations — p x and q x will U Note that for each x, p x and
q x sum to 1 — can you ex-
help us calculate the expected present value of payments contingent on survival
plain why this is?
and death in the next sections.

11.2.3 k px and k q x
We can also define k p x as the probability someone now aged exactly x will survive ? Can you see that
at least k years, and k q x as the probability a person aged exactly x now dies
k px = 1 −k q x ?
within k years’ time.
What if we want to calculate k p x ? How would you go about it?
The expected number of people alive at age x is l x . Out of this group, l x+k
are expected to survive to age x + k. So we can calculate the probability that the
person survives from age x to age x + k as:
" Useful formula:
l x+k
k px = for k = 0, 1, 2, . . . . l x+k
lx k px =
lx
For example, using ELT 12 (Males),:
U Note that p x is the same as
l 41 93 570
p 40 = = = 0.99765; 1px .
l 40 93 790
l 43 93 060
3 p 40 = = = 0.99222;
l 40 93 790
l 53 87 868
3 p 50 = = = 0.97539.
l 50 90 085
What about k q x ? You could calculate k q x using the sum of all the deaths
between ages x and x + k. For example, what is the probability that a person aged
40 dies before they turn 43? This could be calculated using the sum of the deaths
aged 40, 41, and 42, divided by the number of people alive at age 40:
d 40 + d 41 + d 42 220 + 242 + 268
3 q 40 = = = 0.00778.
l 40 93 790
But an easier method is to use the relationship between k p x and k q x :

3 q 40 = 1 −3 p 40 = 1 − 0.99222 = 0.00778.
212 Chapter 11 Life Contingencies

In other words, the probability that someone aged x dies before age x + k is 1
minus the probability that they are still alive at age x + k.
We can now calculate the probabilities of all sorts of events from the life tables.
Consider the following example.

Example 11.2
What is the probability of a life currently aged 50 exact dying between ages 60 and
T Note that when we say “be-
65? Use the ELT12 (Male) life tables.
tween age 60 and 65”, we
mean “between the 60th and
Solution: Loosely speaking, if the life currently aged 50 were to die between ages
65th birthdays” — i.e. while
60 and 65, their death would be counted amongst those lives dying between those
they are 60, 61, 62, 63, or 64
ages, i.e. their death would be counted in d 60 , d 61 , d 62 , d 63 , and d 64 . The probability
years old.
of a life aged 50 dying between ages 60 and 65 is then

d 60 + d 61 + d 62 + d 63 + d 64
P(Life aged 50 dies between 60 and 65) =
l 50
l 60 − l 65
=
l 50
78 924 − 68 490
=
90 085
= 0.115824.

Another way of thinking about this question is to reason along the following lines.
For a life currently aged 50 to die between ages 60 and 65, they first need to survive
for 10 years and then die within the following 5 years. Thus we can write

P(Life aged 50 dies between 60 and 65) = 10 p 50 5 q 60

l 60 l 65
µ ¶
= × 1−
l 50 l 60
l 60 − l 65
= .
l 50

The reasoning above was quite informal, but we can see that this approach arrives
at the same answer.

Exercises for 11.2.3 k p x and k q x

Ex.11.7 Which of the following is least likely to occur to a 40 year old according
to the A1967–70 (ultimate) mortality table1 ?
1. The life dies before reaching age 42.
2. The life dies between age 41 and age 42.
3. The life survives to age 99.
1 This is the life table based on the mortality experience of assured lives in the United Kingdom

during the years 1967–70. Ignore for the purposes of this course ‘select’ mortality, and any symbols
like l [x] , l [x]+1 , d [x] , ä [x] , etc., which have the age x enclosed in square brackets to indicate that
they refer to select mortality. We will use only quantities like l 40 , l 41 , d 40 , ä 40 , which refer to what
is known as ‘ultimate’ mortality; much simpler! These values can be found in Appendix 11.9.
11.3 Curtate future lifetime 213

4. The life dies between age 97 and age 99.

Ex.11.8 On the basis of ELT 12 (Males):
1. What proportion (of the original 100 000 lives) are expected to die
in the first year of life?
2. What proportion (of the original 100 000) are expected to die in
the first two years of life? Equivalently, what is (1 – the proportion
expected to survive the first two years of life)?
3. What proportion of those reaching 40 are expected to survive to
42?
4. What proportion of the original 100 000 are expected to die be-
tween 40 and 42?
Ex.11.9 You can now work out the probability of your own survival. Use the
different life tables to check how likely you are to be still alive at 80.
What about your chances of living to 99? Note that the tables are old,
? Improving mortality means
and mortality is improving all the time. Your actual chances of living
that for a given age, the risk
that long are much better now than for people in the 1960s. of a person dying that age is
decreasing over time. Can
you think of reasons why
11.3 Curtate future lifetime mortality has been improving
over time?
11.3.1 Defining K x
Before we start using the life tables to calculate expected present values of pay-
ments, we need to introduce one more concept: the curtate future lifetime.
Simply said, the curtate future lifetime of a life is the number of full years that U We use curtate lifetimes be-
cause they are easier to work
the life will live from now until they die. So if a person aged 20 dies at age 44 and
with — our life tables are tab-
8 months, their curtate future lifetime is 24 years. A life’s complete future lifetime ulated in whole years, and
is the total time they survive for. In our example the life’s complete future lifetime the maths is simpler if the
is 24.67 years. ages are whole years.
We denote the curtate lifetime of a life by K , or by K x if we wish to emphasise
that the person concerned is now aged x. T The notation [. . . ] is often
used for ‘integer part of’,
e.g. [5.75] = 5, but another
Example 11.3 For instance, if someone takes out a life assurance policy and notation you might meet
dies 5 years and 9 months later, we say that the curtate future lifetime, from is the ‘floor notation’ b. . .c,
the inception of the policy, is K = 5. (The complete, as opposed to curtate, which means the same thing.
future lifetime is 5.75 years.) In Excel, you would be using
the function TRUNC() to
calculate this!
We can describe the curtate future lifetime as the complete future lifetime
rounded down, or as the ‘integer part of’ the complete future lifetime.

11.3.2 K x as a discrete random variable

Consider a person now aged x years exactly, often denoted just by (x). Let
K (or K x ) denote the curtate future lifetime of that person. This implies that T So we can say “(31) buys a
whole life assurance policy”,
or “(57) is receiving an annu-
ity”.
214 Chapter 11 Life Contingencies

death occurs between K x and K x + 1 years from now and so the age at death lies
between x + K x and x + K x + 1. We do not know in advance what K x will be: it is
uncertain. All we know is that it takes only non-negative integer values (0, 1, 2, . . .).
We therefore treat K x as a discrete random variable.
? What is the sample space Can we find from our life tables the probability that someone now aged 40 will
of K 50 , the curtate future
die within one year? That is, P(K 40 =0)?) Yes, for example using ELT 12 (Males):
lifetime of a person now
aged 50? I.e. what values can d 40 220
P(K 40 =0) = = = 0.00235.
K 50 take on? l 40 93 790
Similarly,
d 41 242
P(K 40 =1) = = = 0.00258,
l 40 93 790
d 50 656
P(K 40 =10) = = = 0.00699, etc.
l 40 93 790
l 45 92 433
and P(K 40 ≤ 4) = 1 − = 1− = 0.01447.
l 40 93 790
All of the probabilities we have found in the previous section can be expressed
in terms of K x . For example, in Example 11.2, we were calculating the probability
that a life aged 50 exactly will die between ages 60 and 65. This could be written
as P(10 ≤ K 50 < 15).
In general, the l x column of a life table contains all the information needed to
specify the probability mass function of K x , which is as follows:
d x+k
P(K x =k) = (k =0, 1, 2, . . .)
lx
= 0 (otherwise).
We can now use this discrete random variable K x , together with the time value
of money, to calculate the expected present value of payments contingent on a
lifetime. You have already met the idea of an expected value in Chapter 3, and the
result that for any function g (X ) of a discrete random variable X :
X
E(g (X )) = g (x)p X (x). (11.1)
x

This result will be very important in our study of life contingencies.

Exercises for 11.3 Curtate future lifetime

Ex.11.10 The table below is called the BUS1003H students’ mortality table.
Complete it using the information supplied.
• P(K 21 ≥ 1) = 0.999423
• The probability that a student aged 20 dies between ages 21 and
23 is 0.001148.
x lx dx qx px
20 9982.2006 0.000582
21
22
11.4 Expected present values of lump sum payments 215

11.4 Expected present values of lump sum payments

Now that you are able to calculate the probability of someone being alive or dead
at a certain age, we can calculate the expected present value of payments contin-
gent on life or death. We begin with products that pay a lump sum benefit. One of
the simplest products that is dependent on life and death is a pure endowment.
U Recall that a pure endow-
ment is a policy which pays
11.4.1 Pure endowments out the sum assured if the
insured life is alive at the end
Example 11.4 of the policy term.
Consider a life currently aged 37 exactly. Suppose that the life will receive R500 in
one year’s time if they are alive. Using an interest rate of 5% p.a. and the ELT 12
(Male), what is the expected present value of this payment now?

Solution: To receive the payment in 1 year’s time the life needs to be alive, so we
need to calculate the probability that they survive to at least age 38. This is
l 38 94 176
P(K 37 ≥ 1) = p 37 = = = 0.99819.
l 37 94 347
To calculate the EPV, we also require the present value of the payment:
EPV = 500 × v 5% × p 37 + 0 × v 5% × q 37 = R475.33.
The zero is included to reflect that there is no payment if the life were to die during
the year. We generally omit the zero term, but it is useful to have here to remind
you that the expected value includes all possible outcomes.

From the above, you can see that the expected present value of a pure endow-
ment has the usual two components, uncertainty and value. The uncertainty is
the probability that the person survives to the end of term, and the value is the
present value of the sum assured, discounted from the date of payment to today.

Example 11.5
A person aged 45 has a pure endowment that will pay R1 million if he survives to
age 65. Calculate the present value of this benefit at 5% p.a.

Solution: We need the probability that he will be alive in 20 years:

l 65 68 490
P(K 45 ≥ 20) =20 p 45 = = = 0.74097.
l 45 92 433
It looks like his chance of making it to age 65 is just less than 75%.
Now we need to calculate the present value of the benefit payable at age 65:
20
R1 000 000 × v 5% = R376 889.48.

So the EPV of the benefit is

R376 889.48 × 0.74097 = R279 263.47
If this policy was purchased now for a single premium, the risk premium (before
expenses and profit) of this policy would therefore be R279 263.
216 Chapter 11 Life Contingencies

We can conclude that the EPV of the benefit from an n-year pure endowment
policy with sum assured of R1, sold to a life aged x and valued using an interest
rate of i % p.a., is:
EPV =n p x × v in% .
The notation A x:n1 is used for the expected present value of the benefit from a
pure endowment policy sold to a life aged x and with a term of n years, i.e.

A x:n1 ≡n p x × v n .

11.4.2 Whole-life policies

A whole-life policy is a life insurance product which pays the policyholder’s es-
tate or beneficiaries the sum assured on the death of the policyholder. We will
assume that the sum assured is paid at the end of the policy year in which the
policyholder dies. The policyholder will usually pay regular premiums in advance
while they are alive to the life insurer for a whole-life policy.
The diagram below illustrates the cashflows, from the point of view of the
policyholder, of a whole-life policy with a sum assured S and an annual premium
of P .

−P −P −P −P −P ··· −P +S
-
time (yrs): 0 1 2 3 4 ··· K 6K + 1
death

? How would you alter this Consider a policyholder aged exactly x years who takes out a whole-life policy
sketch to represent the point
and whose estate will be paid R1 at the end of the (policy) year of death. This
of view of the life assurance
means that R1 will be paid to the policyholder’s estate at time K x + 1. If we knew
company that provides the
insurance? what K x was, this would be a simple calculation — the present value is simply
v K x +1 . But K x is a random variable, and can take on a range of different values.
U For convenience we take the What we are looking for is the expected present value, that is E v K x +1 .
¡ ¢
sum assured to be R1 here, In principle we can calculate it as follows:
but in practice it would of
course be much larger. E v K x +1 =
¡ ¢
vP(K x =0) + v 2 P(K x =1) + v 3 P(K x =2) + . . .
= v q x + v 2 1 p x q x+1 + v 3 2 p x q x+2 + v 4 3 p x q x+3 + . . .
∞
v k+1 k p x q x+k .
X
=
k=0

You can see that this would be tedious to calculate by hand. Fortunately this has
been tabulated for many life tables, and is denoted by A x .
That is,

A x ≡ expected p.v. of R1 payable at the end of the year of death of (x).

Have a look at the mortality tables listed in Appendix 11.9. You will notice that
" It is important to ensure that there is a column labelled A x for all ages x. This column is the expected present
you use the correct tables
when calculating such ex-
pected present values. In
particular, you should make
sure you are using the table
with the correct interest rate.
The tables in Appendix 11.9
11.4 Expected present values of lump sum payments 217

value of a whole-life insurance benefit of R1 paid at the end of the year of death
to a policyholder currently aged x.

Example 11.6 For instance, if a life aged 50 takes out a whole-life policy
now, with a death benefit of R10 000 payable at the end of the year of death,
the expected present value of that payment is 10 000A 50 . If the interest rate
is 4% and the mortality is that of the A1967–70 (ultimate) table, the expected
present value of the death benefit is 10 000 × 0.38450 = R3 845.

Exercises for 11.4.2 Whole-life policies

Ex.11.11 Consider a life (65) subject to the mortality of the A1967–70 (ultimate)
table, who takes out a whole-life policy with sum assured of R2 000 000
payable at the end of the year of death. The relevant interest rate is 4%
p.a. What is the expected present value of the sum assured?

11.4.3 Endowment assurance policies

An endowment assurance policy pays out the sum assured to a policyholder’s T Recall that an endowment
estate at the end of year of death, provided the death occurs within the term, or is a combination of a term
assurance and a pure endow-
the sum assured is paid out at the end of the term if the policyholder survives to
ment.
the end of the term. Premiums on an endowment policy are generally payable
annually or monthly, in advance, as long as the policyholder is alive and the term
of the policy has not expired.
218 Chapter 11 Life Contingencies

Below are two diagrams illustrating the potential outcomes of an endowment

assurance where the policyholder, initially aged 40, pays a premium of P at the
start of each year for 20 years (if alive), and the sum assured of S is paid at the end
of year of death, or on survival to age 60, whichever is earlier.
In the first diagram, the policyholder dies between K and K + 1 years from the
inception of the policy. The sum assured S is paid at time K + 1 and premiums
cease at that time.
−P −P −P −P ··· −P +S
-
time (yrs): 0 1 2 3 ··· K K +1
6 ··· 20
death

In the second diagram, the policyholder survives beyond the end of the term
of the policy and receives the sum assured at time 20. In this case the policyholder
pays all 20 premiums.
−P −P −P −P ··· −P +S
-
time (yrs): 0 1 2 3 ··· 19 20 6
death

Now consider the general example, where (x) who has purchased an endow-
ment assurance policy with a sum assured of R1 and a term of n years. The
payment of the sum assured will happen at time K x + 1 if K x < n, i.e. if the poli-
cyholder dies within n years, or at time n if K x ≥ n. So the present value of the
payment is v min(K x +1,n) .
We can define W as a random variable representing the present value of the
benefit of this policy, and then we can write:

v K x +1
½
if K x < n
W= .
vn if K x ≥ n

W is obviously dependent on K x .
The expected present value of the benefit is

E(v min(K x +1,n) ) = E(W ) = vP(K x =0) + v 2 P(K x =1) + v 3 P(K x =2) + . . . +
+v n P(K x =n − 1) + v n P(K x ≥ n)
n−1
X ³ k+1 ´
n
= v k p x q x+k + v n p x .
k=0

We define this value as follows:

A x: n ≡ expected p.v. of R1 payable at end of year of death of (x),

or at time n if sooner.

Again these values are tabulated for certain x +n, but it is important to know how
" For example, A1967–70 (ulti- to calculate the above quantity for values x + n that are not tabulated.
mate) gives A x: n for x + n =
55, 60, 65, and 70.
11.4 Expected present values of lump sum payments 219

Example 11.7 Example of an endowment assurance policy:

Consider again a life (50) subject to the mortality of the A1967–70 (ultimate) table,
who takes out an endowment assurance policy with sum assured of R10 000 payable
at the end of the year of death or at age 65, whichever is sooner. The relevant interest
rate is 4%. What is the EPV of the benefit?

Solution: The EPV of the benefit is R10 000A 50: 15 = 10 000 × 0.57711 = R5 771.
? Why is the endowment ben-
efit (R5 771) worth so much
more than the corresponding
Example 11.8 whole-life insurance benefit
in Example 11.6 (R3 845)?
Consider a life (31) subject to the mortality of ELT 12 (Males), who takes out an
endowment assurance policy with sum assured of R35 000 payable at the end of the
year of death or at age 35, whichever is sooner. The relevant interest rate is 9.5%.
What is the EPV of the benefit?

Solution:
A 31:4 is not tabulated, so we are going to have to do this the long way.

A 31:4 = vP(K 31 =0) + v 2 P(K 31 =1) + v 3 P(K 31 =2) +

+v 4 P(K 31 =3) + v 4 P(K 31 ≥ 4).
d 31 d 32 d 33 d 34 l 35
= v + v2 + v3 + v4 + v4
l 31 l 31 l 31 l 31 l 31
115v + 122v 2 + 129v 3 + 137v 4 + 94 652v 4
=
95 155
= 0.696104

The EPV is therefore: R35 000 × 0.696104 = R24 363.66.

11.4.4 Term assurance

A temporary assurance policy, also referred to as a term insurance policy, is a life
assurance policy which pays out at the end of year of death if the policyholder
dies within a specified term agreed at inception.

Example 11.9
A very simple term assurance contract for a life currently aged 37 exactly pays R500
in one year’s time if the life dies during the year. Using an interest rate of 5% p.a.
and the ELT 12 (Male) life table, what is the expected present value of this payment
now?

Solution: The payment will be made in 1 year’s time if the life has died during the
coming year, so we need to calculate the probability that they die before turning 38.
This is
l 38 94 176
P(K 37 =0) = q 37 = 1 − = 1− = 0.001812.
l 37 94 347
220 Chapter 11 Life Contingencies

Now the expected present value of the payment is

EPV = 500v 5% × q 37 + 0 × v 5% × p 37 = R0.86.

The zero is again included to remind you that there is no payment on survival to
U Can you see how similar this
the end of the year.
is to the example at the start
of Section 11.4.1?
Now let us look at a term longer than one year. Consider a life (x) who has
a term insurance policy with a sum assured of R1 and a term of n years. The
death benefit will be paid at the end of year of death, provided the death is before
the end of the term, i.e. at time K x + 1 provided K x < n. Nothing is paid if the
policyholder survives to the end of the term. The present value of this benefit can
then be written as: ½ K x +1
v if K x < n
Z= .
0 if K x ≥ n
Z is a random variable that depends on K x . It is similar to an endowment assur-
ance, except there is no benefit paid on survival to the end of the term.
The expected present value of the benefit is

E(Z ) = vP(K x =0) + v 2 P(K x =1) + v 3 P(K x =2) + . . . + v n P(K x =n − 1)

n−1
X r +1
= v r p x q x+r .
r =0

1
This expected present value is denoted by A x:n , i.e.

n−1
1
v r +1 r p x q x+r .
X
A x:n =
r =0

These values are usually not tabulated, but there is a link between A x:n , the ex-
1
pected present value of an endowment assurance benefit, and A x:n , the expected
present value of a term insurance benefit:
? What is the intuitive reason-
ing for this result? 1
A x:n = A x:n +n p x v n . (11.2)

" Useful formula: So we can use the tabulated A x:n values, along with the l x s, to find values for
1
A x:n = 1
A x:n n
+n p x v . A x:n .

Example 11.10
Consider a life (61) subject to the mortality of the A1967–70 (ultimate) table, who
takes out a term insurance policy with a sum assured of R500 000 and a term of 4
years. The sum assured is payable at the end of the year of death provided the death
occurs within 4 years. The relevant interest rate is 4% p.a. What is the expected
present value of the benefit?

Solution: A 1 is not tabulated. The easiest way to find the value is to first find
61:4
A 61:4 and then from this we can find A 1 .
61:4
11.4 Expected present values of lump sum payments 221

It is easy just to read A 61:4 = 0.85838 off the table on page 249.

l 65 4
A1 = A 61:4 − v
61:4 l 61
27 442.681 4
= 0.85838 − v
29 606.239
= 0.85838 − 0.79234
= 0.06604.

We can now calculate the expected present value the benefits:

EPV = 0.06604 × 500 000 = R33 021.50.

We can also work out the value of A 1 from first principles:

61:4

3
A1 v r +1 r p 61 q 61+r
X
=
61:4
r =0
3 l 61+r d 61+r
v r +1
X
= ×
r =0 l 61 l 61+r
d 61 v + d 62 v 2 + d 63 v 3 + d 64 v 4
=
l 61
474.101291v + 517.087295v 2 + 562.418523v 3 + 609.951190v 4
=
29606.239
= 0.066044179
= 0.06604.

This is obviously more work. It’s a good thing that so many values are tabulated!

1
Another way of calculating A x:n is to use A x :

1
A x:n = A x −n p x v n A x+n .

Intuitively, the above can be explained as follows: a term assurance policy is

equivalent to a whole-life assurance policy less all the cashflows which occur after
the end of the term of the term assurance policy.

Exercises for 11.4.4 Term assurance

Ex.11.12 Show
1
A x:n = A x −n p x v n A x+n .

Ex.11.13 Suppose that a man aged 20 years now is considering taking out a life
insurance policy. He is considering a term assurance or an endowment
assurance policy. The sum assured is R100 000 and the term is 45 years
on both policies. Using A1967–70 (ultimate) mortality tables and an
interest of i = 0.04, find the expected present value of the benefits
of each policy. Which policy do you think will have higher annual
premiums?
222 Chapter 11 Life Contingencies

Ex.11.14 A life aged 45 now purchases a term assurance policy with a sum
assured of R10 000 and a term of 5 years. If this policy were to be
purchased for a single premium, find this premium (ignoring expenses
and profits) for a life subject to A1967–70 (ultimate) mortality, and
using an interest rate of 4% p.a.
Ex.11.15 Consider a life aged exactly 62, subject to A1967–70 (ultimate) mortal-
ity. Find from first principles (i.e. by using tables of l x and or d x ) the
EPV of the benefits from:
1. an endowment assurance policy;
2. a pure endowment; and
3. a temporary assurance policy.
In all cases the sum assured is R100 000, the term is three years, and
the relevant interest rate is 5% per annum.

11.5 Expected present values of series of payments

" When you think of annuities, So far we have dealt with a single sum assured payable on survival, death or
the first thing that comes to a combination of those. Now we need to consider series of payments payable
mind is products which pay while someone is still alive — life annuities.
a regular income to some-
one. But don’t forget that the
premiums paid by policy-
holders for a life insurance
product also take the form
of an annuity with payments
going from the policyholder
to the insurer!
 11.5 Expected present values of series of payments 223

11.5.1 Whole-life annuities

A whole-life annuity is an annuity which makes regular payments while a life is
alive and stops once they die.
Below is a diagram of annuitant who receives a payment of R100 at the start
of each year provided he or she is still alive on that date.

+R100 +R100 +R100 +R100 +R100 ··· +R100

-
time (yrs): 0 1 2 3 4 ··· K 6K + 1
death

If a policyholder buys a whole-life annuity from a life assurance company, ? Describe the cashflows that
they will typically pay over a large initial sum to receive regular payments while happen when you buy a
they are alive. whole-life annuity. If you
struggle to remember, have a
Suppose that (x), an annuitant initially aged x, receives an annuity of R1
look at Section 9.1.2.
annually in advance, as long as he or she is alive. A total K x + 1 annuity payments
will be paid to the annuitant. (Have a look at the above diagram again to confirm T A person who is receiving an
that K x + 1 payments will occur.) Again, if we knew what the value of K x was, we annuity is referred to as an
could calculate the present value of these payments as ä K x +1 . annuitant.
But we do not know what value K x is going to take since it is a random variable.
What we really need is the expected present value of those annuity payments.
One way to think of the present value of a life annuity is that it is the expected
value of the sum of all possible annuities that might be paid to a person, multiplied
by the probability of each of them occurring: an annuity of 1 payment only,
multiplied by the probability that the person dies in the first year; plus an annuity
of two payments, multiplied by the probability that the person dies in the second
year; and so forth. If we write this out we have:
³ ´ ∞
X
E ä K x +1 = ä k+1 P(K x =k)
k=0
= ä 1 P(K x =0) + ä 2 P(K x =1) + ä 3 P(K x =2) + . . .
= ä 1 × q x + ä 2 ×1 p x q x+1 + ä 3 ×2 p x q x+2 + . . .
X∞
= ä k+1 k p x q x+k .
k=0

This would be extremely

³ tedious
´ to calculate by hand, but again for many life
tables, the values of E ä K x +1 are tabulated — or can be calculated using a com-
puter program or spreadsheet. The quantity ä x is defined as:

ä x ≡ expected p.v. of R1 p.a. in advance while (x) is alive.

Example 11.11
Suppose that (50) pays a premium of R100 p.a. in advance for life, and is subject to
the mortality of the table A1967–70 (ultimate). Suppose also that present values are
224 Chapter 11 Life Contingencies

calculated here at the interest rate i =0.04. What is the expected present value of
these payments?

Solution:

100ä 50 = 100 × 16.003 = R1600.30.

Similarly, but for ages 30 and 90 respectively:

100ä 30 = R2106.10

100ä 90 = R352.40.

Exercises for 11.5 Expected present values of series of payments

Ex.11.16 Suppose that (37) pays a premium of R900 p.a. in advance for life, and
is subject to the mortality of the A1967–70 (ultimate) table. Suppose
also that present values are calculated here at the interest rate i =0.04.
What is the expected present value of these payments?
Ex.11.17 For x = 10, 20, 35, 50, and 99, find the expected present value of
R1 000 paid annually in advance to (x) while they are alive. Suppose
(x) is subject to the mortality of the A1967–70 (ultimate) table and that
the interest rate is i =0.04. What do you notice?

An alternative way to find the EPV of life annuities

Another way of thinking of whole life annuities is as the sum of the expected
present values of single payments paid at the start of each future year if the
annuitant is still alive:

ä x = P(K ≥ 0)v 0 + P(K ≥ 1)v 1 + P(K ≥ 2)v 2 + . . .

∞
P(K ≥ k) v k
X
=
k=0
∞
vk.
X
= k px
k=0

The intuitive explanation of this method is that a payment of R1 is made at the

start of each year if the annuitant is alive at the time. The present value of the
first payment is 1 (because it is paid at time 0, and the person is alive then). The
expected present value of the second payment is 1 p x × v 1 , because the probability
of being alive to receive the payment is 1 p x and the present value of the payment
is v 1 . The expected present value of the third payment is 2 p x × v 2 , and so forth.
In a later course, you will prove this more formally using the definition of ä x .
11.5 Expected present values of series of payments 225

A handy link between A x and ä x : A x = 1 − d ä x

Recall from Module 1 that

i
d≡ .
1+i
This quantity allows us to find a very useful link between A x and ä x .
For convenience we first repeat some notation. With K denoting the curtate
future lifetime of a life now aged x:
¡ ¢
ä x = E ä K +1 = expected p.v. of R1 p.a. in advance while (x) is alive;
¡ K +1 ¢
Ax = E v = expected p.v. of R1 payable at end of year of death of
(x).
Since we know that v K +1 = 1 − d ä K +1 (can you see why?), taking expectations of
both sides gives us the following:

E v K +1 = E 1 − d ä K +1 = 1 − d E ä K +1 .
¡ ¢ ¡ ¢ ¡ ¢

Hence,
A x = 1 − d ä x , " Useful formula:

which is a very handy link between A x and ä x . So if you have a table of values of A x = 1 − d ä x .
ä x , you don’t really need a table of A x .
Check for yourself that, using A1967–70 (ultimate) mortality and at 4%, this
formula gives A 50 = 0.3845 — the figure we used in Example 11.6.

Expected curtate future lifetime

When you tell people that you are studying actuarial science, they often expect you
to be able to work their expected future lifetime. We are now in a position to do this
— that is, predict the expected, not actual, curtate future lifetime of an individual.
The method for doing this is similar to the annuity derivation at the start of this
section. We are looking for the expected value of K x , the curtate future lifetime of a
person. The expected value of K x is
∞
X
E(K x ) = kP(K x =k)
k=0
= 0 × P(K x =0) + 1 × P(K x =1) + 2 × P(K x =2) + . . .
= 0 +1 p x q x+1 + 2 ×2 p x q x+2 + . . .
X∞
= k k p x q x+k
k=0

Basically, we are summing the potential number of years a person can live, multi-
plied by the probability that they live exactly this number of years. This value is also
tabulated and it is called e x , i.e.

e x ≡ expected curtate future lifetime of a life aged x.

Use e x to look up your own expected curtate future lifetime in the A1967–70 (ulti-
mate) table.
226 Chapter 11 Life Contingencies

11.5.2 Temporary annuities

A temporary annuity makes regular payments for as long as a life is alive but stops
if the life survives beyond the end of a specified term. A common example of a
temporary annuity is the premiums payable for any insurance policy that has a
finite term.
So policyholder (x), who has a term assurance policy, pays RP p.a. in advance
while he or she is alive, but for at most n years. In this case the number of
premiums paid is K x + 1, provided that number does not exceed n, i.e. min(K x +
1, n) — the smaller of K x + 1 and n. The easiest way to consider the expected
present value of such annuities is by way of an example.

Example 11.12
Suppose that the policyholder (62) pays premiums of R1 p.a. in advance while he
or she is alive, but for 3 years at most. Assume that (62) is subject to the mortality
of the A1967–70 (ultimate) table and i = 0.04. What is the expected p.v. of these
premiums?

Solution: Remember that at most three premiums are paid annually in advance. So
if the policyholder survives 2 years or more, i.e. if K 62 ≥ 2, exactly 3 premiums are
paid. It follows that in that case the present value of premiums is R1× ä 3 . Therefore,
at interest rate of 4% p.a., the expected p.v. of the premiums is:

ä 1 P(K 62 =0) + ä 2 P(K 62 =1) + ä 3 P(K 62 =2) + ä 3 P(K 62 =3) + ä 3 P(K 62 =4) + . . .
= ä 1 P(K 62 =0) + ä 2 P(K 62 =1) + ä 3 P(K 62 ≥ 2)
= 1 × (d 62 /l 62 ) + 1.96154 × (d 63 /l 62 ) + 2.88609 × (l 64 /l 62 )
517.08730 + 1.96154 × 562.41852 + 2.88609 × 28052.632
=
29132.138
= 2.83477

Fortunately this sort of thing is also tabulated, but only for specific values of
x + n. That is, we have tables of:
" As the values of ä x: n are
only tabulated for specific
ä x: n ≡ expected p.v. of R1 p.a. in advance while (x) is alive,
x + n, it is important that you
understand how to calculate but for at most n years.
them in case the value you
need is not tabulated. For example, ä 62: 3 = 2.835 in tables based on A1967–70 (ultimate) mortality and
4% interest, which confirms our calculation above.

Exercises for 11.5.2 Temporary annuities

Ex.11.18 Write down a formula for ä x: n in terms of ... p ... s, q ... s, and ä ... s.
Ex.11.19 Prove that
" Useful formula:
A x: n = 1 − d ä x: n .
A x: n = 1 − d ä x: n .
Hint: in the above proof for A x and ä x , replace K + 1 by min(K + 1, n).
11.5 Expected present values of series of payments 227

Ex.11.20 R100 is paid annually in advance to (60) while they are alive but for
at most 5 years. Assume that (60) is subject to the mortality of the
A1967–70 (ultimate) table and i =0.04.
Find the expected present value of this annuity by first principles
as done above.
Check your answer using the tables in Appendix 11.9.
Ex.11.21 R100 is paid annually in advance to (59) while they are alive but for
at most 3 years. Assume that (59) is subject to the mortality of the
A1967–70 (ultimate) table and i =0.05. Find the expected present value
of this annuity.
Ex.11.22 R500 is paid annually in advance to (37) while they are alive but for
at most 3 years. Assume that (37) is subject to the mortality of the ELT
12 (Males) table and i =0.08. Find the expected present value of this
annuity.
Ex.11.23 A life insurance company sells a two-year endowment policy to a life
aged exactly 33. The basic sum assured of R10 000 is payable at the
end of the year of death or on survival to age 35. What is the expected
present value of the benefit from this policy at the outset? Assume the
life is subject to mortality in ELT 12 (males) and that the interest rate is
6% p.a.

Alternative ways of calculating the EPV of temporary annuities

Again, there is an easier way of calculating the expected present value of tempo-
rary annuities:

Example 11.13 Consider again Example 11.12, and calculate (at 4%):

1 + v 1 1 p 62 + v 2 2 p 62 .

You should again get 2.83477.

Can you prove that this way of calculating ä x: 3 always works? That is, can
you prove that the following statement is true in general? (Hint: start with the
right-hand side of the equation below, and write ä 1 , ä 2 and ä 3 as (respectively)
1, 1 + v and 1 + v + v 2 ).
2
v r P(K x ≥ r ) = ä 1 P(K x =0) + ä 2 P(K x =1) + ä 3 P(K x ≥ 2).
X
r =0

We have already seen the intuitive reasoning for this result when we encountered
a similar formula for whole-life annuities. Again, we are simply saying that if the
person is alive at the start of the year (here we calculate the probability of being
alive), he will receive an annuity payment (so we multiply the probability by the
present value of one annuity payment).
228 Chapter 11 Life Contingencies

You will see in a later course that ä x: n and ä x can (for all x and n) be written
as follows:
n−1
X r n−1
X r
ä x: n = v P(K x ≥ r ) = v r px ; (11.3)
r =0 r =0
∞ ∞
v r P(K x ≥ r ) = v r r px .
X X
ä x =
r =0 r =0

Example 11.14
R500 is paid annually in advance to (31) while they are alive but for at most 4 years.
Assume that (31) is subject to the mortality of the ELT 12 (Males) and i =0.095. Find
the expected present value of this annuity.

Solution: We will use Equation (11.3) to find the expected present value of this
annuity.

E(500ä min(K 31 +1,4) ) = 500ä 31:4

3
v r r p 31
X
= 500 ×
r =0
= 500 × (0 p 31 + v p 31 + v 2 2 p 31 + v 3 3 p 31 )
l 31 l 32 l 33 l 34
µ ¶
= 500 × +v + v2 + v3
l l 31 l 31 l 31
µ 31 ¶
95 155 95 040 2 94 918 3 94789
= 500 × +v +v +v
95 155 95 155 95 155 95 155
95 155 + v95 040 + v 2 94 918 + v 3 94 789
= 500 ×
95 155
= 500 × 3.50280
= R1 751.40.

There is yet another way to calculate temporary annuities:

ä x:n = ä x −n p x v n ä x+n .

Use a cashflow diagram to show how this method works. You should also be
able to prove the above equation mathematically using ä x: n = n−1 r
P
P∞ r r =0 v r p x and
ä x = r =0 v r p x .

Exercises for 11.5.2 Temporary annuities

Ex.11.24 R100 is paid annually in advance to (57) while they are alive but for
at most 8 years. Assume that (57) is subject to the mortality of the
A1967–70 (ultimate) table and i =0.04.
1. Find the expected present value of this annuity using Equation
(11.3).
2. Check your answer using the tables in Appendix 11.9.
 11.6 Pricing 229

Ex.11.25 R100 is paid annually in advance to (65) while they are alive but for
at most 5 years. Assume that (65) is subject to the mortality of the
A1967–70 (ultimate) table and i =0.05. Find the expected present value
of this annuity.
Ex.11.26 R500 is paid annually in advance to (22) while they are alive but for
at most 4 years. Assume that (22) is subject to the mortality of ELT 12
(Males) and i =0.075. Find the expected present value of this annuity.
Ex.11.27 R100 is paid annually in advance to (32) while they are alive but for
at most 8 years. Assume that (32) is subject to the mortality of the
A1967–70 (ultimate) table and i =0.04. Find the expected present value
of this annuity.

11.6 Pricing
You are now ready to use life contingencies to calculate premiums for various
policies. Keep in mind that a regular premium payable until death or the end of
the policy term, whichever is sooner, is a life annuity, and pricing is as simple as
putting together an equation of value.
" A note on vocabulary: life
insurance uses the term
11.6.1 Net Premiums net premium instead of the
term risk premium we used
“Net premiums”, which we also call “risk premiums”, ignore expenses and profits
in general insurance. The
completely, and only allow for the cost of providing the benefit to the policyholder.
office premium is also called
Have another look at the cashflows of a regular premium whole-life assurance the gross premium.
policy sold to a life aged x exact:

−P −P −P −P −P ··· −P +S
-
time (yrs): 0 1 2 3 4 ··· Kx 6K x + 1
death

The premiums are payable from inception until the beginning of the period ? Why do you think the pol-
in which the policyholder dies. So how many annual premiums are paid? icyholder starts making
The answer is K x + 1, as the policyholder pays the first premium at inception premium payments at in-
ception and not at the end of
and then a premium at the start of each year that they survive.
every year?
For the whole-life assurance, now that we know how many premiums are paid
and when the sum assured is paid, we can write down expressions in terms of K x
for the present values of the premiums and sum assured. The present value of the
premiums is
P ä K x +1
and the present value of the sum assured is

Sv K x +1 .

It follows that the expected present value of the premiums is:

P ä x ,
230 Chapter 11 Life Contingencies

and the expected present value of the sum assured is

S Ax .

To calculate P , we equate the expected present value of the premiums with

the expected present value of the benefits. Considering the whole-life assurance
that pays the policyholder RS at the end of the year of death, this equation is:

P ä x = S A x ,

and rearranging we have

Ax
P =S .
ä x
In the case S = 1, the following is often tabulated

Ax
Px ≡ .
ä x

If the policy was a term assurance (sold to a life aged x exact) with a term of
n years, the premium payments will be limited to the length of the policy term,
and the net premiums for a term assurance policy are such that:
1
P ä x:n = S A x:n

and so
1
A x:n
P =S .
ä x:n

Exercises for 11.6.1 Net Premiums

Ex.11.28 Write down similar expressions for the net premiums of endowment
and pure endowment contracts, both of which have level annual pre-
miums paid in advance while the life is alive, but for at most the term
of the policy.

For an endowment assurance, when S = 1, the following is often tabulated for

you:
A x:n
P x:n ≡ .
ä x:n

Example 11.15
Suppose a life aged 43, who is subject to A1967–70 (ultimate) mortality, takes out
a term insurance policy with sum assured of R100 000 and term of 22 years. The
policyholder pays level annual premiums in advance while they are alive and for at
most 22 years. The interest rate used to price the policy is i = 0.04.
(a) What is the expected present value of the sum assured at inception?
11.6 Pricing 231

(b) What is the expected present value of the premiums? Use P to represent the
annual premium amount.
(c) Finally, what is the annual net premium?

Solution:
(a) From the A1967–70 (ultimate) table, we can find that A 43:22 = 0.44711, but we
need A 1 . We can use the relationship between a term assurance and an
43:22
endowment assurance:

A1 = A 43:22 −22 p 43 v 22
43:22
l 65 22
= 0.44711 − v
l 43
27 442.681 22
= 0.44711 − v
33 378.285
= 0.10019.

Therefore the expected present value of the benefit is

100 000 A 1 = 100 000 × 0.10019 = R10 019.02.

43:22

(b) The expected present value of the premiums is P ä 43:22 = 14.375P.

(c) Equating the expected present values of the benefits and the premiums, we
have
10 019.02 ? Find the premium for the
P= = R696.98 p.a.
14.375 corresponding endowment as-
surance. What do you notice?

Exercises for 11.6.1 Net Premiums

Ex.11.29 In Section 11.4 you completed a number of exercises in which you

calculated the expected present values of benefits from a range of
policies. Return to this section now, and calculate the net premium
payable for each policy using the same set of assumptions as you used
to calculate expected present value of the benefits. Assume that annual
premiums are payable in advance while the policyholder is alive, but
for at most the term of the policy, if the policy has a term.

11.6.2 Gross premiums

So far we have calculated net premiums. These are the premiums that do not allow
for expenses and profit. An insurer needs to be reimbursed for their expenses and
to earn profits on the business it sells. The premium that allows for expenses and
profits is called the gross premium or the office premium. In this section we are
concerned with calculating the gross premium.
Gross premiums are those which satisfy the equation

EPV of gross premiums = EPV of benefits + EPV of expenses + EPV of profits.

232 Chapter 11 Life Contingencies

As before, the abbreviation EPV stands for ‘expected present value’. Strictly speak-
ing, ‘expenses’ should read ‘estimated expenses’ or ‘assumed expenses’, as ex-
penses are not known in advance. Compare the above equation with

EPV of net premiums = EPV of benefits.

Which will be bigger, net premiums or gross premiums?

Example 11.16
Suppose that a life aged 35 and subject to A1967–70 (ultimate) mortality takes out a
whole-life policy with sum assured R100 000 payable at the end of the year of death.
Premiums are payable annually in advance (for life). The relevant interest rate is
4% p.a.
(a) What is the net premium for such a policy?
(b) Suppose that, in determining the gross premium for this policy, the life assurer
uses the following estimate of expenses:
• each time a premium is paid, R5 plus 2% of gross premium;
• additional initial expenses, R100 plus 70% of the gross premium; and
• claim expenses of 1% of the sum assured, incurred when the benefit is
paid.
The insurer requires that the profit earned on the policy equals 10% of the
expected present value of the gross premiums. Find the gross premium.

Solution:
(a) The net premium is

A 35 0.22810
P = R100 000 × = 100 000 × = R1 136.58.
ä 35 20.069

(b) Let G be the gross premium. By equating the expected present value of the pre-
miums with the expected present value of the benefits, profits, and expenses,
we have the following equation:

G ä 35 = R100 000 × 1.01 × A 35 + R100 + 0.7G + (R5 + 0.02G)ä 35 + 0.1G ä 35 .

Rearranging this, we have

R100 000 × 1.01 × A 35 + R100 + R5ä 35

G= .
0.88ä 35 − 0.7

Reading the relevant values off the life tables, we can find

R100 000 × 1.01 × 0.22810 + R100 + R5 × 20.069

G =
0.88 × 20.069 − 0.7
= R1 370.13.

We can see that the gross premium is significantly larger than the net premium.
11.6 Pricing 233

Example 11.17
Suppose a life aged exactly 66, subject to A1967–70 (ultimate) mortality, buys an
annuity, payable annually in arrear for life. The interest rate used is 4% p.a. and
expenses are as follows:
• each time an annuity payment is made, R100 plus 1.25% of the annuity actually
paid to (66);
• initial expenses, R500 plus 1.5% of the purchase sum.
Ignore profits in this case.
(a) Assume that the purchase sum is R2 million. How big is the annuity?
(b) Suppose the amount of the annuity is R150 000 (payable annually in arrear).
How big is the purchase sum?

Aside:

We have not dealt with life annuities paid in arrear. Suppose (x) has to make
payments annually in arrear while alive. The present value of these payments is
a K x and the expected present value is

E(a K x ) = a 1 P(K x =1) + a 2 P(K x =2) + a 3 P(K x =3) + . . . .

This would be tedious to calculate by hand, but fortunately there is a simple

relationship between a K x and ä K x +1 :
? Show that äK x +1 = 1 + aK x .
ä K x +1 = 1 + a K x ,

and so
E(ä K x +1 ) = 1 + E(a K x ). (11.4) U Recall that ä x = E(äK x +1 )

We denote E(a K x ) by a x , and from (11.4), we have

ä x = 1 + a x .

The tabulated ä x values can thus be used to find the a x values.

Example 11.17 continued.

Solution:
(a) Let the annuity payment be X . Then:

2 × 106 = X (ä 66 − 1) + 500 + 0.015(2 × 106 ) + (100 + 0.0125X )(ä 66 − 1)

= 1.0125X × 9.376 + 500 + 30 000 + 937.60.

Hence X = R207 365.52, i.e. approximately R207 366. Or even just ap-
proximately R207 400 (since the tables give you four significant figures
only for ä 66 ).
234 Chapter 11 Life Contingencies

(b) Purchase sum is G, where:

G = 150 000(ä 66 − 1) + 500 + 0.015G + (100 + 1875)(ä 66 − 1)

= 1 425 417.6 + 0.015G.

Hence G = R1 447 124.47, i.e. approximately R1 447 124.

Exercises for 11.6.2 Gross premiums

Ex.11.30
T Initial expenses: these ex-
penses are incurred by the in- Suppose a life aged exactly 32, subject to A1967–70 (ultimate) mortality,
surer at the outset of the pol- takes out an endowment assurance policy. The term on the endowment
icy. They are usually treated assurance is 33 years and the sum assured is R150 000. The sum assured
as happening at the same will be paid at the end of year of death or after 33 years, whichever is
time that the first premium is sooner. Premiums are payable annually in advance while the life is
paid.
alive, but for at most 33 years. The expenses on the policy are:
Renewal expenses: these
expenses are incurred by an • Initial expenses: R2 500 plus 10% of the gross premium,
insurer on a regular basis. • Renewal expenses: R100 plus 1.5% of the gross premium, incurred
They are usually assumed when each premium is paid, but from the start of the second year.
to happen when a regular
premium is due or when an
• Termination expenses: R500 incurred when the benefit is paid.
annuity payment occurs. The insurer wants to earn a profit equal to 12% of the expected present
Termination expenses: these value of the gross premiums. The interest rate is 4% p.a.
expenses are incurred when
a policy comes to an end, 1. Find the annual net premium.
e.g. on the death of a policy- 2. Find the annual gross premium.
holder or annuitant, or on the
Ex.11.31 Suppose a life aged exactly 41, subject to A1967–70 (ultimate) mor-
survival of the policyholder
tality, takes out a term assurance policy. The term is 29 years and the
to the end of the term of a
term insurance policy or an sum assured is R250 000. The sum assured will be paid at the end of
endowment assurance policy. year of death provided that this happens within 29 years. Premiums
are payable annually in advance while the life is alive but for at most
29 years. The expenses on the policy are
• Initial expenses: R750 plus 5% of the gross premium,
• Renewal expenses: R250 plus 1.5% of the gross premium, incurred
when each premium is paid, but from the start of the second year.
• Termination expenses: R250 incurred only if and when the benefit
is paid.
The insurer aims to earn a profit equal to 15% of the expected present
value of the gross premiums. The interest rate is 4% p.a.
1. Find the annual net premium.
2. Find the annual gross premium.
11.7 Reserving: Gross premium policy values 235

Ex.11.32 Consider a life aged (55) exact who buys a life annuity with a cost
of R2 000 000 (this is the gross premium). The annuity pays annual
amounts in arrear and the following expenses are allowed for in the
calculation of the gross premium:
• Initial expenses: R5 000 plus 0.5% of the gross premium,
• Renewal expenses: R250 plus 1.5% of the annuity amount.
Ignore profits.
Using the A1967–70 (ultimate) mortality table and an interest rate of
4% p.a., find the annual annuity amount.

11.7 Reserving: Gross premium policy values

The concept of a policy value is important as it is used to determine the reserves
a life insurer must hold. They are also used to determine the profit or loss of a life
insurance company. We will only consider gross premium policy values here. For
our purposes the gross premium policy value is defined as U “Policy value” is another
term for reserve in life insur-
ance.
Gross premium policy value at time t = EPV at time t of future benefits+
EPV at time t of future expenses − EPV at time t of future premiums,

where the premiums are the gross premiums.

The easiest way to see how to calculate the gross policy value is to do some
examples.

Example 11.18
Consider a 10-year endowment assurance policy purchased by a life aged 50 exactly.
Level annual premiums are payable throughout the term of the policy, and the sum
insured, R150 000, is payable at the end of the year of death or at the end of the
term, whichever is sooner.
The basis used for finding the gross premium and the policy value is:
• Mortality: A1967–70 (ultimate),
• Interest: 4% p.a.,
• Initial expenses: R250 plus 1.5% of the gross premium,
• Renewal expenses: R150 plus 0.5% of the gross premium, incurred when each
premium is paid, but from the start of the second year.
• Termination expenses: R250 incurred when the benefit is paid.
(a) Find the gross premium p.a.
(b) Find the gross premium policy value at times 5 and 7 (in both cases just before
the premium due at that time is paid).

Solution:
236 Chapter 11 Life Contingencies

(a) Let G be the gross premium. Equating the expected present value of the
premiums with the expected present value of the benefits and the expenses,
we have:

G ä 50:10 = (150 000 + 250)A 50:10 + 250 + 0.015G + (150 + 0.005G)(ä 50:10 − 1).

Hence
(150 000 + 250)A 50:10 + 250 + 150(ä 50:10 − 1)
G= .
ä 50:10 − 0.015 − 0.005(ä 50:10 − 1)
From the tables we have

A 50:10 = 0.68436 ä 50:10 = 8.207,

and so
(150 000 + 250) × 0.68436 + 250 + 150 × (8.207 − 1)
G= = R12 770.55.
8.207 − 0.015 − 0.005(8.207 − 1)

(b) We will use the notation PVt to denote the gross policy value at time t . We
need to find PV5 and PV7 . Remember

PVt = EPV at time t of future benefits + EPV at time t of future expenses

− EPV at time t of future premiums,

and so in this case

PV5 = 150 000A 55:5 + 250A 55:5 + (150 + 0.005 × 12 770.55)ä 55:5 −
| {z } | {z }
EPV at time 5 of future benefits EPV at time 5 of future expenses

12 770.55ä 55:5 .
| {z }
EPV at time 5 of future premiums

From the tables we have

A 55:5 = 0.82513 ä 55:5 = 4.547,

and so

PV5 = 150 000 × 0.82513 + 250 × 0.82513 + (150 + 0.005 × 12 770.55) × 4.547
− 12 770.55 × 4.547
= R66 880.48

Similarly,
? Calculate PV8 and PV9 for this
policy. What do you notice? PV7 = (150 000 + 250)A 57:3 + (150 + 0.005 × 12 770.55)ä 57:3 − 12 770.55ä 57:3
= (150 000 + 250) × 0.89017 + (150 + 0.005 × 12 770.55) × 2.856
−12 770.55 × 2.856
= R97 886.12.
11.7 Reserving: Gross premium policy values 237

Example 11.19
Consider a life aged 55 exact who purchases an annuity with a single premium.
The annuity pays level annual payments in arrear of R150 000 for as long as the
annuitant is alive. The basis used for finding the gross premium and the policy
value is:
• Mortality: A1967–70 (ultimate),
• Interest: 4% p.a.,
• Initial expenses: R1 500 plus 1.5% of the gross premium,
• Renewal expenses: R100 plus 0.5% of the annuity amount.
Ignore profit.
(a) Find the single gross premium.
(b) Find the gross premium policy value at times 10 and 15, in both cases just after
the annuity payment due at that time is paid.

Solution:
(a) Let G be the gross premium. Equating the expected present value of the
premium with the expected present value of the benefits and expenses, we
have:

G = 150 000(ä 55 − 1) + 1 500 + 0.015G + (100 + 0.005 × 150 000)(ä 55 − 1)

150 000(14.327 − 1) + 1 500 + (100 + 0.005 × 150 000)(14.327 − 1)
⇒G =
0.985
= R2 042 515.69.

(b)

PV10 = 150 000(ä 65 − 1) + (100 + 0.005 × 150 000)(ä 65 − 1)

= 150 000(10.737 − 1) + (100 + 0.005 × 150 000)(10.737 − 1)
= R1 468 826.45.
? Calculate PV20 and PV30 for
this policy. What do you no-
PV15 = 150 000(ä 70 − 1) + (100 + 0.005 × 150 000)(ä 70 − 1) tice?
= 150 000(8.957 − 1) + (100 + 0.005 × 150 000)(8.957 − 1)
= R1 200 313.45.

Exercises for 11.7 Reserving: Gross premium policy values

Ex.11.33 Suppose a life aged exactly 35, subject to A1967–70 (ultimate) mortal-
ity, takes out a whole-life policy. The sum assured is R150 000 and will
be paid at the end of year of death. Premiums are payable annually in
advance while the life is alive. The expenses on the policy are:
• Initial expenses: R2 500 plus 10% of the gross premium,
• Renewal expenses: R100 plus 1.5% of the gross premium, incurred
when each premium is paid, but from the start of the second year.
238 Chapter 11 Life Contingencies

• Termination expenses: R500 incurred when the benefit is paid.

The interest rate is 4% p.a. Ignore profits.
1. Find the annual net premium.
2. Find the annual gross premium.
3. Find the gross premium policy value at times 10 and 20.
Ex.11.34 Consider a 20-year term insurance policy sold to a life now aged 45.
The sum assured is R100 000. The policy has annual premiums payable
in advance while the life is alive, but for at most 20 years. The basis to
be used for the premiums and the policy values is:
• Mortality: A1967–70 (ultimate),
• Interest: 4% p.a.,
• Initial expenses: R500 plus 10% of the gross premium,
• Renewal expenses: R100 plus 1.5% of the gross premium, incurred
when each premium is paid.
• Termination expenses: R300 incurred if and when the benefit is
paid.
Ignore profits.
1. Find the annual net premium.
2. Find the annual gross premium.
3. Find the gross premium policy value at times 5, 10 and 15.
4. Comment on your answer to part 3 of this question.

11.8 Some more examples

Example 11.20 Increasing payments

Consider a life aged 45 exact who takes out a special term assurance with a term of
5 years and level annual premiums. The premiums are paid annually in advance
while the life is alive, but for at most 5 years. The sum assured is R100 000 in the
first year and increases by R50 000 each year after that, and is payable at the end
of year of death within the 5 year term. The life is subject to A1967–70 (ultimate)
mortality and the interest rate is 5% p.a.
Find the annual net premium for this special term assurance.

Solution:
We will find the expected present value of the benefits and premiums separately.
EPV(benefits) = 100 000(0 p 45 q 45 v 1 + 1.5 ×1 p 45 q 46 v 2 + 2 ×2 p 45 q 47 v 3
+2.5 ×3 p 45 q 48 v 4 + 3 ×4 p 45 q 49 v 5 )
d 45 v 1 + 1.5 × d 46 v 2 + 2 × d 47 v 3 + 2.5 × d 48 v 4 + 3 × d 49 v 5
= 105 ×
l 45
= 105 × (87.639072v 1 + 1.5 × 98.666249v 2 + 2 × 110.960099v 3 +
2.5 × 124.619139v 4 + 3 × 139.746592v 5 )/33231.486
= R2 991.76.
11.8 Some more examples 239

Let P be the annual net premium.

U The mortality function
EPV(premiums) = P ä 45:5 5% ä 45:5 5% is not tabulated in
4 Appendix 11.9.
r
X
= P r p 45 v
r =0
l 45 + l 46 v + l 47 v 2 + l 48 v 3 + l 49 v 4
= P
l 45
33231.486 + 33143.847v + 33045.181v 2 + 32934.221v 3 + 32809.601v 4
= P
33231.486
= 4.5202P.

By equating the expected present value of the benefits and the premiums, we have

R2 991.76
P= = R661.87.
4.5202

Example 11.21 Deferred payments

Consider a life aged 27 exact who purchases a retirement annuity. The retirement
annuity will pay annual amounts of R150 000 from age 65 provided the life is alive
at that time. The payments will continue for as long as the life is alive. Annual
premiums are paid in advance while the life is alive and before they reach age 65.
The life is subject to A1967–70 (ultimate) mortality and the interest rate is 4% p.a.
T Recall that a retirement an-
Find the annual net premium for retirement annuity.
nuity is an annuity that starts
paying out when a person
Solution: We will find the expected present value of the benefits and premiums
retires, and that is purchased
separately.
with regular premium pay-
EPV(benefits) = 150 00038 p 27 v 38 ä 65 Why do we multiply by 38 p 27 v 38 ? ments paid until the retire-
27442.681 38 ment date.
= 150 000 v × 10.737
33905.397
= R293 673.68.

Let P be the annual net premium.

EPV(premiums) = P ä 27:38 4%
= 19.613P.

By equating the expected present value of the benefits and the premiums, we have

R293 673.68.
P= = R14 973.42.
19.613

Example 11.22 Limited-premium whole-life assurance

Suppose a life aged 31 exact takes out a limited-premium whole-life policy which
pays R1 000 000 at the end of year of death. The premiums are payable annually in
advance while the life is alive but for at most 34 years.
The basis used to find the gross premium and the gross premium policy values is:
240 Chapter 11 Life Contingencies

• Mortality: A1967–70 (ultimate),

• Interest: 4% p.a.,
• Initial expenses: R500 plus 0.5% of the sum assured,
• Renewal expenses: R150 plus 0.5% of the gross premium, incurred when each
premium is paid, but only from the second year.
• Termination expenses: R150 incurred when the benefit is paid.
Ignore profits.
(a) Find the gross annual premium.
(b) Find the gross premium policy values at times 20 and 40.

Solution:
(a) We will find the expected present value of the benefits, expenses, and premi-
ums separately.

EPV(benefits) = 1 000 000A 31

= 1 000 000 × 0.19704
= R197 040.

Let G be the annual gross premium.

6
EPV(expenses) = 500
| + 0.005
{z × 10} + |(0.005G + 150)( ä − 1) + 150A 31
{z 31:34 } | {z }
Initial expenses Renewal expenses Termination expenses

= 5 500 + (0.005G + 150) × (18.581 − 1) + 150 × 0.19704

= R8 166.71 + 0.087905G.

EPV(premiums) = G ä 31:34
= 18.581G.

Equating the sum of expected present value of benefits and expenses with the
expected present value of premiums, we have

18.581G = R197 040 + R8 166.71 + 0.087905G.

Hence,
R197 040 + R8 166.71
G= = R11 096.40.
18.581 − 0.087905
(b)

PV20 = 1 000 000A 51 + 150A 51 + (0.005 × 11 096.40 + 150)ä 51:14 − 11 096.40ä 51:14
= 1 000 000 × 0.39699 + 150 × 0.39699 + (0.005 × 11 096.40 + 150) × 10.445
−11 096.40 × 10.445
= R283 293.91.

PV40 = 1 000 000A 71 + 150A 71

= 1 000 000 × 0.66876 + 150 × 0.66876
= R668 860.31.
11.8 Some more examples 241

Exercises for 11 Life Contingencies

Ex.11.35 Suppose that A x = 0.2, A x+20 = 0.35, and A x:20 = 0.55.

Find v 20 20 p x .
Ex.11.36 An insurance company issues a special term assurance policy to a life
aged exactly 60. The sums assured in respect of the first, second and
third years of the policy are R200 000, R150 000 and R50 000 respectively.
Death benefits are payable at the end of the year of death. In addition, a
claim expense is incurred at the time a benefit is paid. The initial level,
at the outset of the policy, is R200; however, this expense increases in
line with inflation.
1. Calculate the expected present value of the outgo for this policy.
2. Premiums are payable annually in advance while the life is alive,
but for at most 3 years. Find the gross premium for this policy.
Basis:
Mortality: A1967–70 (ultimate),
Interest: 8% p.a.,
Expense inflation: 5% p.a.
Life Tables

11.9 A1967–70 (ultimate)

A1967–70 (ult.): l x , d x , q x , e x and ä x , A x , P x for i = 4%

x lx dx qx ex ä x Ax Px
0 34489.000 25.176970 0.00073000 73.305 24.235 0.06788 0.00280
1 34463.823 23.435400 0.00068000 72.358 24.182 0.06992 0.00289
2 34440.388 21.697444 0.00063000 71.407 24.126 0.07208 0.00299
3 34418.690 19.962840 0.00058000 70.452 24.066 0.07438 0.00309
4 34398.727 18.231325 0.00053000 69.493 24.003 0.07682 0.00320
5 34380.496 16.846443 0.00049000 68.530 23.935 0.07941 0.00332
6 34363.650 15.463642 0.00045000 67.564 23.864 0.08214 0.00344
7 34348.186 14.426238 0.00042000 66.594 23.790 0.08501 0.00357
8 34333.760 13.733504 0.00040000 65.622 23.711 0.08803 0.00371
9 34320.026 13.041610 0.00038000 64.648 23.629 0.09118 0.00386
10 34306.985 12.693584 0.00037000 63.673 23.543 0.09449 0.00401
11 34294.291 12.688888 0.00037000 62.696 23.454 0.09793 0.00418
12 34281.602 12.684193 0.00037000 61.720 23.361 0.10152 0.00435
13 34268.918 13.707567 0.00040000 60.743 23.264 0.10525 0.00452
14 34255.210 16.099949 0.00047000 59.767 23.163 0.10910 0.00471
15 34239.110 20.885857 0.00061000 58.795 23.061 0.11305 0.00490
16 34218.225 27.716762 0.00081000 57.831 22.957 0.11703 0.00510
17 34190.508 36.083978 0.00105538 56.878 22.854 0.12100 0.00529
18 34154.424 34.045471 0.00099681 55.938 22.752 0.12492 0.00549
19 34120.378 32.120924 0.00094140 54.994 22.645 0.12905 0.00570
20 34088.257 30.320482 0.00088947 54.045 22.532 0.13339 0.00592
21 34057.937 28.654305 0.00084134 53.094 22.413 0.13796 0.00616
22 34029.283 27.134610 0.00079739 52.138 22.288 0.14276 0.00641
23 34002.148 25.773968 0.00075801 51.180 22.157 0.14779 0.00667
24 33976.374 24.587343 0.00072366 50.219 22.020 0.15306 0.00695
25 33951.787 23.590041 0.00069481 49.255 21.877 0.15857 0.00725
26 33928.197 22.800087 0.00067201 48.289 21.727 0.16433 0.00756
27 33905.397 22.235837 0.00065582 47.322 21.571 0.17035 0.00790
28 33883.161 21.918678 0.00064689 46.353 21.408 0.17662 0.00825
Continued on Next Page. . .
243
244 Life tables

x lx dx qx ex ä x Ax Px
29 33861.242 21.871653 0.00064592 45.383 21.238 0.18316 0.00862
30 33839.370 22.120120 0.00065368 44.412 21.061 0.18996 0.00902
31 33817.250 22.691713 0.00067101 43.441 20.877 0.19704 0.00944
32 33794.559 23.616313 0.00069882 42.470 20.686 0.20439 0.00988
33 33770.942 24.927346 0.00073813 41.500 20.488 0.21201 0.01035
34 33746.015 26.660702 0.00079004 40.531 20.282 0.21991 0.01084
35 33719.354 28.856012 0.00085577 39.563 20.069 0.22810 0.01137
36 33690.498 31.555531 0.00093663 38.597 19.849 0.23657 0.01192
37 33658.943 34.806376 0.00103409 37.633 19.621 0.24533 0.01250
38 33624.136 38.658678 0.00114973 36.672 19.386 0.25437 0.01312
39 33585.478 43.166743 0.00128528 35.714 19.144 0.26370 0.01377
40 33542.311 48.390486 0.00144267 34.760 18.894 0.27331 0.01447
41 33493.920 54.392787 0.00162396 33.810 18.637 0.28321 0.01520
42 33439.528 61.242823 0.00183145 32.865 18.372 0.29339 0.01597
43 33378.285 69.013943 0.00206763 31.925 18.100 0.30385 0.01679
44 33309.271 77.784809 0.00233523 30.992 17.821 0.31459 0.01765
45 33231.486 87.639072 0.00263723 30.064 17.534 0.32560 0.01857
46 33143.847 98.666249 0.00297691 29.144 17.241 0.33687 0.01954
47 33045.181 110.960099 0.00335783 28.231 16.941 0.34841 0.02057
48 32934.221 124.619139 0.00378388 27.326 16.635 0.36020 0.02165
49 32809.601 139.746592 0.00425932 26.430 16.322 0.37223 0.02281
50 32669.855 156.449401 0.00478880 25.543 16.003 0.38450 0.02403
51 32513.405 174.837587 0.00537740 24.666 15.678 0.39699 0.02532
52 32338.568 195.022261 0.00603064 23.799 15.348 0.40969 0.02669
53 32143.546 217.115508 0.00675456 22.943 15.012 0.42260 0.02815
54 31926.430 241.227167 0.00755572 22.099 14.672 0.43569 0.02970
55 31685.203 267.463670 0.00844128 21.268 14.327 0.44896 0.03134
56 31417.739 295.924315 0.00941902 20.449 13.978 0.46238 0.03308
57 31121.815 326.698763 0.01049742 19.643 13.626 0.47594 0.03493
58 30795.116 359.861258 0.01168566 18.851 13.270 0.48962 0.03690
59 30435.255 395.467485 0.01299373 18.074 12.912 0.50340 0.03899
60 30039.787 433.548031 0.01443246 17.312 12.551 0.51726 0.04121
61 29606.239 474.101291 0.01601356 16.566 12.189 0.53118 0.04358
62 29132.138 517.087295 0.01774972 15.835 11.826 0.54515 0.04610
63 28615.051 562.418523 0.01965464 15.122 11.463 0.55913 0.04878
64 28052.632 609.951190 0.02174310 14.425 11.099 0.57310 0.05163
65 27442.681 659.475345 0.02403101 13.745 10.737 0.58705 0.05468
66 26783.206 710.705757 0.02653550 13.084 10.376 0.60094 0.05792
67 26072.500 763.270092 0.02927491 12.440 10.016 0.61476 0.06138
68 25309.230 816.701010 0.03226890 11.816 9.660 0.62847 0.06506
69 24492.529 870.42676 0.03553846 11.210 9.307 0.64206 0.06899
70 23622.102 923.764510 0.03910594 10.623 8.957 0.65550 0.07318
71 22698.338 975.916616 0.04299507 10.055 8.612 0.66876 0.07765
Continued on Next Page. . .
A1967–70 (ult.) 245

x lx dx qx ex ä x Ax Px
72 21722.421 1025.970799 0.04723096 9.507 8.272 0.68183 0.08242
73 20696.450 1072.905636 0.05184008 8.978 7.938 0.69469 0.08751
74 19623.545 1115.602828 0.05685022 8.469 7.610 0.70730 0.09294
75 18507.942 1152.867282 0.06229041 7.979 7.289 0.71966 0.09873
76 17355.074 1183.456935 0.06819083 7.509 6.975 0.73173 0.10491
77 16171.618 1206.121769 0.07458263 7.059 6.669 0.74351 0.11149
78 14965.496 1219.654833 0.08149779 6.628 6.371 0.75498 0.11851
79 13745.841 1222.951387 0.08896883 6.216 6.081 0.76611 0.12598
80 12522.890 1215.077817 0.09702855 5.823 5.800 0.77691 0.13394
81 11307.812 1195.345162 0.10570968 5.449 5.529 0.78735 0.14241
82 10112.467 1163.382955 0.11504443 5.093 5.267 0.79743 0.15141
83 8949.084 1119.208465 0.12506402 4.755 5.014 0.80714 0.16097
84 7829.875 1063.282955 0.13579821 4.434 4.772 0.81648 0.17111
85 6766.592 996.546350 0.14727447 4.131 4.539 0.82543 0.18186
86 5770.046 920.423984 0.15951762 3.845 4.316 0.83400 0.19324
87 4849.622 836.796582 0.17254883 3.574 4.103 0.84219 0.20525
88 4012.825 747.930364 0.18638499 3.320 3.900 0.84999 0.21793
89 3264.895 656.367491 0.20103786 3.080 3.707 0.85741 0.23128
90 2608.527 564.781016 0.21651335 2.855 3.524 0.86446 0.24531
91 2043.746 475.805955 0.23281064 2.644 3.350 0.87114 0.26002
92 1567.940 391.862192 0.24992160 2.447 3.186 0.87746 0.27541
93 1176.078 314.988929 0.26782987 2.262 3.031 0.88342 0.29146
94 861.089 246.711348 0.28651079 2.090 2.885 0.88904 0.30817
95 614.378 187.956835 0.30593023 1.929 2.748 0.89433 0.32550
96 426.421 139.032704 0.32604547 1.779 2.619 0.89929 0.34343
97 287.388 99.667530 0.34680441 1.640 2.498 0.90394 0.36193
98 187.721 69.108564 0.36814494 1.510 2.384 0.90829 0.38093
99 118.612 46.258687 0.38999885 1.390 2.279 0.91236 0.40040
246 Life tables

A1967–70 (ult.): ä x:n , A x:n and P x:n for x + n = 55 and i = 4%

x ä x:n A x:n P x:n n

54 1.000 0.96154 0.96154 1
53 1.955 0.92481 0.47304 2
52 2.869 0.88967 0.31015 3
51 3.743 0.85602 0.22868 4
50 4.582 0.82376 0.17978 5
49 5.387 0.79280 0.14717 6
48 6.160 0.76306 0.12387 7
47 6.904 0.73448 0.10639 8
46 7.618 0.70699 0.09280 9
45 8.306 0.68054 0.08193 10
44 8.968 0.65508 0.07305 11
43 9.605 0.63057 0.06565 12
42 10.219 0.60697 0.05940 13
41 10.810 0.58424 0.05405 14
40 11.379 0.56235 0.04942 15
39 11.927 0.54126 0.04538 16
38 12.455 0.52095 0.04183 17
37 12.964 0.50139 0.03868 18
36 13.454 0.48255 0.03587 19
35 13.925 0.46442 0.03335 20
34 14.379 0.44696 0.03108 21
33 14.816 0.43017 0.02903 22
32 15.236 0.41400 0.02717 23
31 15.640 0.39846 0.02548 24
30 16.029 0.38351 0.02393 25
29 16.402 0.36914 0.02251 26
28 16.761 0.35534 0.02120 27
27 17.106 0.34208 0.02000 28
26 17.437 0.32935 0.01889 29
25 17.755 0.31713 0.01786 30
24 18.059 0.30541 0.01691 31
23 18.352 0.29416 0.01603 32
22 18.632 0.28339 0.01521 33
21 18.900 0.27307 0.01445 34
20 19.157 0.26319 0.01374 35
19 19.403 0.25374 0.01308 36
18 19.638 0.24469 0.01246 37
17 19.863 0.23605 0.01188 38
16 20.083 0.22756 0.01133 39
15 20.299 0.21926 0.01080 40
A1967–70 (ult.) 247

A1967–70 (ult.): ä x:n , A x:n and P x:n for x + n = 60 and i = 4%

x ä x:n A x:n P x:n n

59 1.000 0.96154 0.96154 1
58 1.950 0.92499 0.47428 2
57 2.856 0.89017 0.31173 3
56 3.720 0.85693 0.23036 4
55 4.547 0.82513 0.18148 5
54 5.339 0.79466 0.14885 6
53 6.099 0.76543 0.12551 7
52 6.829 0.73735 0.10798 8
51 7.531 0.71035 0.09433 9
50 8.207 0.68436 0.08339 10
49 8.857 0.65934 0.07444 11
48 9.484 0.63522 0.06697 12
47 10.089 0.61196 0.06066 13
46 10.672 0.58954 0.05524 14
45 11.235 0.56790 0.05055 15
44 11.777 0.54703 0.04645 16
43 12.301 0.52689 0.04283 17
42 12.806 0.50746 0.03963 18
41 13.294 0.48871 0.03676 19
40 13.764 0.47062 0.03419 20
39 14.217 0.45318 0.03187 21
38 14.655 0.43635 0.02978 22
37 15.077 0.42013 0.02787 23
36 15.483 0.40449 0.02612 24
35 15.875 0.38943 0.02453 25
34 16.252 0.37491 0.02307 26
33 16.616 0.36094 0.02172 27
32 16.965 0.34748 0.02048 28
31 17.302 0.33454 0.01934 29
30 17.626 0.32209 0.01827 30
29 17.937 0.31012 0.01729 31
28 18.236 0.29862 0.01638 32
27 18.523 0.28758 0.01553 33
26 18.799 0.27698 0.01473 34
25 19.063 0.26681 0.01400 35
24 19.316 0.25706 0.01331 36
23 19.559 0.24771 0.01266 37
22 19.792 0.23876 0.01206 38
Continued on Next Page. . .
248 Life tables

x ä x:n A x:n P x:n n

21 20.015 0.23020 0.01150 39
20 20.228 0.22200 0.01097 40
19 20.432 0.21417 0.01048 41
18 20.626 0.20668 0.01002 42
17 20.812 0.19954 0.00959 43
A1967–70 (ult.) 249

A1967–70 (ult.): ä x:n , A x:n and P x:n for x + n = 65 and i = 4%

x ä x:n A x:n P x:n n

64 1.000 0.96154 0.96154 1
63 1.943 0.92528 0.47630 2
62 2.835 0.89097 0.31430 3
61 3.682 0.85838 0.23312 4
60 4.489 0.82733 0.18429 5
59 5.261 0.79767 0.15163 6
58 5.999 0.76926 0.12823 7
57 6.708 0.74200 0.11062 8
56 7.389 0.71580 0.09687 9
55 8.045 0.69058 0.08584 10
54 8.677 0.66627 0.07678 11
53 9.287 0.64281 0.06922 12
52 9.876 0.62016 0.06279 13
51 10.445 0.59827 0.05728 14
50 10.995 0.57711 0.05249 15
49 11.527 0.55664 0.04829 16
48 12.042 0.53685 0.04458 17
47 12.540 0.51769 0.04128 18
46 13.022 0.49916 0.03833 19
45 13.488 0.48123 0.03568 20
44 13.939 0.46389 0.03328 21
43 14.375 0.44711 0.03110 22
42 14.797 0.43089 0.02912 23
41 15.205 0.41521 0.02731 24
40 15.599 0.40005 0.02565 25
39 15.980 0.38540 0.02412 26
38 16.347 0.37126 0.02271 27
37 16.702 0.35761 0.02141 28
36 17.045 0.34443 0.02021 29
35 17.375 0.33172 0.01909 30
34 17.694 0.31947 0.01806 31
33 18.001 0.30767 0.01709 32
32 18.296 0.29630 0.01619 33
31 18.581 0.28536 0.01536 34
30 18.854 0.27483 0.01458 35
29 19.118 0.26471 0.01385 36
28 19.370 0.25499 0.01316 37
27 19.613 0.24565 0.01252 38
26 19.846 0.23669 0.01193 39
25 20.070 0.22810 0.01137 40
24 20.284 0.21986 0.01084 41
Continued on Next Page. . .
250 Life tables

x ä x:n A x:n P x:n n

23 20.489 0.21197 0.01035 42
22 20.685 0.20442 0.00988 43
21 20.873 0.19720 0.00945 44
20 21.052 0.19031 0.00904 45
19 21.223 0.18372 0.00866 46
18 21.387 0.17744 0.00830 47
17 21.542 0.17145 0.00796 48
16 21.697 0.16550 0.00763 49
15 21.850 0.15962 0.00731 50
A1967–70 (ult.) 251

A1967–70 (ult.): ä x:n , A x:n and P x:n for x + n = 70 and i = 4%

x ä x:n A x:n P x:n n

69 1.000 0.96154 0.96154 1
68 1.931 0.92575 0.47954 2
67 2.802 0.89223 0.31844 3
66 3.623 0.86067 0.23758 4
65 4.400 0.83078 0.18883 5
64 5.138 0.80237 0.15615 6
63 5.844 0.77524 0.13266 7
62 6.519 0.74926 0.11493 8
61 7.168 0.72430 0.10105 9
60 7.793 0.70027 0.08986 10
59 8.396 0.67708 0.08065 11
58 8.979 0.65467 0.07291 12
57 9.543 0.63298 0.06633 13
56 10.089 0.61196 0.06065 14
55 10.619 0.59157 0.05571 15
54 11.134 0.57178 0.05136 16
53 11.633 0.55257 0.04750 17
52 12.118 0.53391 0.04406 18
51 12.589 0.51579 0.04097 19
50 13.047 0.49818 0.03818 20
49 13.492 0.48108 0.03566 21
48 13.924 0.46446 0.03336 22
47 14.344 0.44833 0.03126 23
46 14.751 0.43266 0.02933 24
45 15.146 0.41746 0.02756 25
44 15.530 0.40271 0.02593 26
43 15.901 0.38841 0.02443 27
42 16.262 0.37455 0.02303 28
41 16.611 0.36112 0.02174 29
40 16.949 0.34812 0.02054 30
39 17.276 0.33553 0.01942 31
38 17.593 0.32336 0.01838 32
37 17.898 0.31160 0.01741 33
36 18.194 0.30023 0.01650 34
35 18.479 0.28926 0.01565 35
34 18.754 0.27868 0.01486 36
33 19.020 0.26847 0.01412 37
32 19.275 0.25864 0.01342 38
31 19.522 0.24917 0.01276 39
30 19.759 0.24005 0.01215 40
29 19.986 0.23129 0.01157 41
Continued on Next Page. . .
252 Life tables

x ä x:n A x:n P x:n n

28 20.205 0.22288 0.01103 42
27 20.415 0.21479 0.01052 43
26 20.617 0.20704 0.01004 44
25 20.810 0.19961 0.00959 45
24 20.995 0.19249 0.00917 46
23 21.173 0.18567 0.00877 47
22 21.342 0.17915 0.00839 48
21 21.504 0.17293 0.00804 49
20 21.658 0.16698 0.00771 50
19 21.806 0.16132 0.00740 51
18 21.946 0.15592 0.00710 52
17 22.080 0.15078 0.00683 53
16 22.213 0.14564 0.00656 54
15 22.346 0.14054 0.00629 55
11.10 English Life Table No. 12 (Males) 253

11.10 English Life Table No. 12 (Males)

x lx dx px qx
0 100000 2449 0.97551 0.02449
1 97551 153 0.99843 0.00157
2 97398 96 0.99901 0.00099
3 97302 67 0.99931 0.00069
4 97235 60 0.99938 0.00062
5 97175 55 0.99943 0.00057
6 97120 51 0.99947 0.00053
7 97069 47 0.99952 0.00048
8 97022 43 0.99956 0.00044
9 96979 40 0.99959 0.00041
10 96939 38 0.99961 0.00039
11 96901 37 0.99962 0.00038
12 96864 37 0.99962 0.00038
13 96827 40 0.99959 0.00041
14 96787 45 0.99954 0.00046
15 96742 57 0.99941 0.00059
16 96685 75 0.99922 0.00078
17 96610 96 0.99901 0.00099
18 96514 108 0.99888 0.00112
19 96406 113 0.99883 0.00117
20 96293 115 0.99881 0.00119
21 96178 113 0.99883 0.00117
22 96065 110 0.99885 0.00115
23 95955 104 0.99892 0.00108
24 95851 98 0.99898 0.00102
25 95753 95 0.99901 0.00099
26 95658 94 0.99902 0.00098
27 95564 96 0.99900 0.00100
28 95468 99 0.99896 0.00104
29 95369 104 0.99891 0.00109
30 95265 110 0.99885 0.00115
31 95155 115 0.99879 0.00121
32 95040 122 0.99872 0.00128
33 94918 129 0.99864 0.00136
34 94789 137 0.99855 0.00145
35 94652 147 0.99845 0.00155
36 94505 158 0.99833 0.00167
37 94347 171 0.99819 0.00181
38 94176 185 0.99804 0.00196
39 93991 201 0.99786 0.00214
Continued on Next Page. . .
254 Life tables

x lx dx px qx
40 93790 220 0.99765 0.00235
41 93570 242 0.99741 0.00259
42 93328 268 0.99713 0.00287
43 93060 297 0.99681 0.00319
44 92763 330 0.99644 0.00356
45 92433 369 0.99601 0.00399
46 92064 412 0.99552 0.00448
47 91652 463 0.99495 0.00505
48 91189 520 0.99430 0.00570
49 90669 584 0.99356 0.00644
50 90085 656 0.99272 0.00728
51 89429 736 0.99177 0.00823
52 88693 825 0.99070 0.00930
53 87868 923 0.98950 0.01050
54 86945 1029 0.98816 0.01184
55 85916 1144 0.98668 0.01332
56 84772 1265 0.98508 0.01492
57 83507 1393 0.98332 0.01668
58 82114 1526 0.98142 0.01858
59 80588 1664 0.97935 0.02065
60 78924 1805 0.97713 0.02287
61 77119 1947 0.97475 0.02525
62 75172 2088 0.97222 0.02778
63 73084 2228 0.96951 0.03049
64 70856 2366 0.96661 0.03339
65 68490 2499 0.96351 0.03649
66 65991 2625 0.96022 0.03978
67 63366 2745 0.95668 0.04332
68 60621 2856 0.95289 0.04711
69 57765 2959 0.94878 0.05122
70 54806 3051 0.94433 0.05567
71 51755 3130 0.93952 0.06048
72 48625 3195 0.93429 0.06571
73 45430 3243 0.92862 0.07138
74 42187 3273 0.92242 0.07758
75 38914 3282 0.91566 0.08434
76 35632 3266 0.90834 0.09166
77 32366 3225 0.90036 0.09964
78 29141 3154 0.89177 0.10823
79 25987 3054 0.88248 0.11752
80 22933 2923 0.87254 0.12746
81 20010 2763 0.86192 0.13808
82 17247 2576 0.85064 0.14936
Continued on Next Page. . .
ELT 12 (Males) 255

x lx dx px qx
83 14671 2365 0.83880 0.16120
84 12306 2137 0.82634 0.17366
85 10169 1897.4 0.81341 0.18659
86 8271.6 1654.1 0.80003 0.19997
87 6617.5 1414.1 0.78631 0.21369
88 5203.4 1184.6 0.77234 0.22766
89 4018.8 971.6 0.75824 0.24176
90 3047.2 779.9 0.74406 0.25594
91 2267.3 612.2 0.72999 0.27001
92 1655.1 470.0 0.71603 0.28397
93 1185.1 352.73 0.70236 0.29764
94 832.37 258.83 0.68904 0.31096
95 573.54 185.74 0.67615 0.32385
96 387.80 130.39 0.66377 0.33623
97 257.41 89.59 0.65196 0.34804
98 167.82 60.30 0.64069 0.35931
99 107.52 39.771 0.63011 0.36989
100 67.749 25.733 0.62017 0.37983
101 42.016 16.349 0.61089 0.38911
102 25.667 10.209 0.60225 0.39775
103 15.458 6.2721 0.59425 0.40575
104 9.1859 3.7949 0.58688 0.41312
105 5.3910
Chapter 12

Pensions and Related Benefits

A person’s lifetime can be seen as having three distinct phases from an income-
earning perspective:
• pre-working age: this is when you focus on your education; your financial
needs are supported by other people (usually your parents);
• working career: this is the stage when you generate income to meet ex-
penses, to acquire wealth and property, to pay for children’s education and
so on; and
U Did you know that retirement
• retirement: this is the stage when you no longer earn income from employ- was only “invented” around
ment and must meet your expenses from savings (or otherwise fall back on 1880? Before then, most peo-
the state or extended family). ple worked until they died.

Each of these phases supports the next. One of the purposes of education, but
most definitely not the only one, is to allow greater income to be generated
throughout our working careers — it’s unlikely that most of you would be reading
these notes if actuaries didn’t earn good salaries. Likewise, it’s crucial during our
working careers to set aside savings which can be used to provide income during
retirement.
This chapter is all about the mechanisms for doing so. Why are actuaries in-
volved in this area? The answer has to do with risk and the uncertainty underlying
these questions:
• How long will we live after we retire?
• What level of income will we require?
• What inflation rate will prevail in retirement?
• What investment returns can be earned on retirement savings?
• How much money will we need to ensure that our needs are taken care of
after we stop working?

Example 12.1 Because of compound interest, starting to save earlier can

have a dramatic effect on the pension you will be able to afford.
" We have used the approxima-
tion
257
1+i
−1 ≈ i − j,
1+ j

to calculate these figures.

258 Chapter 12 Pensions and Related Benefits

Assume that you enter the workforce at the age of 25 and are planning to
retire at 65, and you start to save 10% of your salary every year. Let’s further
assume that you can expect to earn an investment return of 5% per annum
in excess of the rate at which your salary escalates. By the age of 65, you will
have saved an amount of 12.08 times your annual salary. However, if you
delayed starting to save by just five years, until you turn 30, the accumulated
multiple would only be 9.03, and if you don’t start until the age of 40 it
reduces to 4.8.

Because retirement funds are

not insurers, the terminology is
Exercises for 12 Pensions and Related Benefits
somewhat different:
• the people participating Ex.12.1 Using the approximation described above, show how the above multi-
in a retirement fund are ples were calculated.
referred to as members
and not “policyholders”— Ex.12.2 What percentage of your salary would you need to save to achieve
there are no “policies” as savings equal to 12.08 times your annual salary if you started saving at
such; age 30? What about if you start at age 40?
• the money paid by mem-
bers and other parties into
the fund is called contribu-
Replacement ratio
tions, not “premiums”;
• the amounts paid to mem- A common method of evaluating a pension is to express the initial retirement
bers on events such as income as a percentage of the pre-retirement income. This is known as the re-
retirement, death and with- placement ratio. The justification for this is that a person who is retiring is likely
drawal are called benefits, to compare his pension income to the income he gets before retirement. The
and not “claims” or “sums replacement ratio is defined as
assured”.
income in the year after retirement
replacement ratio = .
income in the year before retirement
Table 12.1 Retirement Fund Vo-
cabulary So a person who used to earn R20 000 per month and who then gets a pension of
R15 000 per month has a replacement ratio of 75%.

Exercises for 12 Pensions and Related Benefits

Ex.12.3 In the example above, you saved up 12.4 times your final salary. You
U An annuity factor is the cost
want to use these savings to buy a life annuity. The annuity factor
per Rand of annuity income
paid to the annuitant while quoted by a life insurer is 15.329. If you buy this annuity, what replace-
they are alive. You can think ment ratio can you achieve?
of it as the mortality function
ä x , but it will allow for the in-
Retirement funds are quite different from the insurance institutions we have
surer’s expenses and profits. considered until now. The main difference is that they are not commercial or-
ganisations, and they do not sell a product or generate a profit. Instead, they are
structured as trusts and operated by trustees who have been tasked to look after
the best interests of members.
12.1 What are people’s needs with respect to retirement, and how can those needs be met? 259

Organisational structure
A retirement fund is a separate legal entity from the sponsoring company. This
means that if, for example, the company goes bankrupt, its creditors cannot make a
claim against the pension fund’s assets; the members’ pensions are protected from
anything that happens to their employer’s business.
Legally, a pension fund is a trust and is governed under trust law. Trusts are man-
aged by trustees, who make decisions on behalf of the ultimate beneficiaries, the
fund’s members.
In South Africa, legislation requires that at least 50% of trustees are elected by fund
members, and no more than 50% are appointed by the employer.

Employer: Employees:
The sponsor of the pension The members of the pension
fund fund

Trustees
Representatives for the
employer and employees
make up the trustees
Employer Employee
contributions contributions

Pension fund

Benefits paid to the

members of the
pension fund from the
pension fund

Figure 12.1 Example of a pension fund structure where the employer

and employees contribute towards the employees’ benefits.

12.1 What are people’s needs with respect to retirement,

and how can those needs be met?
The major need in retirement is to have a source of income. Most people would T The amount of income
which is “sufficient” for a
like this income to be:
comfortable retirement
• inexpensive to fund for; is contentious. Some re-
searchers maintain that
• predictable to fund for, i.e. contributions should not vary from year to year;
post-retirement expen-
• as large as possible, but at least comparable to their pre-retirement income; diture is lower than pre-
retirement, because of a
• payable for life; the older a person is, the more likely it is that they need a
reduction in work related
secure source of income.
expenses and the need to
• predictable and stable, i.e. people would like to know what level of income save. Others claim that these
their pension fund will generate before they retire, making it possible to reductions are offset by in-
plan for retirement; creasing leisure costs and
healthcare costs. Generally,
• increase in line with inflation or some similar index so that the pensioner it is thought that 70%–100%
can maintain their standard of living in the long term. of the income before retire-
ment should be sufficient to
fund post-retirement spend-
ing.
260 Chapter 12 Pensions and Related Benefits

Some of these objectives conflict with each other. For example, trying to
make the pension as large as possible conflicts with the objective of making it
inexpensive to fund, and possibly making the income predictable if the pension
fund contributions are invested in risky investments.
T In addition to income, peo- Pension funds are vehicles for saving for retirement. The basic constituents of
ple also look to their retire-
a pension can be summarised as follows:
ment fund to provide some
level of lump sum on retire-
ment (to settle debt and the pension =
like), as well as death and contributions − expenses + investment returns − pre-retirement withdrawals
disability benefits. ,
annuity factor

where
• contributions (C ) are the payments made into the pension fund by the
member and possibly a sponsor, e.g. the member’s employer,
• expenses (E ) are the costs associated with running the pension fund, e.g. the
costs associated with making the benefit payments and managing the in-
vestments,
• investment returns (I ) are the returns earned on the invested contributions,
and
• pre-retirement withdrawals (W ) are the withdrawals made by a member
from the pension fund before retirement — these reduce the final pension a
person will receive.
? Can you think of any reason
why money might be taken
out of a pension fund before
retirement?

Figure 12.2 Illustration of the flows of contributions, benefits, ex-

penses, and investment returns in a pension fund.
To maximise the pension, you would need to maximise the contributions
you make, minimise the expenses incurred, maximise the investment returns
achieved, and minimise any withdrawals before retirement. In addition, you
? What are the two main fac- would also want to make the annuity factor as small as possible!
tors that will determine the
From equation above, you can see that investment returns play a role in the
value of the annuity factor?
final pension that you will receive. Seeking larger investment returns, in the
hope of maximising the final pension you receive, will usually exposes you to
12.1 What are people’s needs with respect to retirement, and how can those needs be met? 261

more risk — remember that the investments with the largest expected returns are
usually the most risky investments. So maximising investment returns may also
increase the possibility that the final pension you receive will be insufficient.

Example 12.2
Consider a person who is saving 15% of their income from age 25 until age 65 for
retirement. Assume there are no costs, salaries are paid annually in arrear, and
there is no pre-retirement mortality. There are two possible investment strategies:
• Strategy A: This is a high risk, high return strategy. The expected return is 4.5%
p.a. above salary inflation, but the returns may be as high as 7% p.a. or as low
as 0% p.a. over salary inflation.
• Strategy B: This is a low risk strategy. The expected return is 3% p.a. over salary
inflation, but the minimum and maximum return will be 2% p.a. and 4% p.a.
over salary inflation, respectively.
Using an annuity factor of 14, calculate the expected, minimum, and maximum
pension expressed as a percentage of salary for each strategy. Which strategy would
you choose? What other information might be useful to you in making this choice?
262 Chapter 12 Pensions and Related Benefits

Solution:
Strategy A:
Minimum return: 0% p.a. over salary inflation:
15%s 40 0% /14 = 42.86%.
Expected return: 4.5% p.a. over salary inflation:
15%s 40 4.5% /14 = 114.68%
Maximum return: 7% p.a. over salary inflation:
15%s 40 7% /14 = 213.89%.

Strategy B:
Minimum return: 2% p.a. over salary inflation:
15%s 40 2% /14 = 64.72%.
Expected return: 3% p.a. over salary inflation:
15%s 40 3% /14 = 80.79%.
Maximum return: 4% p.a. over salary inflation:
15%s 40 4% /14 = 101.81%

Your choice would depend on how tolerant you are of the risk of Strategy A — can
you survive on 42.86% of your pre-retirement salary? Do you have other savings
or investments that you can use? You would also want to know what the actual
probability distribution of returns is — how likely is it that you will get the minimum
return under each strategy, how likely is it that you will achieve less than say 2% p.a.
on each strategy (2% delivers a 65% replacement ratio which you could consider
acceptable?).

Exercises for 12.1 What are people’s needs with respect to retirement, and how
can those needs be met?

Ex.12.4 Consider someone who starts saving at age 45. What are their potential
outcomes under each strategy? Do you think a 45 year old may make a
different decision than a 25 year old?
Two types of retirement funds have been developed to meet these retirement
needs: defined contribution and defined benefit funds.

12.1.1 Defined contribution funds:

A defined contribution fund operates very much like a savings account with
regular amounts being saved on an ongoing basis. These pension funds are
? Why do you think contribu- called “defined contribution” or DC funds because the contribution amounts
tions are generally defined
are agreed upfront and are usually defined as a percentage of salary. The benefit
as a percentage of salaries?
amount which will be received at retirement is not known before the retirement
date.
12.1 What are people’s needs with respect to retirement, and how can those needs be met? 263

Example 12.3
Why is the amount of benefit unknown in a DC fund?

Solution: Consider the simplified funding equation we have defined:

C −E + I −W
P= .
a
In a DC fund the contributions are fixed. The other elements of this equation, I ,
a, W and E , are all variable. The annuity factor a is uncertain as it will depend ? Which of I , E , W and a do
on the prevailing interest rates and mortality at the date of retirement. As a result you think might have the
the pension a person will receive, P , will be variable — the value will depend on greatest effect on the pension
the investment returns achieved, the actual expenses incurred, the pre-retirement in practice?
withdrawals made by the member, interest rates, and mortality rates.

DC funds meet the following criteria:

• they can be used to try maximise the final pension received, e.g. by choosing
to invest the assets in high return investments, and
• they have contributions which are known ahead of time and so the contri-
butions are predictable.
The primary downfall of DC funds is that the final pension received is not reliably
predictable in advance. The main reason for this is that there is usually uncertainty
around the investment returns achieved, especially if the assets are invested in
high return, high risk investments, e.g. if the investment returns are poor the final
pension received may be insufficient.

Investments in a DC fund
Choosing the investment strategy for the retirement savings in a DC fund is obvi-
ously a key decision — it affects both the expected value of the pension that can be
obtained and the variability of possible pensions, i.e. the more risky the investment
strategy, the wider the range of potential outcomes.
The investment decisions can be made for the fund as a whole, by the board of
trustees of the fund, or they can be made by each member individually. Individual
member choice is becoming more and more popular as it is seen as giving mem-
bers flexibility and control over their own investments. However, it is becoming
increasingly apparent that few members have the financial skills to be able to select
appropriate investments. One solution to this problem are so called “lifestyle” or
“life stage” portfolios, where a member’s portion of the fund is moved from high
risk, high return to lower risk, lower return portfolios as they get closer to retirement.
This helps to make the final pension received more certain.

Example 12.4
Consider a fund with 3 investment options: “Conservative”, “Aggressive” and “Lifestyle”.
The Lifestyle option invests in the Aggressive option until age 60 and then switches
to the Conservative option until retirement at age 65. A person aged 30 joins the
264 Chapter 12 Pensions and Related Benefits

fund and is required to select an investment strategy. The expected returns are as
follows:
Conservative: This is a guaranteed fund which is expected to match salary
inflation.
Aggressive: This is an equity and bond fund and it is expected to outperform
salary inflation by 4% p.a.
(a) If the contribution to retirement, net of expenses, is 15%, what multiple of
salary will the member expect to accumulate at retirement if she selects each
strategy? Assume salaries are paid annually in arrear for simplicity, and ignore
pre-retirement mortality.
(b) Which strategy would you recommend? What other information would you
ask for to make your recommendation? What is the advantage of the life style
option? How could you improve on the lifestaging mechanism?

Solution:
(a) Accumulation under Aggressive option:

15% × s 35 4% = 11.048 times the final salary.

Accumulation under Conservative option:

15% × s 35 0% = 5.25 times the final salary.

Accumulation under Lifestyle option:

15% × (s 30 4% (1 + 0%)5 + s 5 0% ) = 9.163 times the final salary.

(b) It looks like the best return is achieved by the Aggressive option (obviously!).
But the above calculations do not take into account risk — the risk that the
portfolio may suddenly drop in value is particularly high with the Aggressive
option. The Conservative option does not have much risk attached to it (we
are told that returns are guaranteed), but the resulting accumulation is very
low.
The objective of the Lifestyle option is to remove the risk of the value dropping
just before retirement. However, this Lifestyle option is poorly designed: if a
100% switch is forced at age 60, the member is still exposed to the risk of a
market crash just before age 60.
A better design for a life staging portfolio is to spread the switch over a few
years. For example, the switch between the funds could involve moving 20%
of the portfolio into the Conservative option at age 60, another 20% at age 61,
and so on until the member is 100% in conservative investments by the age of
64. This staggered switch is intended to reduce the risk of switching just after
a market crash.

Exercises for 12.1.1 Defined contribution funds:

Ex.12.5 Calculate the expected multiple of salary on retirement if the switch is

spread over 5 years as suggested above.
12.1 What are people’s needs with respect to retirement, and how can those needs be met? 265

12.1.2 Defined Benefit Funds

The alternative to defined contribution funds are “defined benefit” funds, or DB
funds.
In the case of DB funds, the benefit at retirement is decided before retirement,
hence the name defined benefit. The benefits at retirement are most often de-
fined as a percentage of salary per year of service. For example, a DB fund may
have the benefit formula ? If the final salary is the salary
in the last year of work, what
α% × final salary × years service, replacement ratio will a per-
son who retires after 27 years
of work have in a DB fund
which is used to determine a members pension. The “α%” is called the “accrual
with a 2% accrual rate?
rate” — it is the rate at which the pension is earned over the service period. It
is usually set to a low number, e.g. 2%. The pension from a DB fund may also
increase in retirement, in which case the rate of increase is usually specified as
well.

Exercises for 12.1.2 Defined Benefit Funds

Ex.12.6 Marco has been a member of a retirement fund with a 2.5% accrual
rate for 17 years. He is now about to retire, and his final salary in the
month before retirement is R56 000 per month. What pension is he
due to receive if final salary is defined as the salary in the month before
retirement?

The equation for the benefit received still holds,

C −E + I −W
i.e. P = .
a
In the case of a DB fund, P is fixed. The E , I , W and a are still expected to vary
unpredictably. The contributions that have to be made into the fund will vary
over time so as to balance this equation, i.e. the contributions in a DB fund have
to be recalculated and possibly changed on a regular basis to ensure that the final
pension received is equal to the defined amount. These contribution calculations
are carried out by actuaries.

Example 12.5
Say a DB fund has members who all joined at age 30, are all 45 years old now, and
are due to retire at 65. The accumulated fund at present is equal to 1.5 times annual
salary for each member. The retirement benefit is equal to 1.5% of final salary
per year of service. If the assumption is that investments will return 3% p.a. over
and above salary increases, and that the annuity factor at retirement is 13, what
contribution rate is needed to fund the remainder of this benefit if the funding is to
be spread evenly over the full period until retirement? Assume no mortality or other
withdrawals pre-retirement, and for simplicity, salaries are paid annually in arrear.
266 Chapter 12 Pensions and Related Benefits

Solution: We start by calculating the total required funds at retirement to purchase

the pension for each member:

lump sum at retirement

pension = accrual rate × years service = .
annuity factor

Hence,

lump sum at retirement = accrual rate × years service × annuity factor

= 1.5% × (65 − 30) × 13
= 6.825 times final salary.

This means that the accumulated fund plus future contributions must end up being
equal to 6.825 times salary at age 65. +Note that we are using i = 3%, since we are
interested in how much more the investment returns are than the salary growth:

C × s 65−45 i + 1.5 × (1 + i )65−45 = 6.825

6.825 − 2.709
⇒C = = 15.32%.
s 20 i

So the contribution rate will have to be 15.32% of salaries to make sure that the
required pension is achieved.

Exercises for 12.1.2 Defined Benefit Funds

Ex.12.7 One year later, the accumulated funds per person have grown to 1.55
times salary.
1. Do you think the experience of the fund was better or worse than
expected?
2. Calculate the required contribution rate now (remember that the
members are all 46 years old now). Is the contribution rate lower
or higher than before?

DB funds meet some of the objectives of retirement funding that we listed

above — pensions are very predictable, and given the right benefit design they
can be sufficient. However, there is no opportunity to maximise the pension, and
the contributions are not stable and predictable from year to year.
This last problem, the unstable contributions, is considered unacceptable
to the members (employees). Members do not like being told “This year, you
will have to save 20% of your salary, even though last year it was 10%”. This
kind of uncertainty around future contribution rates does not typically fit in with
members’ budgeting and spending patterns. So how can DB funds be made to
work?
DB funds require that one of the sponsoring parties be willing to make variable
contributions. It is usually the employer that accepts to pay variable contributions,
while the member makes defined contributions. The reasons for this are:
12.2 Accessing retirement funds 267

• employers are more able to undertake the variable payments;

• employers are also larger, and therefore they can benefit somewhat from
the law of large numbers — what would be highly variable contribution rate
for one member over time, may be considered less variable for an employer
with thousands of employees.
However, nowadays employers are less tolerant of the variability of contributions,
and as a result DB funds are no longer very common in many countries.

DB Funds then and now

DB used to be the most common method of retirement funding in South Africa
and world-wide. This type of fund fits very well with the view of the employer as a
paternalistic, protective figure, and with the old model of “job for life” where many
workers had stable careers with just one or two employers over their working life.
As work practices changed, employees became more mobile, and also more inter-
ested in maximising investment returns for themselves, DC Funds became more
common. In South Africa today, only a handful of DB funds remain, and DC funds
have become the norm.
However, DC does carry more risk for the employee: the benefit is not known in
advance, and it is difficult for someone to know whether they are saving enough.

12.2 Accessing retirement funds

Retirement funds are not for profit institutions, and have no formal sales strategies.
However, there are different types of providers, and the nature of the provider
determines how different retirement vehicles can be accessed. The three main
providers of retirement benefits: the state, employers and similar organisations,
and insurers providing retirement annuities. In almost all countries, the state
has some role to play in retirement provision, although the extent of this role
varies widely between countries.
State benefits are often supplemented by benefits from occupational pension T Organisations which can
provide pensions include
funds provided by employers and similar organisations. In addition to this, in-
trade unions, bargaining
dividuals may choose to save more, either through retirement-specific savings, councils, and other industry
such as through retirement annuity funds in South Africa or through general bodies.
savings. If these sources of savings combined are insufficient to meet needs in
retirement, the additional burden will have to be carried either by the state (e.g.
through use of state healthcare resources rather than private medical care) or by
the extended family or community of the individual. One of the principal aims
of the retirement funds industry is to assist people to make sufficient savings for
retirement, thus reducing this risk.
268 Chapter 12 Pensions and Related Benefits

12.2.1 State retirement provision

The state often provides a pension to its citizens. Such pensions differ from
country to country — in some countries it is a flat amount per person, elsewhere
it may be related to an individual’s earnings before retirement or some other
criteria. Sometimes this benefit is fully funded by the state, and sometimes
people are required to contribute as well.
Some states offer no pensions at all. Even in those countries, the government
will still have an influence on the retirement savings market through regulations
and through setting a tax regime which is likely to include incentives for retire-
ment savings. Regulation is necessary because pension funds are vehicles which
accumulate very large assets, and governments will want to ensure that only fit,
proper, and appropriately qualified persons are managing them, and that their
management is monitored.
At the other extreme, the state could provide all post-retirement benefits,
without making any allowance for private pension funds. In practice, most states
are on a spectrum between these two extremes. In South Africa, the level of
U Means-tested means that state provision is currently very low, extending only to the State Old Age Pension
in order to receive a benefit,
(SOAP) which in 2013 was R1 260 per month. The SOAP is means-tested, with
the applicant must demon-
strate that they have less maximum levels of assets and income set in order for individuals over the age of
than a specified amount 60 to qualify to receive it.
of assets and/or income.
Means-tests can lead to 12.2.2 Occupational retirement funds
morale hazard — can you
see why? It has become very common around the world for employers to provide a vehicle
and a contribution towards their employees’ retirement savings, called an occu-
pational fund. The government often encourages this, for example by making
U The terms “retirement fund”, contributions to employer funds tax-exempt. Employers who decide to start up a
“pension fund”, and “retire-
pension fund generally make it compulsory for employees to join the fund.
ment scheme” are generally
used interchangeably. South
Africa also has “provident 12.2.3 Individual provision
funds” as distinct from “pen-
sion funds”. The main differ- Finally, people who do not have an employer fund, or who want to save more
ence is that pension funds than what the rules of their occupational fund allow, can save up for retirement
pay out a benefit that is at themselves. They could do this by putting money into a savings or investment
least 2/3rds pension, while account, but it is generally more tax efficient to use a retirement annuity product
provident funds can pay out from an insurer. Those products are life insurance policies, and they are marketed
the retirement benefit as
and distributed in the same way as other life insurance products. Legislation
100% cash. The government
allows for tax relief on contributions made to such vehicles.
is currently proposing to
phase out these provident We can extend Figure 12.1 to include the additional sources of retirement
funds — can you think why? provision we have mentioned here; see Figure 12.3. In this Chapter, we mainly
focus on employer sponsored retirement schemes, also called “occupational
funds”.
12.3 Underwriting 269

12.3 Underwriting
Since retirement funds generally pay out the accumulated savings either on U Consider also that occupa-
retirement or earlier exit, there is not much risk involved — if everyone decided to tional funds are effectively
group arrangements, which
leave a retirement fund at once, the amount due to them would be exactly covered
generally do not suffer from
by the assets of a DC fund, for example. Even for a DB fund, the mismatch would
anti-selection.
be limited. The opportunity for anti-selection is therefore limited for retirement
products, and there is no need for underwriting.

12.4 Setting contribution rates

For occupational DC funds, this process is relatively simple. The contribution T The term “pricing” is not
rate is set by the employer when the fund is set up. The factors that are taken into used in retirement funding
— the correct terminology
account are:
here is “setting contribution
• the amount the employer and members are willing to contribute; rates”.
• the expenses of the fund;
• the cost of other benefits offered by the fund, which will also be funded from
the contribution (for example additional death benefits);
• the expected investment returns which can be earned by the fund, and thus
the contribution rate required to generate a target replacement ratio;

Figure 12.3 This is an extension of Figure 12.1 which includes the additional sources of
retirement provision mentioned here. The grey shaded area encompasses the compo-
nents that form part of the occupational fund.
270 Chapter 12 Pensions and Related Benefits

• what the normal contribution rates are in the market, for example, what are
the contribution rates in other occupation funds?
U For example, in SA only con-
tributions up to a certain • Any regulatory restrictions or limitations on the contribution rate.
maximum are tax exempt —
employers therefore set up
their funds in such a way as Example 12.6
not to exceed this limit. PesCo is an employer who wants to set up a retirement fund for its staff. An internal
study has been conducted with the following results:
• PesCo is willing to contribute 15% of salaries. Employees are willing to con-
tribute 7.5% of salaries.
• There is demand for a death benefit of 3 times annual salary on death. The
best quote from an insurer for this benefit is 2.4% of salaries p.a.
• Administration and other costs will amount to 1.5% of salaries p.a.
• The fund is going to invest in a moderate risk portfolio with an expected return
of CPI+5% p.a.
• Employees experience salary increases of around CPI+1.5% p.a.
• The employer and employees feel that a person with 40 years service should
retire with a pension of at least 85% of their salary.
Calculate the recommended rate of contribution given these constraints. Ignore
pre-retirement mortality, and assume that a pension can be purchased at retirement
using an annuity factor of 13.75.

Solution: Starting from the target, and assuming all cashflows happen annually in
arrear, for simplicity:

pension = final salary × target replacement ratio.

Now we equate the accumulated contributions with the pension

salary × contribution rate × s 40 j

= salary × 85%, (12.1)
13.75
where j = 0.05 − 0.015 = 3.5% using the approximate net interest rate.
Salary cancels out on both sides of Equation (12.1), giving:

contribution rate × 84.550

= 0.85
13.75
⇒ contribution rate = 13.823%.

This is the contribution that will generate a sufficient replacement ratio. In addition,
we need contributions to cover death benefits and administration costs:

gross contribution rate = 13.823% + 1.5% + 2.4% = 17.723%.

The employer and employees together are willing to pay 15% + 7.5% = 22.5%, which
is sufficient to achieve all the objectives required by the client.
12.4 Setting contribution rates 271

Exercises for 12.4 Setting contribution rates

Ex.12.8 An alternative approximation for the interest rate j in Example 12.6

would be j = 1.05/1.015 − 1. Recalculate the contribution rate in the
example, and compare your answer to the rate we found above.
Ex.12.9 What replacement ratio will be achievable after 40 years of service with
a contribution of 22.5% in the above fund?

For DB funds, setting contributions is a more complex actuarial process.

Contributions are calculated regularly, and increased or reduced based on:
• the benefit structure of the fund: the accrual rate, the expenses and the cost
of other benefits;
• the profile of the membership of the fund: this determines how much must
be contributed to fund the benefits that will accrue in the future (future
service benefits).
• the funding level of the fund: this determines if extra contributions are
needed to fund for benefits that have accrued in the past that are not suffi-
ciently funded for by existing assets.

Accrual
One of the most interesting actuarial concepts in DB pension funding is accrual.
Accrual refers to the gradual build up of benefits over time: a member who just
joined the fund today has built up no benefit yet. If he leaves tomorrow, he will
receive no benefit from the fund.
A member who was in the fund for one year will qualify for a pension based on one
year of service. A member who worked for 10 years can get a pension based on
10 years of service, and so on. Therefore, at any given point in time, the accrued
liabilities refer only to the past service of the member; and liabilities grow for each
member over the years until the full benefit is accrued by the time they retire.
This is very different to insurance, where the policyholder generally qualifies for
the full sum assured from the onset of the policy. This means that for example for
whole-life insurance, any claims payable early in the term of the policy are much
higher than the total premiums received from the policyholders, but claims payable
late in the policy may be comparable to total premiums received.

The funding level is the ratio of the value of the assets to the value of the
liabilities of the fund,

value of the assets

i.e. funding level = .
value of the liabilities
The funding level is affected by the actual experience of the fund and the assump-
tions used to value the liabilities. The actual experience consists of the investment
returns earned, the actual expenses of the fund, and the actual benefits paid (par-
ticularly pre-retirement benefits).
272 Chapter 12 Pensions and Related Benefits

12.5 Benefits
The main purpose of a fund is to pay retirement benefits. Some funds pay the
retirements benefit as a pension from the fund, i.e. regular payments to the
member while they are alive. Other funds pay out the retirement benefit as a
U Paying pensions from the lump sum on retirement. Depending on legislation and people’s preferences, this
fund was common in DB
lump sum may then be used to purchase an annuity from an insurer.
funds, but it is rare for DC
funds in South Africa to pay It is important to realise that not every member who joins a company, and
pensions. hence its pension fund, is going to continue in that employment until retirement.
In particular, there is a risk of death before retirement, of ill-health or disability
preventing them from continuing to work. Most commonly, people leave employ-
ment, either voluntarily (going to another job) or involuntarily (retrenchment or
dismissal), before reaching retirement. It is only reasonable that members will
expect that in the event of these contingencies, they or their dependants will be
entitled to some form of benefit from the fund. We consider three types of benefit
that may be paid from a fund before retirement, namely death benefits, ill-health
and disability benefits, and withdrawal benefits.

Death benefits
U For example, a fund may Pre-retirement death benefits may be provided in the form of a lump sum, or
offer the accumulated mem-
as an annuity benefit to dependants (usually paid to the member’s spouse and
ber share plus 5 times the
annual salary of the member children). In most funds, the benefit payable on death is more than the accrued
on death. pension benefit under the fund. This means that the fund faces mortality risk;
the risk to the fund is that more members die than expected, and so the fund
U A very large fund can choose will have to pay out more death benefits than expected. This risk is most often
to self-insure, in other words
transferred to an insurance company. The fund does this by buying a group life
start its own internal risk
policy for the members.
pool, to provide for the
death benefits. But most
funds are too small to do Ill-health and disability benefits
this, and there is a trend
in the market to make re- Many DB pension funds offer enhanced benefits on early retirement for reasons
tirement funds as risk free of ill health. The sponsoring employer is thus exposed to the risk of more ill-
as possible for employers, health early retirements than expected. An alternative approach, used by most
which means that using in- DC funds and an increasing number of DB funds is to take out group disability
surance has become the insurance contracts with life insurers to provide disability income protection or
norm. lump sum benefits for the members. This again transfers the risk from the fund
to the insurer.

Withdrawal benefits

Fifty years ago, it was commonplace for individuals to spend their entire working
careers with one employer. These days, however, a typical working career might
span five or six (or many more) employers, and an individual saving for retire-
ment who is leaving one employer for another will wish to take their accumulated
12.5 Benefits 273

retirement savings with them. These could be transferred to another fund, for ex-
ample the new employer’s fund, or a retirement annuity product. Or, if legislation
permits, they may be withdrawn as cash.

Preservation of retirement savings in South Africa

Unfortunately, South African legislation permits the payment of withdrawal bene-
fits in cash. The government encourages transfers ahead of cash withdrawals by
taxing cash withdrawals at a high rate. Despite this incentive, however, an alarm-
ingly high proportion of withdrawing members choose to be paid their benefits in
cash because of the obvious attraction of more consumption power today, rather
than at an uncertain future date. The net result is that many individuals arrive at
retirement with hopelessly insufficient savings to be financially independent.
Simply legislating a requirement for pension benefits to be preserved would seem
to be the straightforward logical answer to this problem, but of course the situa-
tion is more difficult than that. In South Africa’s history, many withdrawals have
been involuntary due to widespread retrenchments, and defenders of the status
quo argue (legitimately) that for many low-income workers, spending retirement
savings may be the only way of supporting themselves and their families in such
circumstances. As a result, it has taken a long time for the government to find a way
to stop the payment of withdrawal benefits in cash — there is now new proposed
legislation that aims to phase out this practice over a period of about 10 years.

A History of South African withdrawal practices

Until a few decades ago, it was common for DB funds to pay withdrawal benefits
which were much lower than the member’s accrued pension benefit. For example,
funds would pay out only the member (and not the employer) contributions, in-
creased with some interest rate, which was often low and bore no relation to the
actual investment returns of the fund. Funds that used a vesting scale to increase
the withdrawing members benefits according to their past service provided slightly
better benefits. A vesting scale works by increasing the proportion of the employer’s
contributions that the member receives on withdrawal according to the length of
their service; the longer the member’s service with the employer, the larger the
proportion of employer’s contributions will be paid out as part of the withdrawal
benefit.
Inevitably, this approach means that withdrawing members do not get the full
benefit that had built up for them. The remainder of the benefit stayed in the fund.
Over time, many funds, especially those with a high staff turnover, accumulated very
large surpluses caused by the withdrawals. Dissatisfaction with this arrangement
was one of the prime motivators for unions to push for the conversion of funds from
DB to DC, where the accumulation of funds is more transparent, and to ensure that
withdrawal benefits more fully reflect what has been put aside for a member.
Legislation now outlaws such reduced withdrawal benefits, requiring that all with-
drawing members receive the full member account.
274 Chapter 12 Pensions and Related Benefits

Exercises for 12.5 Benefits

Ex.12.10 List all the different benefits that may be paid from a retirement fund.
Ex.12.11 A retirement fund offers a death benefit of 3 times annual salary plus
the member’s accumulated retirement savings. The death benefit takes
the form of an annuity payable to the spouse and/or children of the
member. What type of life insurance product would the retirement
fund need to purchase to transfer the mortality risk to the insurance
company?

12.6 Monitoring
Retirement funds need to be regularly monitored, just like any other financial
institution. Actuaries monitor the performance of retirement funds by conducting
regular valuations of the funds in order to:
• monitor the funding level, i.e. the ratio of assets to liabilities;
• in the case of DB funds, recommend the appropriate employer contribution
rate;
• in the case of DC funds, review the allocation of member and employer
contribution to individual member accounts;
• monitor the investment performance of the fund against its targets, and
check that assets are matched to liabilities;
• monitor the experience of the fund. For a DC fund, this is limited to in-
vestment returns and expense experience. For DB funds, salary increases,
mortality and withdrawals (i.e. pre-retirement benefits), and if pensions are
paid from the fund, also the post retirement mortality need to be measured
and compared against assumptions.
The valuation process is an invaluable tool for pension fund trustees to be able
to assess if the fund is operating according to expectations, and to be able to
make timely interventions should they be required. A DB fund valuation requires
actuarial judgement in setting the assumptions about the future experience of
the fund. Valuations are usually either performed on a prudent basis, or on a best
estimate basis with a separate margin for prudence.

Example 12.7 An actuary has valued the liabilities of a pension fund on a

conservative basis to be R120.34 million. A valuation of the same liabilities
on a best estimate basis gives a result of R99.20 million. The fund assets are
R150 million.
The actuary can report this either as:
Liabilities: R120.34 million
Assets: R150.00 million
Surplus: R29.66 million
12.6 Monitoring 275

Funding level: 124.65%

Or alternatively:
T The second approach makes
Liabilities: R99.20 million the margin for prudence
Reserve for prudence: R 21.14 million more explicit and easier to
identify, but the end result is
Total liabilities: R120.34 million
the same.
Assets: R150.00 million
Surplus: R29.66 million
Funding level: 124.65%

Exercises for 12 Pensions and Related Benefits

Ex.12.12 Calculate the funding level for a fund with:

1. Assets of R1.17 billion, liabilities on a best estimate basis of R1.23 bil-
lion and a reserve for prudence of R200 million;
2. Assets of $305 million, liabilities on a prudent basis of $250 million
which include an allowance for prudence of $20 million.
Chapter 13

Ethics and Professionalism

13.1 Introduction
Now we’re going to step aside from the development of technical skills and com-
mercial awareness, on which we’ve focussed our attention so far, and talk about
an element which is just as important to your professional careers. Throughout
this programme, and in your future career, enormous weight will be put on actu-
aries and quants being trustworthy professionals who can be relied on to act with
integrity at all times.
As an indicator of the importance of ethical behaviour, this is what the CFA
Society South Africa says about its objectives:
U The CFA, which stands for
“CFA Society South Africa promotes ethical and professional stan- Chartered Financial Analyst,
dards within the investment industry, encourages professional de- is an international qualifi-
velopment through the CFA Programme, and facilitates the open ex- cation administered by the
CFA Institute based in the
change of information and opinions. . .
United States which is com-
“Established in 1962, the CFA Programme sets the global standard for pleted by many quants, as
investment knowledge, standards, and ethics. Earning the credential well as many others in the as-
can serve as a passport to entry or advancement within the investment set management and related
profession around the world. The designation tells clients, employers, industries.
and colleagues that the charterholder has mastered a rigorous curricu-
lum covering a broad range of investment topics and that he or she is
committed to the highest ethical standards in the profession.”

You can see that in addition to technical skills and investment knowledge,
there is an overwhelming emphasis on ethical conduct. Let’s take a moment to
remind ourselves of the actuarial control cycle, which is at the heart of actuarial
practice:

277
278 Chapter 13 Ethics and Professionalism

Figure 13.1 Actuarial Control Cycle.

You will see that professionalism underlies all actuarial practice; every single
phase in the control cycle ought to be underpinned by professional behaviour. So
why exactly are ethics and professionalism so important in our field of work?
Of course, anyone providing a service of any kind should act with integrity
and honesty. You would like your plumber and auto mechanic to give best advice
at reasonable cost, and you would hope that your waiter doesn’t add extra items
to the bill at the end of a long night. But professionals, like doctors, lawyers,
engineers, actuaries and quants are often held to an even higher standard. Why is
that?

Exercises for 13.1 Introduction

Ex.13.1 Before reading further, why do you think this is the case?
Like much in this course, it relates back to risk. You will find that some types of
professionals operate within a higher risk environment (remember that high risk
means high uncertainty or high value, or both): when a plumber makes a mistake,
the kitchen may get flooded; but when a doctor makes a mistake, a person may
suffer or die (that’s very high value!). Actuaries and quants deal with events that
are very uncertain and generally have high value attached to them; and both the
uncertainty and value are complex to measure.

Example 13.1 If actuaries and quants did not behave professionally, there
could be many opportunities for personal gain:
• Inflate the value of income lost as a result of an accident, to get your
client more money from the person who caused an accident;
• Use a low estimate for the risk of a business venture, in order to secure
funding from the parent company;
13.2 Professional societies and codes of conduct 279

• Use information obtained as part of your job as a share analyst to trade

shares before the market gets the information (insider trading);
• Hide information about a company’s financial problems or misconduct
Actuarial Society of South Africa -
in order to prevent a scandal which may cause the company to collapse
Introduction to the Code of Con-
and you to lose your job. duct

Members are expected to render

quality services to their clients
Exercises for 13.1 Introduction through:
1. The application of special-
Ex.13.2 Can you think of other opportunities that might exist for actuaries ist and up-to-date actuarial
to benefit personally by acting unethically, in the practice areas that knowledge and expertise;
you’ve learned about on this course so far? 2. The demonstration of ethi-
cal behaviour, especially in
In all of the above cases in Example 13.1, it would be very difficult for a non- doing actuarial work; and
expert to identify the deceit — only another skilled financial professional may be 3. The member’s account-
able to see through the complexities of the calculations involved and realise that ability to the Society for
you have been acting improperly. professional oversight.
The principles of professional
Exercises for 13.1 Introduction conduct therefore include:
1. Knowledge and expertise:
Ex.13.3 Do you think that ethical issues for actuaries and quants arise only in A member shall perform
situations where there is an opportunity for personal financial gain? If only those professional
not, what other situations do you think might lead to ethical decisions services for which the
member is competent and
having to be made? appropriately experienced.
So how can we make sure that actuaries and quants are trusted not to abuse 2. Values and behaviour: A
member shall act honestly,
their position?
with integrity, competence
and due care, and in a
manner that fulfils the
13.2 Professional societies and codes of conduct profession’s responsibility
to the public.
For most professions, this is accomplished by professional societies which govern
3. Professional accountability:
their members using codes of conduct. A member is subject to the
We have already referred to some of these Societies, but there are many more, professional requirements
for example: and oversight of the Soci-
ety, and shall do nothing
• the Actuarial Society of South Africa ([Link]) – the box that brings the actuarial
on the right outlines the Society’s Code of Conduct profession into disrepute.

• the Chartered Financial Analysts (CFA) Institute ([Link])

Table 13.1 ASSA Code of Conduct
• the Casualty Actuarial Society ([Link]) and the Society of Actuaries
([Link], in the US (the US still has two separate bodies, with CAS
specialising in short-term insurance and the SoA covering the other areas of
actuarial practice))
• the Institute and Faculty of Actuaries ([Link]) in the UK (this
used to be two bodies, The Institute in England and the Faculty in Scotland;
when they merged, they could not agree on a combined name!)
280 Chapter 13 Ethics and Professionalism

• the Actuarial Society of India ([Link])

• the Institute of Actuaries of Australia ([Link])

Exercises for 13.2 Professional societies and codes of conduct

Ex.13.4 Preparation for ethical behaviour should start early. When it comes
time for you to graduate, you will probably want your lecturers to write
letters of reference on your behalf for potential employers, and you
will need the Head of Actuarial Science to certify that you are a fit and
proper person to enter your chosen profession. How willing do you
think your lecturers will be to sign off for you in each case if during
your university career you are found guilty of the following:
1. Stealing money from a classmate
2. Cheating on a test
3. Being arrested following a peaceful protest against government
corruption
4. Signing the register for an absent classmate at a compulsory tuto-
rial (or being the absent classmate and being signed for)
5. Plagiarising part of your fourth-year research project from a Mas-
ter’s thesis submitted in Albania

Despite the number of actuarial societies around the world, there is actually a
trend towards more unified standards for education and conduct: the Interna-
tional Actuarial Association ([Link]), which has as its members the
actuarial societies of most other countries, is driving the adoption of core values
and standards.

Exercises for 13.2 Professional societies and codes of conduct

Ex.13.5 What do you think are the benefits of more unified standards through
the efforts of the IAA? Are there any risks or downsides to this?

Such societies take on the role of a supervising body for their members, ensur-
ing that all members meet standards of expertise and professional conduct. If the
societies consistently implement this, over time other parties, such as potential
clients and the government, come to trust that any member of a particular society
is likely to know what they are doing, and do it with integrity. For example, in
South Africa, the Actuarial Society operates as follows:
• In order to become a “fellow” of the Society, you have to pass (or be exempted
from) exams set by the Society. This ensures all actuaries in the Society have
the required technical skills.
• Once you are a member of the Society, you must follow the Code of Conduct
of the Society. This requires, amongst other things, that you act ethically
and professionally at all times.
13.2 Professional societies and codes of conduct 281

• If you do not follow the Code, the Society may discipline you: this includes
anything from a warning, to a fine, a requirement for relevant training, a
suspension or even expulsion from the Society.
• Members of the public or companies looking for an Actuary may check
whether anyone claiming to be an actuary actually is a member of the
Society. If the actuary is a member of the Society, the person hiring them
can be confident that the actuary has the technical and ethical credentials
promised by the Society.
• The government has gone one step further: there are a number of regula-
tions, for example the Pension Funds Act, that specify certain tasks that may
only be performed by an Actuary. The regulations specifically refer to an
Actuary as someone who is a Fellow of the Actuarial Society of South Africa.
This implies that the government trusts the Actuarial Society to supervise its
members and ensure that we are qualified to do the job.
This system ensures that anyone who is a member of the Society can be
trusted to have a certain level of expertise and ethics. As long as clients and
employers confirm that they are dealing with a member of the Society, they
are protected by the Code of Conduct and can have certain expectations of the
actuaries they deal with. But note that anyone can claim to be an actuary: the Continuing Professional Develop-
Society has no power to stop someone who is not a member from doing or saying ment (CPD)
anything. That’s why it is important to check the credentials of any professional You might think that once you
you interact with. have graduated and written a few
We conclude this section by considering the conduct frameworks of two other more actuarial exams to qualify,
bodies which regulate many graduates of this programme, namely the Institute that would be it for actuarial edu-
cation, but you’d be very wrong!
and Faculty of Actuaries in the UK and the CFA Institute.
Part of ensuring that actuaries live
up to their professional promise
13.2.1 The Actuaries’ Code to the public is requiring qual-
ified actuaries to demonstrate
The Institute and Faculty of Actuaries have put forward the Actuaries’ Code as the that they are continuing to keep
standard to which they expect their members to aspire. Its aim is to ensure ethical their skills sharp by showing com-
pliance with CPD requirements
conduct principles which members are expected to follow, in the public interest on an ongoing basis. You can
(a concept to which we will return in the next section), so that the public will have read more about this, if you like,
confidence in the actuarial profession and with work of individual actuaries. at [Link]
The core principles outlined in the Code are: [Link]/Professionalresources/
[Link].
• Integrity: members should act honestly and with the highest standards of
integrity.
Table 13.2 Continuing Profes-
• Competence and Care: members must perform their duties competently sional Development
and with care.
• Impartiality: members must not allow bias, conflict of interest or the undue
influence of others to override their professional judgement.
• Compliance: members must comply with all relevant legal, professional
and statutory requirements, take reasonable steps to ensure they are not
282 Chapter 13 Ethics and Professionalism

placed in a position where they are unable to comply, and will challenge
non-compliance by others.
• Communication: members must communicate effectively and meet all
applicable reporting standards.
The similarities between these requirements and the Code of Conduct of the
Actuarial Society of South Africa are apparent; this is not surprising since both
bodies regulate actuaries who work in very similar environments, and both bodies
have very similar aims in terms of the professional standards to which they expect
their members to adhere.
T More information can be
found at [Link].
[Link]/regulation/pages/ Exercises for 13.2.1 The Actuaries’ Code
actuaries-code.
Ex.13.6 What similarities do you see between the South African and UK require-
ments? Are there any differences you find interesting?

13.2.2 The CFA Institute

The CFA Institute has a two-pronged approach to ensuring professional conduct:
members must adhere to a Code of Ethics and to Standards of Professional Con-
duct. For those considering pursuing qualification as CFAs after graduation, it’s
worth noting that these apply equally to candidates for the CFA qualification;
in other words, you will be expected to behave professionally from the first day
you’re associated with the Institute.
The Code of Ethics maintains that members must:
• place the integrity of the profession and the interests of clients above their
own interests;
• act with integrity, competence, and respect; and
• maintain and develop their professional competence.
These are high-level objectives, whereas the Standards of Professional Con-
duct set out fairly detailed requirements for professional behaviour by CFA char-
terholders and candidates. While this detail is beyond this course, you may note
that the Standards cover the following areas:

• professionalism and integrity of the capital markets;

• duties to clients and employers;
• investment analysis and recommendations; and
• conflicts of interest and responsibilities of CFA Institute members and can-
didates.
U More information can be
found at [Link].
org/ethics/codes/Pages/ Interesting to note:
[Link].
The CFA Institute has a much more formalised approach to introducing ethics into
the curriculum from an early stage than the actuarial bodies. From Level I of the CFA
13.2 Professional societies and codes of conduct 283

exams, 20% of the marks are devoted to ethics and professionalism, and candidates
are expected to know and understand the detailed provisions of the Code and the
Standards of Professional Conduct, and be able to apply them to practical situations.
The actuarial bodies have historically adopted a more pyramid-style approach,
with technical skills at the base on which are built commercial and environmental
knowledge and then professionalism. The UCT Actuarial Science programme has
for several years made an effort to introduce our students to ethical issues from
first year (which is why you’re reading this now!) and the Actuarial Society of
South Africa is revising its educational approach to ensure that so-called normative
skills (which include professionalism as well as skills such as communication) are
incorporated into the curriculum, to the appropriate degree and at the appropriate
level, through all phases of actuarial education.

Exercises for 13 Ethics and Professionalism

Ex.13.7 Let’s suppose that you work for a consultancy which advises pension
fund trustees on their investment strategy; you make investment man-
ager recommendations from the entire market. You are advising a large
client looking to place R500 million of assets (that’s a lot of money!)
with an asset manager. Your company also owns a 50% stake in a
prominent asset manager called BAM (Brilliant Asset Management).
BAM is a top-quartile performer over 3 and 5 years, although it has
bottom-quartile performance in the last year. You would objectively
consider BAM on a shortlist of 5 managers, but there are at least two
other managers you think are better.

Terminology note:
As you know, asset managers manage money in investment portfolios
on behalf of external clients, whether individuals or, as in this case,
financial institutions such as pension funds or insurance companies. It’s
common practice to divide all asset managers being considered by their
performance relative to the market: a top-quartile performance simply
means that the investment returns delivered by that asset manager were
in the top 25% compared to all their competitors, and bottom-quartile
is the opposite. Clearly you’d like your clients’ performance to be in
the top quartile, but as illustrated in this example, there’s no guarantee
that top-quartile past performance will necessarily result in top-quartile
performance going forward.

Your options are as follows:

1. strongly recommend BAM to your client, with no disclosure of
your company’s ownership stake
2. strongly recommend BAM, but disclose the ownership stake
3. present BAM as one of five options, with no disclosure of the
ownership stake
284 Chapter 13 Ethics and Professionalism

4. present BAM as one of five options, with disclosure

5. present BAM as one of five options, disclose ownership and explic-
itly recommend other managers
6. exclude BAM from your shortlist because of the conflict of interest
Which choice would you make? How would you justify it?

13.3 What does it mean to be “professional”?

Having examined codes of conduct, let’s return to the rationale for them. So
far, we have used the term “professional behaviour” quite casually. Most of us
have some sense of what this means, but it’s useful to think about it a little more
deeply. Obviously, the words “professional” and “profession” share a common
root, and when we talk about professional behaviour, we are talking about the
sort of actions that we would reasonably expect from a member of a profession
(such as an actuary or a quant). So that begs the question: what do we mean by a
“profession”? What does it mean to be a member of a profession?

Exercises for 13.3 What does it mean to be “professional”?

Ex.13.8 What other professions do you know of?

The first, obvious point is that members of a profession are members of a

fairly exclusive club; there are requirements which have to be met before entry
to the profession, and the professional body then represents members of the
profession. One area in which this representation takes place is in the negotiation
of areas of work which may only be performed by members of the profession. So
in some ways, you could consider a profession to be very similar to a trade union:
representing the collective interests of members and bargaining on their behalf.
But our common vision of a “profession” is somehow different to that of a trade
union. What, then, sets professions apart from trade unions?

Exercises for 13.3 What does it mean to be “professional”?

Ex.13.9 What do you think is the defining, distinguishing characteristic of pro-

fessions that sets them apart from trade unions?

You might well have come up with a variety of answers to that question. For
example, you might have referred to the educational requirements of entry to a
profession. But many trades also require some skill or qualification in order to be
admitted to the union. A useful way to think about this is to consider the list of
characteristics of a profession proposed by Lord Henry Benson, which are that a
T Baron Benson was a famous profession:
accountant who played a
significant role in the devel- • is controlled by a governing body which directs the behaviour of its mem-
opment of the accounting bers;
profession in the UK (and,
for interest, was born and
raised in Johannesburg!).
13.4 Summary 285

• sets adequate standards of education as condition of entry and ensures that

professional competence is attained;
• sets ethical rules and professional standards;
• is designed for the public benefit;
• takes disciplinary action, including expulsion;
• reserves work by statute in the public interest;
• allows fair and open competition in the practice of the profession;
• has members who are independent in thought and outlook; and
• gives leadership to the public in its particular field of learning.

Example 13.2 Do you notice any consistent theme running through that
list?
Look at how many times there is (direct or indirect) reference to the public.
The actuarial profession in the UK has in fact identified that what distin-
guishes professional bodies from trade bodies is that one of their primary
objectives is to act explicitly in the public interest, in addition to supporting
the needs of their members. And if this is at the heart of what it means to
be a profession, then acting in the public interest must be at the heart of
professional behaviour.

In this context, the codes of conduct discussed above make even more sense:
these are the standards which we must uphold if we are to act in the public
interest, and if we do so, then the public ought to have confidence in the work
we do. The difficulty, of course, is that in many circumstances it’s not entirely
clear what action is in the public interest, and this is a term which is open to
interpretation. In our lectures, we will discuss some of the difficulties that arise.

13.4 Summary
Hopefully the discussion above has got you thinking about what it means to
act ethically and professionally, and the importance of doing so (in addition
to having all the technical skills and commercial awareness that actuaries and
quants are expected to have). Many ethical dilemmas are not clear-cut, and it
is very likely that at some point in your career you will be exposed to a situation
(perhaps many) where you will be forced to ask yourself: what is the ethically
correct and professional action in this situation? It is worthwhile cultivating a
network of fellow professionals whom you respect and trust so that you have a
sounding-board to help you make what can be very difficult decisions.
286 Chapter 13 Ethics and Professionalism

Exercises for 13 Ethics and Professionalism

Ex.13.10 Coming back to our BAM decision, now let’s make it more compli-
cated: your boss makes it clear to you that your bonus in this year and
the future is going to be partly, but significantly, linked to the volume
of client assets which are placed with BAM.
Ex.13.11 What if you know that there is a shake-up of the BAM management
team in the offing and that recent poor performance is the result of
poor management decisions, but you have been asked to keep this
information confidential?
Ex.13.12 And suppose that BAM have a strong relationship with a large govern-
ment pension fund to which you would like the consulting appoint-
ment; they view a recommendation to this client as a quid pro quo for
securing the assets of your client.
U Latin phrase alert! Quid pro
quo means “a favour in ex- Ex.13.13 And how would things change if your best friend is Chief Investment
change for another” - ba- Officer of BAM? Let’s say that you’ve just been to the christening of her
sically they’re saying that second child and know that her position is under pressure as a result of
they’ll recommend you for assets under management not having grown sufficiently over the past
the consulting appointment
year.
if you place the assets of your
client with BAM. Ex.13.14 And finally, how would your response change if you knew that the
Financial Services Board (the regulator of financial services in South
Africa) and the President of the Actuarial Society of South Africa were
going to be reading your report to your client?
Solutions 287

Selected Numerical Solutions for Part 3

Chapter 8 Solutions
Ex. 8.5 1. Public Liability
2. Product Liability
3. Professional Liability
4. Product Liability and 3rd Party Liability

Ex. 8.6 Both claims will be recovered

Ex. 8.8 1. R178.57

2. R450
3. R20.24

Ex. 8.9 1. Square meter per month

2. a. R471.24
b. R68 544
c. R2 448

Ex. 8.10 1. R450.39

2. a. R443.32
b. R380.32
c. R161.92

Chapter 9 Solutions
Ex. 9.4 1. 65 year old
2. man
3. 50 year old
4. flat
5. no guarantee

Ex. 9.14 R 3 014 706

Chapter 10 Solutions
Ex. 10.3 33.3% quota share, R500 000 limit

Ex. 10.4 1. Insurer: 500 000; Reinsurer R1 5000 000

2. Insurer: R7.5m; Reinsurer: R22.5m
3. Insurer: R62.5m; Reinsurer: R187.5m

Ex. 10.12 1. R175 000

2. R280 000
3. R800 000
4. R800 000
288 Chapter 13 Ethics and Professionalism

Chapter 11 Solutions
Ex. 11.1 1. (a)
l 1 = 97 551,
(b)
l 3 = 97 302,
(c)
l 99 = 107.52,
2.
l 2 = 97 398,
3.
l 44 = 92 763.

Ex. 11.2
l 42 = 93 570 − 242 = 93 328.
x lx dx
60 78924 1805
61 77119 1947
62 75172 2088
Ex. 11.3
63 73084 2228
64 70856 2366
65 68490 2499
66 65991 2625
Ex. 11.4 1. 730
2. 2 602
3. 22 881
4. 3 512
5. 99 957.984
6. 99 957.984

Ex. 11.5
42 187
p 73 = = 0.92862.
45 430
3 130
q 71 = = 0.06048.
51 755
Ex. 11.6
2 959
q 69 = = 0.05122
57 765
p 69 = 1 − q 69 = 0.94878.

Ex. 11.7 2.

Ex. 11.8 1.
d0
= 0.02449
l0
Solutions 289

2.
d0 + d1
= 0.02602
l0
3.
l 42
= 0.99507
l 40
4.
d 40 + d 41
= 0.00462
l0
Ex. 11.9 Assuming the student is 19 years old, on the basis of A1967–70 (ulti-
mate):
61 p 19 = 0.36702 and 80 p 19 = 0.00348,

and on the basis of ELT 12 (Males):

61 p 19 = 0.23788 and 80 p 19 = 0.00112.

x lx dx qx px
20 9982.2006 5.8096 0.000582 0.999418
Ex. 11.10
21 9976.3910 5.7564 0.000577 0.999423
22 9970.6346 5.7032 0.000572 0.999428
Ex. 11.11
EPV(benefits) = 2 000 000A 65 = R1 174 100.

Ex. 11.13 Endowment assurance:

EPV(benefits) = 100 000A 20:45 = R19 031.

Term assurance:

EPV(benefits) = 100 000A 1 = R5 248.36.

20:45

The endowment assurance will have larger premiums, all else equal.

Ex. 11.14
EPV(benefits) = 10 000A 1 = R149.11.
45:5

Ex. 11.15 1.
EPV(benefits) = 100 000A 62:3 = R86 624.31.
2.
1
EPV(benefits) = 100 000A = R81 374.12.
62:3
3.
EPV(benefits) = 100 000A 1 = R5 250.19.
62:3

Ex. 11.16
EPV(premiums) = 900ä 37 = R17 658.90.
290 Chapter 13 Ethics and Professionalism

Ex. 11.17
EPV(benefits) = 1 000ä 10 = R23 543.
EPV(benefits) = 1 000ä 20 = R22 532.
EPV(benefits) = 1 000ä 35 = R20 069.
EPV(benefits) = 1 000ä 50 = R16 003.
EPV(benefits) = 1 000ä 99 = R2 279.

Ex. 11.20
EPV(benefits) = 100ä 60:5 = R448.94.

Ex. 11.21
EPV(benefits) = 100ä 59:3 5% = R282.23.

Ex. 11.22
EPV(benefits) = 500ä 37:3 8% = R1 389.18.

Ex. 11.23
EPV(benefits) = 10 000(1 − d ä 33:2 6% ) = R8 900.69.

Ex. 11.24
EPV(benefits) = 100ä 57:8 = R670.80.

Ex. 11.25
EPV(benefits) = 100ä 65:5 5% = R432.22.

Ex. 11.26
EPV(benefits) = 500ä 22:4 7.5% = R1 797.46.

Ex. 11.27
EPV(benefits) = 100ä 32:8 = R698.35.

Ex. 11.30 1.
Net premium = R2 429.22.
2.
Gross premium = R3 101.57.

Ex. 11.31 1.
Net premium = R2 031.33.
2.
Gross premium = R2 777.72.

Ex. 11.32
Annual annuity amount = R150 960.15.

Ex. 11.33 1.
Net premium = R1 704.87.
Solutions 291

2.
Gross premium = R1 967.97.
3.
PV10 = R16 766.33.
PV20 = R41 228.19.

Ex. 11.34 1.
Net premium = R773.65.
2.
Gross premium = R933.97.
3.
PV5 = R2 086.03.
PV10 = R3 982.05.
PV15 = R3 988.32.
PV19 = R1 277.00.
4. The first three policy values suggest that the policy value in-
creases with term at a decreasing rate. If we calculate PV19 (not
asked for), we see that it actually has decreased relative to PV15 .

Ex. 11.35
20 p x v 20 = 0.53846.

Ex. 11.36 1.
EPV(benefits) = R5 394.49.
2.
Gross premium = R1 965.93.

Chapter 12 Solutions
Ex. 12.2 Using j = 5%, 13.4% and 25.3% respectively.

Ex. 12.3 80.9%

Ex. 12.4 A 0%: RR = 26.79%

4.5%: RR = 47.75%
7%: RR = 67.77%
B 2%: RR = 34.32%
3%: RR = 39.06%
4%: RR = 44.62%

Ex. 12.5 9.875465

Ex. 12.6 23 800

292 Chapter 13 Ethics and Professionalism

Ex. 12.8 13.989% vs. 13.823%

Ex. 12.9 1.1437 using j = 3.5%

Ex. 12.12 1. 0.818182

2. 1.22
Module 4

Assets

293
Chapter 14

Overview of Assets

14.1 Why study assets?

If you are studying quantitative finance, the analysis and evaluation of various
types of assets is likely to form a large part of your future career. Actuaries are
also involved in the analysis and evaluation of assets. But for most “traditional”
actuaries, a significant part of your future work will have to do with liabilities,
i.e. the (usually uncertain) financial obligations owed by financial institutions to
their members or clients, and their valuation.

Exercises for 14.1 Why study assets?

Ex.14.1 List the types of financial institutions actuaries may work for.
Ex.14.2 For each institution, give some examples of the liabilities that the insti-
tution may have.
Ex.14.3 Can individuals have liabilities too? Give some examples.

All financial institutions have liabilities. This section of the course deals with U You will have come across
the other side of the balance sheet, i.e. the assets which these entities have in- “balance sheets” in your ac-
counting courses. Recall that
vested in so as to meet their liabilities. It is good advice for this section to keep
they compare the liabilities
the liability side of the balance sheet in mind when considering the various asset of a company to its assets.
classes and their properties. Of course, it is also equally good advice when study- When we looked at the fund-
ing actuarial liabilities to give some thought to the appropriate assets. Actuarial ing level of a pension fund,
and quantitative practice is often about solving a joint asset-liability puzzle. for example, we were using
the balance sheet of the fund
to find the values of the as-
14.2 Matching assets to liabilities sets and liabilities.

A common starting point for deciding on an investment strategy is to look to

match assets and liabilities as far as possible.

Matching

295
296 Chapter 14 Overview of Assets

Perfect matching is a situation in which all of the liability cash outflows are exactly
offset by the asset cash inflows, removing all the risk of not being able to meet your
liabilities.

Example 14.1 Say you have a certain commitment to pay out R100 to some-
body in 12 months’ time. This constitutes a liability to you, which you could
perfectly match by purchasing a zero-coupon bond with a capital repayment
of exactly R100 in 12 months’ time.

Example 14.2 Consider a hypothetical financial institution which has a

range of commitments over the next 15 years. The institution invests their
assets so as to perfectly match their liabilities. Figure 14.1 illustrates the
institution’s liability and asset cashflows. We can see that asset cashflows in
any year are sufficient to meet the liability cashflow requirements.

Matching assets and liabilities

1000

800

600

900
400 840 850 840
760
650
590 570
510 510
460
200
280 310 300
200
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
-200
-280 -310 -300
-200
-460
-510 -510
-590 -570
-650
-400 -760
-840 -850 -840
-900

-600

-800

-1000

Assets Liabilities

Figure 14.1 Asset and liability cashflows.

U “Nominal” in this context In practice, actuarial liabilities are uncertain in nature, and so it is impossi-
refers to liabilities that are ble to predict cash outflows with complete accuracy. Short-term, fairly certain
fixed in Rand terms, and nominal liabilities are the easiest to predict, while long-term highly uncertain real
“real” refers to liabilities that liabilities are much harder to estimate, and any estimates are subject to significant
are linked to inflation.
error.

Exercises for 14.2 Matching assets to liabilities

Ex.14.4 For a life insurer, list some examples of liabilities that are:
14.3 The investment market 297

• short-term nominal;
• long-term nominal;
• long-term real.
Are there any short-term real liabilities?

Similarly, the timing and amount of any cashflows from an asset can be very
uncertain, though the degree of uncertainty varies for different asset classes.

Exercises for 14.2 Matching assets to liabilities

Ex.14.5 Which asset classes have predictable cashflows?

Ex.14.6 Which asset classes do you think have uncertain cashflows?
Ex.14.7 Under what circumstances can the cashflows from a bond be uncer-
tain?

Because of the uncertainty in the liability and asset cashflows, perfect match-
ing is generally not possible in practice, but perfect matching remains a useful
benchmark when setting an investment strategy.

14.3 The investment market

The investment market may be divided into the following sub-markets:
• Money market which comprises of short-term investments such as cash,
bank deposits, Treasury bills, bills of exchange, and commercial paper.
• Bond market which comprises of government-issued bonds and those is-
sued by companies, parastatals and local government. Bonds can be fixed
interest, meaning that their coupons and redemption values are fixed in
nominal terms, or index-linked, meaning that their cashflows are linked to
inflation and grow over time.
• Equity market, where shares in companies are traded.
• Property market, also commonly referred to as the real estate market, is a
market where properties are traded.
T The term security is a general
• Derivatives market where instruments such as futures and options, which term for assets that can be
derive their value from that of another underlying security, are traded. traded, like equities, bonds,
money market instruments
The above markets are the main market types that we will focus on. The assets
and also derivatives.
traded in these markets are by no means the only assets that you can buy in an
investment market; for example, commodities are things such as oil, coal, maize,
and coffee, and these are also often traded in an investment market. Another
interesting broad asset class is a class often referred to as alternative investments.
This asset class includes hedge funds, private equity, and private debt.
298 Chapter 14 Overview of Assets

Exercises for 14.3 The investment market

Ex.14.8 Which of the asset classes traded in the above markets do you think
provide real returns? Which provide nominal returns?

14.4 Characteristics of investment markets

14.4.1 Domestic vs. international investment
? Can you think why a pen- Note that for many of these asset classes, investors may consider investing ei-
sion fund might have some
ther domestically or internationally. The primary reasons for choosing to invest
foreign-denominated liabili-
in international bonds or equities, say, would be to match foreign currency-
ties?
denominated liabilities, or to provide diversification, which will be discussed later
in this Module.

14.4.2 Direct vs. indirect investments

Investors may also choose to invest in these assets either directly, or indirectly via
collective investment vehicles such as unit trusts. A collective investment vehicle
is managed by an asset management house. Thousands of different vehicles exist,
where the asset managers select the underlying assets to match a certain “theme”
or objective.

Example 14.3 An asset manager may manage portfolios such as:

• A property portfolio which buys real estate;
• An equity portfolio with an aggressive mandate designed to maximise
returns while taking on fairly high risks;
• Another equity portfolio designed to deliver moderate growth but not
to lose money as much as possible;
• An equity portfolio designed to deliver high income but not a lot of
capital growth;
• A bond portfolio designed to deliver high income and not a lot of capital
growth;
• A money market portfolio designed to give a stable, predictable return
higher than a bank account;
• Any number of balanced portfolios (also called multi-asset) which have
holdings across a number of asset classes.
Investors can invest in these portfolios, and leave the management of the
assets to the asset manager. Indirect investment vehicles provide market
? Which assets may not be ac-
expertise, diversification of investments and access to asset classes which
cessible to an individual in-
vestor? What about a small
may not be accessible to an investor; in return, a fee is paid to the asset
pension fund? manager, usually a percentage of assets under management.
14.5 Income vs. capital gains 299

The decision to invest directly or indirectly mostly depends on the size of the
investor, the unit size of the assets the investor wants to buy, the expertise of the
investor, and the availability of the right collective investment vehicle. Large
pension funds and life insurance companies, for example, would tend to invest ? Which of the four major as-
set classes above has assets
directly, and often employ their own professional fund managers to manage
with (very) large unit sizes?
their investments. This approach is justified for large financial institutions as
their transactions tend to be very large, and investing directly allows them to
ensure that the investment strategy is perfectly suited to their needs. For smaller
investors, however, transaction costs as well as a lack of expertise often makes
direct investment not viable, and these investors often use collective investment
vehicles to invest.

Exercises for 14.4.2 Direct vs. indirect investments

Ex.14.9 For each of the following investors, would direct or indirect investment
be more suitable?
• An individual investor who has little expertise in the stock market.
• An individual with a lot of expertise in property and who wants to
invest R10 000 in property.
• A small pension fund.
• A large pension fund.

14.4.3 Primary vs. secondary markets

Bonds and equities owe their existence to the need of corporate and government
entities to raise capital in order to carry out their planned activities. For example,
a government may wish to build a new road, or a company may wish to expand
its business into a new market. Governments can borrow money by issuing
bonds or money market instruments. Companies can also issue bonds (or borrow ? What are the other ways
for governments to raise
money from banks), but may alternatively look to raise capital by issuing new
money?
shares which entitle the holders to part-ownership of the company.
The primary markets are where newly issued bonds and shares are first T Think of buying a car as
brought to the investor market. Secondary markets are where these securities an analogy: you can buy a
are subsequently traded between investors. The existence of these secondary new car from the dealership
markets makes it possible to buy and sell assets daily, which is essential for a good (that’s the primary market),
investment strategy. or you can buy and sell sec-
ond hand cars online, or in
second hand car dealerships.
14.5 Income vs. capital gains These are examples of the
secondary car markets.
Assets can deliver returns via income and capital gains. Income refers to regular
cashflows paid to the holders of the assets, such as dividends on shares or coupons
on bonds. Capital gains, on the other hand, refer to the difference between the
capital value received when an asset is sold or matures, and the price that was T It is possible for the final
capital value to be less than
the original purchase price,
and in this case there is a
capital loss!
300 Chapter 14 Overview of Assets

paid to obtain the asset originally. Since income is paid regularly, it can be a good
match for liabilities with regular cashflows.

Exercises for 14.5 Income vs. capital gains

Ex.14.10 What life insurance product has regular benefit cashflows and could
be matched using assets with regular income?

In most countries, both income and capital gains are taxed, often at different
rates. This may affect how interested investors are in certain types of assets. In
South Africa, for example, capital gains are taxed at a lower rate than income, in
an effort to promote long-term savings and investment.

14.6 The SYSTEM T framework

A mnemonic which has served generations of actuarial students, and which you
may find useful in thinking about the properties of asset classes now and in your
future studies, is SYSTEM T, standing for the following characteristics:
• Security, i.e. the risk of default — how likely is the issuer of this security to
go under?
U Recall that the more risky the
asset, the higher the i will be • Yield, i.e. the return on an investment — this is the i that we have been using
due to investors demanding a to calculate the price of bonds and equities in Module 2.
risk premium.
• Spread, i.e. the volatility or uncertainty of market values.
U Marketability refers to how
• Term, i.e. short, medium or long.
easy it is to sell an asset in
general, i.e. a marketable • Expenses, i.e. transaction costs. The E may also refer to Exchange rate risk
asset has an active “market” for foreign assets.
where it is traded. Liquid-
ity is a term which refers to
• Marketability and liquidity, i.e. the ease with which the asset can be bought
how easily an asset can be or sold.
sold for cash without a signif- • Tax
icant change to the price, i.e.
a liquid asset is a marketable
asset that can be sold easily Exercises for 14.6 The SYSTEM T framework
without having to reduce the
price significantly to attract Ex.14.11 Which of the four main asset classes (money market, bonds, equities,
buyers. and property) do you think are marketable? Which ones are liquid?
Ex.14.12 Based on what you already know about the four main asset classes,
try and complete a SYSTEM T analysis on each one of them.

Don’t worry if the importance of some these characteristics eludes you for
now. You can refer back to this mnemonic once you have learnt more about the
various asset categories later in this module, and use it as a framework for the
evaluation of asset properties.
Chapter 15

Money Market Instruments

Money market instruments are assets which have a very short term. Most have
maturities ranging from one day to one year. The most common terms are three
months or less. The fact that they mature so soon makes them easily convertible
into cash. In general, money market instruments are regarded as the best match
for very short-term liabilities. U Another name used for
money market instruments
is “cash”.
Example 15.1 An example of a short-term liability for a life insurer are
endowment assurances which are nearing the end of their term, e.g. a few
months before they mature. Another example are expenses that need to be
paid in the near future.
Can you think of short-term liabilities that a pension fund may have? What
about a general insurer?

Direct money markets instruments include:

• bank account deposits, including term certificates of deposit,
• interbank loans (loans between banks),
• commercial paper,
• treasury bills,
• repurchase agreements (repos).
It is also possible to invest in money markets indirectly through money market
collective investments; for example, money market mutual funds allow investors
to invest indirectly in money markets. As money market investments are highly
liquid and have short terms, they are generally close to being risk-free in the sense ? Why does the risk of default
reduce the shorter the term
that there is very little risk of default.
of the asset?
The major participants in the money market are commercial and investment
banks, insurance companies, governments, corporations, government-sponsored
enterprises, money market mutual funds, brokers and dealers, and the Reserve
Bank. Some of these are issuers, some are investors, and others facilitate trading
in the money markets.

301
302 Chapter 15 Money Market Instruments

Exercises for 15 Money Market Instruments

Ex.15.1 For each of the market participants above, describe what role or roles
they might play.

15.1 Cash and bank deposits

Banks are in the business of making loans. In order to make loans, they need
capital. They can borrow capital from different sources: one of them is the
Reserve Bank , and the other is people who are willing to deposit their money
T The South African Reserve with the bank. The SARB lends money to banks at the repo rate, which is the
Bank (SARB) is the central
most important short term interest rate in the economy.
bank of the country.

Repo rate
The repo rate is the interest rate at which the Reserve Bank lends money to private
banks. When a bank experiences a cash shortfall or a need for liquidity on a daily
basis, they can always get a loan from the Reserve Bank. A formal system is in place
to guide the process through which banks borrow from the Reserve Bank and it is
called the repurchase transactions system (repo system). The repo rate is set by
the Monetary Policy Committee, and it has a powerful effect on the economy — a
high repo rate raises the cost of borrowing in the country, for example, making it
more expensive to get loans and so slowing economic growth because it is more
expensive for companies to grow and expand. Figure 15.1 contains a plot of the
? If you had a company which historic repo rate. The repo rate in 2014 is very low at only 5.75%; back in 2008 the
you wanted to expand, when repo rate was much higher at 12%.
would you have been more
likely to take out a bank loan
to do that, now or 2008? Historic Repo Rate
12

10
Repo Rate

2004 2006 2008 2010 2012 2014

Date

Figure 15.1 Historic repo rate.

Banks make loans to individuals and companies at rates higher than the repo
rate. The difference between the rates (called the spread) is one of the ways that
15.1 Cash and bank deposits 303

banks make profits. Another source of profits is that banks pay depositors a rate
lower than the repo rate.
A deposit in a bank, which can be thought of as a loan from the depositor to
the bank, is one of the most common and familiar money market instruments.
Deposits held in banks typically earn interest at a predefined rate, which is likely
to vary with the prevailing repo rate. T Bank deposits are not con-
sidered securities, as they
The various types of accounts where cash can be placed on deposit include:
are not traded.
• Call deposits: where money is available to the investor on demand, i.e.
the depositor has “instant access” to withdraw the capital deposited. Call
deposits can be for as short a term as overnight.
• Notice deposits: where the investor has to give specified notice (e.g. 32 days)
of the intention to withdraw the cash from the account.
• Term deposits (or fixed-term deposits): where no withdrawal of funds is
allowed prior to expiry of the term. Term deposits are typically for terms of
up to a year but can be for longer.

Exercises for 15.1 Cash and bank deposits

Ex.15.2 You probably have a bank account of your own, can you identify which
type of bank account it is?

The interest rate obtained by the depositor on these deposits can be:
U A fixed term deposit will gen-
• fixed for the term of the deposit, erally have a fixed-interest
• variable from day to day, or rate throughout the term
and all the cashflows will
• be fixed for an initial period and then vary after the initial period. be certain and known in ad-
Often, the interest rate also varies with the amount you deposit in the bank vance. At the other extreme,
account. a call account will have a
variable interest rate and the
amount and timing of the
Exercises for 15.1 Cash and bank deposits cashflows will be unknown.
Ex.15.3 What type of account do you think is likely to get higher interest, all
else equal?
• A call deposit or a 60-day notice deposit?
• A 32-day notice deposit or a 3-month fixed term deposit?
• An account with R500 in it or an account with R25 000 in it?

Once the interest rate has been determined, it is then possible to determine
the interest payments on these bank deposits. In most cases, the interest is usually
compounded daily and paid monthly into the account.

Example 15.2 Consider a call deposit account. The depositor can make
deposits and withdrawals when they wish. Figure 15.2 contains two plots.
The first plot (the upper panel) illustrates the cashflows into and out-of the
304 Chapter 15 Money Market Instruments

account over time. The second plot (the lower panel) illustrates the total
accumulated value of the deposit in the account over time.

10 000
Interest Payments
Withdrawls
8 000
Deposits (withdrawls)

6 000

4 000

2 000

-2 000

-4 000

50 000
Account Value
45 000

40 000

35 000

30 000

25 000

20 000

15 000

10 000

5 000

Figure 15.2 Call deposit account example.

Exercises for 15.1 Cash and bank deposits

Ex.15.4 Marcia has a call deposit with the following interest rates:
• for deposits of <R1 000: 1.5% p.a. nominal, convertible daily,
• for deposits ≥R1000 but <R5 000: 2.5% p.a. nominal, convertible
daily, and
• for deposits ≥R5000: 4% p.a. nominal, convertible daily.
Marcia opens an account on 1 June 2013 and deposits R1 000 per month
for 7 months, with the first deposit occurring on 1 June 2013. The fees
on the account are R15 per month, and are deducted on the first day of
every month starting 1 June 2013. Interest is paid into the bank account
on the last day of each month. What is Marcia’s balance at the end of
31 December 2013?
15.2 Money market securities 305

Ex.15.5 PapayaInc is a small business with a bank account. Their starting

balance on 1 January is R50 000. On the 3rd, 10th, 17th and 24th of
January, they deposit R15 000. On the 25th January, they withdraw
R100 000 for wages. What is their closing balance on 31 January if the
account pays interest of 5% nominal, convertible daily, and the interest
is deposited in the bank account on the last day of each month?

15.2 Money market securities

Most money market securities are promissory notes which have terms of 1 to 18
months. They include treasury bills, money market strips, commercial paper, T A promissory note is much
like an IOU: “A promise to
banker’s acceptances, and bearer deposit notes.
pay the bearer RX at a speci-
Promissory notes: fied date”.
• Pay out to the bearer, so they can be traded (unlike a bank account, that will
only pay out to the account holder); ? If you write a promissory
note, are you making or get-
• Are unsecured, which means that they are just a promise, there are no ting a loan?
specific assets backing the promise;
• Pay no regular interest; the only cashflows are the purchase price and the
redemption amount.
These money market instruments are bought and sold at a discount and
mature at face value. The difference between the purchase price and the face
value is the interest that you have earned on the transaction. T The face value (also referred
to as the par value) is the
amount that would be paid
Example 15.3 out at maturity when the
You decide to purchase a treasury bill with a three-month term and a face value of money market security is
R10 000. It is sold at a discount, so it costs you R9 850. At maturity, you will receive redeemed, i.e. it is the re-
R10 000, which means you’ve earned R150 in interest on your investment. demption amount.
What is the effective annual interest rate you earned on this transaction?

Solution: The equation of value is:

3
9 850 × (1 + i ) 12 = 10 000,
and solving for i , we find
i = 6.2319%.

Money market securities are considered to be equivalent to cash. Their main

advantages are:
U “Principal” means the
• fully guaranteed principal by issuer if held to maturity, amount originally invested.
• usually higher rates of return than bank deposits with similar terms,
• they can be traded,
• they have a variety of maturity dates,
• they are highly liquid, and
• are low risk.
306 Chapter 15 Money Market Instruments

Exercises for 15.2 Money market securities

Ex.15.6 Find the correct price for R1 000 face value of each of the following
instruments:
1. 10-day term, valued at 5%;
2. 1-month term, valued at 6%;
3. 6-month term, valued at 4%.
The interest rates stated are effective annual rates.
Can you see why money market instruments have stable values?

15.2.1 Treasury bills

Treasury bills (also called T-bills) are short-term (usually 3-month) debt obliga-
tions of the central government. The investor who purchases a treasury bill is in
U Treasury bills are effec- effect lending money to the government for a short period of time in exchange for
tively very short term, zero-
a guaranteed return.
coupon, government bonds!
The prices of treasury bills are traditionally quoted using a simple discount
? In what circumstances is the rate.
return not guaranteed?
Simple discount rates
120 000
Price
100 000

80 000
Face Value / Redemption Amount In Module 2, we introduced simple interest rates. Remember that an amount RC
60 000 accumulated for n years at a simple interest rate of i p.a. accumulates to
40 000

20 000

-
RC × (1 + i n).
-20 000

-40 000 In Module 1, we introduced d , the effective discount rate, which is defined as
-60 000 i
-80 000
d = 1+i . What does d mean in practice?
0.05
-100 000
If i = 5% p.a., then d = 1.05 = 4.762%. You can then use d to calculate the present
-120 000
value of R100 paid in one year’s time as follows:
Figure 15.3 Cash flows from a
R100(1 − d ) = R100(1 − 0.04762) = R95.24.
Treasury Bill.
Note that this is the same as using R100v, since:
i 1
1−d = 1− = = v;
1+i 1+i
so 1 − d is just another way of expressing v.
So finally, what happens if we use a simple discount rate? The present value of RC
received in n years’ time using an annual simple discount rate d is:

RC × (1 − d n).

Simple interest and discount rates are not used often in the mathematics of finance
because they are not realistic. T-bill returns are quoted as a simple discount rate
because of historical reasons. You will need to be able to calculate the effective
interest rate from the simple discount rate in order to assess the returns earned.
15.2 Money market securities 307

Example 15.4
If a 6-month treasury bill with a face value of R100 is priced using a 7% simple
discount rate, what is the price?

Solution: The price is calculated as

1
100 × (1 − 7% × ) = R96.50.
2
Can you see why we multiplied by 1/2 in this case?

With d as the simple discount rate, the amount P that the market is prepared
to lend for a repayment of R100 after n days (usually n is 91 days) is found as
follows: ³ n ´
P = 100 1 − d .
365
How would you calculate the effective rate of return from the price?

Example 15.5
If a bill with n days to maturity was bought for P and has a repayment of R100, the
effective annual interest rate i can be found as follows. First we write down the
equation of value
n
P (1 + i ) 365 = 100.
From this we can find
µ ¶ 365
100 n
i= − 1.
P
Find the price of a 91-day (3 month) treasury bill with a simple discount rate of 10%,
and a face value of R150, and calculate the annual effective rate of return of this bill.

Solution: The price can be found as follows:

µ ¶
91
P = 150 1 − 0.1 × = R146.26,
365

and using the formula above we can find the annual effective rate of return earned
as
µ ¶ 365
150 91
i= − 1 = 10.657%.
146.26

15.2.2 Commercial Paper

Commercial paper is a short-term promissory note issued by highly rated banks
and some large non-financial corporations. The debt is usually issued at a
discount reflecting the prevailing market interest rates. ? Commercial paper is very
similar to treasury bills - ex-
Commercial paper usually offers the highest interest rates in the money mar-
cept it is issued by a com-
ket. The actual interest rate offered depends on the risk of the issuer defaulting.
pany, not the government!
Since it is not backed by any collateral, only firms with excellent credit ratings
308 Chapter 15 Money Market Instruments

from a recognized rating agency will be able to sell their commercial paper at
a reasonable price. Commercial paper is a security, since it can be traded. A
company that is able to issue commercial paper because of its creditworthiness,
may prefer to issue the paper than to take out a loan from a bank. The paper will
usually be liquid, and as a result may be more cost effective than a bank loan.
? Will commercial paper have Commercial paper is usually issued with a fixed maturity of between one day
a lower or higher issue price
and 9 months, and is normally used to raise financing for accounts receivable
compared to a treasury bill
and to meet short term liabilities.
of the same term and face
value? Which one has a
higher risk of default? Example 15.6 Say, for example, a company has accounts receivable of
R1 million with a credit period of 6 months. The company will not be able
U Accounts receivable are
to liquidate its receivables before 6 months. Suppose that the company is
moneys that a company ex-
pects to receive in the future.
in need of funds. It can issue commercial paper at simple discount of 10%
The company can access with a total face value of R1 million and a 6 month maturity.
this money sooner by issu- The company has strong credit rating and finds buyers easily. In this way
ing commercial paper with a the company is able to liquidate its receivables immediately, and the buyer
face value equal to the value of the commercial paper is able to earn interest of R50 000 over a period of
of the accounts receivable,
6 months. The company uses the accounts receivable to pay the maturity
and with a maturity date
value of the commercial paper in 6 months’ time.
on the same day that the
accounts receivable are due.

? Can you show how the

R50 000 was calculated? 15.2.3 Banker’s acceptances and bearer deposit notes
Banker’s acceptances (BAs) and bearer deposit notes (BDNs) are similar to com-
mercial paper, except they are guaranteed by a third party — a bank. This makes
them more secure than commercial paper because in the event that the issuer
cannot pay, the bank will step in and pay on their behalf.
? Would a BA be more or less The difference between BAs and BDNs is that a BA is issued by the company,
expensive than commercial
but carries the “acceptance” (i.e. the guarantee) of a bank. A BDN is issued by the
paper issued by the same
bank, on behalf of the company.
company?
The most common term for these instruments is 90 days, though the terms
can vary from 30 to 180 days.

15.2.4 Money market strips

Strips are created when the individual interest coupons are separated (“stripped”)
from the maturity value of a government bond. The interest coupons and maturity
value are sold as individual investments called strips. Money market strips have
terms from one month to 18 months. The risk associated with money market
strips is similar to that of treasury bills offered by the same issuer.
T A building society is a fi-
nancial institution which
is owned by its members and 15.2.5 Negotiable certificate of deposit
operates as a mutual orga-
A negotiable certificate of deposit (NCD) is a short-term security issued by banks
nization. Building societies
offer banking and related
and building societies showing that a stated sum of money has been deposited for
financial services, in particu-
lar mortgage lending.
15.2 Money market securities 309

Sells a security

Pays R100

A 2 weeks later B
“REPO” “REVERSE
Buys back the same security REPO”

Pays R102

NET EFFECT:

A Pays R2 for borrowing R100 for 2 weeks B

Figure 15.4 Example of a repo agreement.

a specified term at a specified rate of interest. The terms to maturity are usually
in the range of 28 days to 6 months, and the interest is payable on maturity.
NCDs are like fixed term bank deposits, but the difference is that they can be
traded in a secondary market. They usually offer a rate of return slightly higher
than banker’s acceptances which makes them extremely popular instruments.

15.2.6 Repurchase rate agreements

A repurchase rate (repo) agreement is the sale of a security (like a bond or equity) T A repo agreement transac-
with a commitment to repurchase the same security at a specified price and on a tion for one party would
mean a reverse repo for the
specified date in the future. A reverse repo agreement involves the purchase of a
second party! See Figure 15.4
security with a commitment to sell it back at predetermined price and date in the for an illustration of a repo
future. When the seller of the security buys back the security at the maturity date, agreement.
they pay more than they originally received for the security at the outset. This
extra amount is the interest payment for the loan of the money, and it is called U This is not the same repo
the “repo rate”. Figure 15.4 illustrates a repo agreement. rate as that charged by the
Repos are a large, important, and complex segment of the money market. reserve bank, although the
names are the same because
Repos offer competitive interest rates for borrowing and lending on a short-term
the transactions underlying
basis. Their terms are usually no more than two weeks and are often overnight.
what the reserve bank does
Repo agreements offer security to the lender of the money because the lender are similar to this repurchase
gets to hold the asset while someone else is using their money. If the borrower of agreement process.
the money defaults at maturity, the lender keeps the asset they have bought and
can recover their money by selling it.
310 Chapter 15 Money Market Instruments

15.3 Money market summary

The characteristics of money market investments can be summarised using the
SYSTEM T framework as follows:
S Security:
This depends on the issuer. For example, money lent to the US Government
is more secure than a deposit with a small company in a developing country.
In most cases, the security will be very good due to the short-term nature of
the instruments, and default is very rare.
Y Yield:
The income from money market investments will be related to the prevailing
short-term interest rate (the repo rate) set by the monetary authorities. This
can vary over time. In general, money market instruments tend to give
a positive real return slightly in excess of inflation, but there are periods
when money market underperforms inflation. The actual level of return
will depend on the type of instrument: who is the issuer, how liquid is the
instrument, and is it secured or guaranteed or not. The highest money
market returns are for instruments that have some risk of default, are not
very liquid and are unsecured.
Compared to other assets, however, money market investments are generally
close to being risk-free, in the sense that there is very little risk of default.
The expected returns from money market investments are therefore lower
than from almost every other type of investment. This does not mean that
cash (i.e. money market) investments will give a lower return than other
investments; for example, money market investments may outperform stock
markets when the stock markets are falling.
S Spread:
The nominal values of money market investments are fixed in cash terms
and, as they are short-term, there is very little volatility in their market values.
For example, if three-month interest rates change by 1% p.a., the value of a
91-day bill would change by only about 0.25%. For call (i.e. instant access or
overnight) deposits with a bank, there is no volatility.
T Term:
The distinctive feature of money market investments is that they are short-
term. The terms are generally less than one year and are often very much
shorter, e.g. one week or even one day.
E Expenses:
The expenses of dealing in and managing money market instruments are
minimal.
E Exchange rate:
Money market investments are available in a wide range of currencies. The
main additional consideration with overseas money market instruments is
the possible fluctuations in exchange rates. Movements in exchange rates
15.3 Money market summary 311

are expected to compensate for investors for differences in interest rates

between countries over the term of the investments. However, currency
movements are very difficult to predict, so there would be a substantial risk
of getting less than the expected return.
M Marketability:
With the exception of call and term deposits, most money market instru-
ments are highly marketable.
T Tax:
The total return from money market instruments is normally treated as
income for tax purposes.
312 Chapter 15 Money Market Instruments

Exercises

Exercises for 15 Money Market Instruments

Ex.15.7 Discuss the reasons why an institution may hold a large portion of their
funds in money market investments.
Ex.15.8 Determine the price of a 91-day (3 month) treasury bill with an annual
simple discount rate of 5% and face value of R250 000.
Ex.15.9 Compute the price of a Treasury bill with 60 days outstanding and a
face value of R10 000, if the current discount rate is 20% p.a simple.
Chapter 16

Bonds

Bonds are loans by investors to the bond issuer in exchange for a given schedule of
repayments, which consist of interest payments (coupons) and capital payments
(redemption). Bonds are commonly regarded as the best match for medium- to
long-term liabilities which are reasonably predictable, e.g. the payments on a life
annuity benefit with fixed escalations.

How can a loan be an asset?

Students are often confused about how a loan can be considered an investment
as this is not quite as intuitive a concept as it is for cash, shares, or property. The
answer is that from the perspective of the lender, the loan provides an entitlement
to future cashflows, which is an asset.
The usefulness of bonds as investments is further enhanced by them being actively
traded, i.e. they are actively bought and sold in a market by investors. Take the
example of Investor A, who buys a newly issued government bond (in the primary
market). A is in effect lending a sum of money to the government in exchange for
a series of fixed repayments. Investor A has the choice of holding on to the bond
until maturity, or alternatively selling it to Investor B (in the secondary market).
Note that the bond entitles the holder to the series of repayments on the original
loan. Once sold, Investor A no longer has anything to do with the contract, and
Investor B inherits all rights to the income stream and the capital repayment at
maturity.

16.1 Bond Markets

The most important sub-markets in a bond market are:
• the domestic government bond market,
• the domestic corporate bond market, and
• the market in overseas government and corporate bonds.
We discuss each of these in turn.
T Until recently, most govern-
ment bonds were seen as
313 risk-free, but the 2008 finan-
cial crisis has cast doubt on
this assumption for a large
number of governments.
314 Chapter 16 Bonds

16.1.1 Domestic government bonds

Government bonds are generally treated as risk-free. This is because governments
can raise money fairly easily to repay their bond obligations. They can do this
either by issuing new bonds, raising tax revenue, or by printing money. Govern-
ments are aware that defaulting on bond commitments would lead to immediate
loss of confidence in the economy, with potentially devastating effects, and so a
government defaulting is quite rare.
T The City of Johannesburg The central government is not the only government entity empowered to
has issued 7 municipal
issue bonds to raise funds. Lower levels of government, e.g. provincial and mu-
bonds since 2004. Read
more at [Link] nicipal governments, may also issue bonds and so may parastatals. The
gk1zUS. fundamental structure of all government bonds is identical; the main distinc-
tion lies in the credit risk associated with issuing entity, and hence the yield that
T A parastatal is a company investors expect on the bonds.
fully or partially owned by
the government. One exam-
ple is Eskom. Can you think
16.1.2 Corporate bonds
of any others? For a corporate, on the other hand, there is a risk that the issuing corporate will
not have enough money to meet repayments as they fall due. This is because
? What would have a higher
a corporate has to fund the repayments from the revenues they earn, and they
yield: a government or a
municipal bond in the same cannot print money or raise taxes. A corporate can also become bankrupt. Thus
country? there is a risk of default with bonds not issued by a government.
Corporate bonds are also often less marketable than government bonds. This
lower marketability is due to the corporate bond issues being much smaller than
government bond issues.
Investors require compensation for bearing these risks, and this results in
higher yields on corporate bonds compared to those on government bonds.
? Do you think government Corporate bonds may either be debentures, which are secured against assets
bonds are secured or unse-
of the company, or unsecured loan stock. If a bond is secured against assets
cured?
of the company and the company defaults on their bond obligations, the bond
holders have a right to the assets and can sell them to recover some of their
investment value. Unsecured loan stock is riskier than secured loan stock and
would command a higher yield from investors.

16.1.3 Overseas bonds

Foreign governments and companies also issue bonds, and some are issued
specifically for investors from other countries to buy. They may be denominated
in the currency of the issuer, or in another currency if required.
Overseas bonds can be bought to match liabilities in overseas countries;
for example, policies sold in those other countries. In that case, the investor
would look for overseas bonds denominated in the correct overseas currency to
minimise the exchange rate risk.
16.2 Types of bonds 315

Example 16.1 A South African insurer has begun selling life insurance poli-
cies in Mauritius. They want to invest in Mauritian bonds, denominated
in Mauritian Rupees, to back those policies. The bonds will pay out fixed
amounts in Mauritian Rupees, and the policies will also pay benefits in Mau-
ritian Rupees, so the liabilities’ currency is matched by the assets currency.
This removes the exchange rate risk from the insurer.

An investor who wants exposure to the bonds of an issuer in another country,

but does not want currency risk or thinks that the foreign currency may lose value
over time, can buy overseas bonds denominated in another currency.

Example 16.2 An asset manager in the US wishes to add some exposure

to South African corporate bonds to her portfolio, but does not want to be
exposed to the Rand (which she feels is too volatile). She buys South African
corporate bonds denominated in US dollars. The SA corporate will pay out
the coupons and redemption values of these bonds in US dollars, regardless
of the dollar-rand exchange rate. In this way, the investor has exposure to
South African corporate bonds and not the Rand.

16.2 Types of bonds

We shall consider three types of bonds:
• zero-coupon bonds (ZCBs),
• fixed-interest coupon paying bonds, and
• inflation-linked bonds.

16.2.1 Zero-coupon bonds

ZCBs are bonds that do not pay interest (coupons) during the term of the bond.
Instead, investors purchase ZCBs at a discount to their nominal value (also called
the face value), which is the amount a bond will pay out at maturity. Figure 16.1
illustrates the cashflows associated with an investment in a ZCB. When a ZCB
matures, the investor will receive the face value of the bond, which can be seen as
the return of the original investment plus the interest earned. ZCBs are available
in a wide range of terms, including some long-term issues. These long-term T A ZCB is basically a long-
term treasury bill.
maturity dates allow an investor to plan for long-term goals, such as funding
for long-term annuity payments. Furthermore, given that there is only a single
cashflow, ZCBs provide a large degree of flexibility when attempting to match
liabilities.

Example 16.3 An investor needs to make the following future payments:

• R100 000 in 3 years,
316 Chapter 16 Bonds

• R200 000 in 5 years, and

• R50 000 in 6 years.
She can purchase ZCBs with three different maturities now to match these
liabilities perfectly. She can buy 3-year ZCBs with a total face value of
R100 000, 5-year ZCBs with a total face value of R200 000, and 6-year ZCBs
with a total face value of R50 000. This will ensure she can meet her liabilities
exactly when they fall due. In this case the investor is perfectly matched.

T Can you explain why a Because ZCBs pay no interest until maturity, their prices are more sensitive to
ZCB is more sensitive to a interest rate changes than other types of bonds, i.e. ZCBs’ prices fluctuate more
change in interest rates than
than other types of bonds when the interest rate changes.
a coupon-paying bond of
the same term?

Redemption amount
Issue date
Maturity
date
Initial price

Figure 16.1 Cashflows from a ZCB.

The price of a ZCB can be found by discounting the cashflows received from
the ZCB by the market yield, the implied interest rate that the market is using to
price the bonds. Consider the following example.

Example 16.4
Calculate the price of a ZCB with a term to maturity of 5 years and a nominal value
of R1 000 000. Assume the interest rate is 7.5% per annum effective.

Solution:

Price = 1 000 000 × (1.075)−5 = R696 558.63.

Example 16.5
A ZCB with a face value of R500 000 and a term to maturity of 10 years is currently
" The gross redemption yield
is the return that you would
expect to get on a bond if you
held it until redemption.
16.2 Types of bonds 317

being sold in the market for R192 771.64. Determine the gross redemption yield
implied by the price of the bond, i.e. calculate the annual effective interest rate
currently being used to price the bond.

Solution:
We can write down the equation of value:
R192 771.64 = 500 000 × (1 + i )−10 ,
and then we can find i as follows
192 771.64 −1/10
µ ¶
i= − 1 = 10%.
500 000

16.2.2 Fixed-interest coupon paying bonds

With fixed-interest (also called “conventional”) coupon paying bonds, the in-
vestor effectively lends the issuer of a bond a sum of money for a fixed period in
return for regular (usually half-yearly) payments of interest (referred to as coupon
payments) over the term of the loan and a lump sum payment (known as the
redemption value) at the end of the term. It is important to remember that all of
these cashflows are known at the outset, so it is possible to predict the nominal ? Under what circumstances
will the cashflows from the
cashflows emerging from the bond with close to complete accuracy.
bond not be predictable?
Figure 16.2 depicts the cashflows from a coupon-paying bond.
Redemption Amount

Issue
Date Coupon Coupon Coupon Coupon Coupon Coupon Coupon Coupon

Maturity
Date

Initial
Price

Figure 16.2 Cashflows from a coupon-paying bond.

When considering government bonds, we generally assume that they are risk-
free, and so the risk-free rate of interest can be used to value them. Corporate
bonds do have a risk of default, and one way to account for this when pricing ? Can you think of any alterna-
these bonds is to increase the interest rate by a risk premium reflecting the risk of tive methods of handling the
additional credit risk?
a particular issuer.
318 Chapter 16 Bonds

Pricing fixed-interest bonds

Bonds are traditionally valued on a discounted cash flow basis, i.e. by discounting
the (known) future coupon and redemption payments to the present date at a
rate of interest reflecting market yields (which should reflect investors’ required
returns). Bond yields are customarily expressed as nominal yields compounded
" If an interest rate is said half-yearly, so that a bond
to be “compounded semi- ³ ¡ yield ¢of 10%
´ per annum is equivalent to an effective
0.1 2
annually” that is the same as annual yield of 10.25% = 1 + 2 − 1 per annum.
saying it is a “nominal rate You have covered the pricing of bonds in Module 1. In this section, we will
convertible semi-annually”. work with bonds which pay interest half yearly. Recall that for a bond with:
• a nominal value of N ;
U Recall that a nominal rate of
interest of j p.a., convertible • term to maturity of n years;
p times p.a., just means an
• and an annual nominal coupon rate D (compounded semi-annually) with
effective rate of j /p per 1/p
of a year. The link between coupons paid semi-annually;
nominal rate convertible the coupon amount paid every 6 months will be
pthly p.a. i (p) and the effec-
tive rate p.a. i is: D ×N
C=
¶p 2
i (p)
µ
1+i = 1+ .
p and there will be a total of 2n coupon payments.
The final redemption value paid by the bond will be determined by the re-
Equivalently:
demption rate R and the nominal value N . Specifically, the investor will receive
i (p) = p (1 + i )1/p − 1 .
¡ ¢ R ×N at redemption. The bond is said to be redeemed at par if R = 1, at a premium
to par if R > 1, and at a discount to par if R < 1.
If the current nominal interest rate is 2i per annum compounded semi-
annually (i.e. an effective rate of i per 6 months), then the price of the bond
is given by:

2n
C (1 + i )−t R × N × (1 + i )−2n
X
Price = +
t =1
| {z }
| {z } pv of the redemption payment
pv of the coupon payments
2n
v it + R × N × v i2n
X
= C
t =1
= C a 2n i + R × N × v i2n .
16.2 Types of bonds 319

Example 16.6
Calculate the price of a bond which pays a coupon of 10% per annum semi-annually
(in arrear) and which is redeemable at par (i.e. R = 1) for a nominal amount of
R100 000 after 5 years. You may assume that the market expects a yield to maturity
of 8% per annum compounded semi-annually.

Solution:
• Nominal value: N = 100 000,
• Redemption value: R × N = 100 000
• Coupon rate: D = 10% p.a., i.e. semi-annually = 5% per half year,
• Coupon amount: D×N 2 =
100 000×0.1
2 = 5 000,
• n = 5 years, i.e., 10 coupon payments in total,
• yield to maturity = 8% compounded semi-annually = 4% per half year.

10
Price = 5 000a 10 4% + 100 000v 4%
1 − (1.04)−10
µ ¶
= 5 000 + 100 000 × (1.04)−10
0.04
= 40 554.48 + 67 556.42
= R108 110.90.

What do we mean by the market expecting a particular yield?

Investors make investment decisions based on the prices of assets and the cash-
flows that these assets offer. If an asset seems overpriced compared to the cashflows
it offers and the risk that the market assigns to it, investors won’t buy it and the
price will decrease. In the same way as the forces of supply and demand will in-
teract to produce an equilibrium in the market for any good, as your studies of
microeconomics will have taught you, so too will supply and demand for bonds
interact to produce an equilibrium price. This in turn implies an equilibrium yield,
which we can take to reflect the required (and expected) return of “the market” as a
whole.

Relationship between bond prices and bond yields

One key characteristic of bond price equation,

Price = C a 2n + R × N × v 2n ,

which you should take particular note of is the inverse relationship between the
price and the yield (i.e. the interest rate used to price the bond). If i increases,
then the price will decrease, and if i decreases, then the price will increase. As
the yield is the effective discount rate in the above pricing formula, the higher the
yield, the lower the present value of future cash flows and hence the lower the
bond price.
320 Chapter 16 Bonds

300.00
i Bond Price
1% 262.41
2% 230.81
250.00
3% 204.14
4% 181.54
5% 162.31
200.00 6% 145.88
7% 131.78
8% 119.64
Bond Price

150.00 9% 109.13
10% 100.00
11% 92.04
100.00 12% 85.06
13% 78.93
14% 73.51
15% 68.70
50.00
16% 64.43
17% 60.61
18% 57.18
0.00
0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20% 19% 54.09
Semi-annual effective interest rate i 20% 51.30

Figure 16.3 The price of a coupon paying bond with 10 years to maturity, coupon rate
of 20% per annum paid semi-annually, nominal value of R100 and varying yield to
maturity.

Put a different way, we know that the future cashflows from a bond are fixed,
therefore the only way that an investor can earn a higher yield is by paying a lower
price for the bond.
The relationship between bond prices and yields is illustrated by an example
in Figure 16.3.

Exercises for 16.2.2 Relationship between bond prices and bond yields

Ex.16.1 Consider a 5 year bond which pays a coupon of R5 every 6 months, and
has a redemption value of R100. Calculate the price of the bond if the
interest rate is:
• 4% effective per 6 months,
• 5% effective per 6 months, and
• 6% effective per 6 months.
Compare the results to the redemption value of the bond. What do you
notice?

Relationship between coupon rates and bond prices

? Can you explain why some- If the same yield applies, and two bonds have different coupon rates but are
one will pay more for a bond otherwise identical (i.e. same issuer, redemption amount and term), the price of
that pays a higher coupon
than a bond that pays a
lower coupon, all else equal?
16.2 Types of bonds 321

the bond with the higher coupon rate will be higher than the price of the bond
with the lower coupon rate.

Exercises for 16.2.2 Relationship between coupon rates and bond prices

Ex.16.2 Consider a 10 year bond which pays a coupon every 6 months, and has
a nominal value of R100. The bond is redeemed at par. Calculate the
price of the bond if the coupon rate is:
• 4% p.a. convertible half-yearly,
• 7% p.a. convertible half-yearly, and
• 10% p.a. convertible half-yearly.
Assume a yield of 7% p.a. convertible half-yearly.
Compare the results to the redemption value of the bond. What do you
notice?

Generally, for a bond that is redeemed at par, pays coupons of C every 6

months, and valued at a semi-annual effective interest rate i :
• when C = i :
the bond price will be equal to its nominal value, i.e. the bond will be trading
at par,
• when C < i :
the bond price will be below its nominal value, i.e. the bond will be trading
at a discount to par,
• and when C > i :
the bond price will be above its nominal value, i.e. the bond will be trading
at a premium to par.
The relationship between the bond price and the coupon rate is illustrated in
Figure 16.4.
322 Chapter 16 Bonds

160 Semi-Annual
Coupon C Bond Price
5 57.43
5.5 61.69
140
6 65.95
6.5 70.20
7 74.46
120
7.5 78.72
8 82.97

Bond Price
8.5 87.23
100
9 91.49
9.5 95.74
10 100.00
80
10.5 104.26
11 108.51
11.5 112.77
60
12 117.03
12.5 121.28
13 125.54
40
13.5 129.80
8 12 16 20 24 28
Annual Coupon Amount 2C 14 134.05
14.5 138.31

Figure 16.4 The price of a coupon paying bond with 10 years to

maturity, yield to maturity of 20% per annum compounded semi-
annually, nominal value of R100, and varying coupon rate paid semi-
annually.
Bonds that are trading below par, i.e. Price < N , are referred to as discount
bonds; bonds trading at par, i.e. Price = N , are referred to as par bonds; and bonds
trading above par, i.e. Price > N , are referred to as premium bonds.
The next example demonstrates that the above relationships do not always
hold for bonds not redeemed at par.

Example 16.7
(a) A 10 year bond issued by the government pays a coupon of 10% p.a. (semi-
annually in arrear) and is redeemable at a premium of 40% (i.e. R = 1.4).
Calculate the price of the bond per R100 nominal value, assuming a yield to
maturity of 12% p.a. compounded semi-annually.
(b) Explain by general reasoning why the price of this bond is above par value.

Solution:
(a) From the question we know that:
• par value N = 100,
• redemption amount R × N = 1.4 × 100 = 140,
• coupon rate = D = 10% p.a., i.e. semi-annually = 5% per half year.
• n = 10 years, i.e. 20 coupon payments, and
• yield to maturity = 12% compounded semi-annually = 6% per half year.
16.2 Types of bonds 323

20
Price = 5a 20 6% + 140v 6%
1 − (1.06)−20
µ ¶
= 5 + 140(1.06)−20
0.06
= 57.35 + 43.65
= 101.00

(b) Since the yield to maturity is greater than the coupon, the bond would trade
at a discount if there were no redemption premium. However, this is coun-
teracted by the effect of the redemption premium which will increase the
price. Since the premium is 40% ≈ 4% per annum, whereas the excess yield to
maturity (over the coupon rate) is approximately 2% per annum, the net effect
is for the bond to trade at a premium to par.

The longer the term to maturity, the more pronounced is the discount or ? Can you explain why this is
premium at which the bond trades. Figure 16.5 illustrates the effect of different the case?
terms to maturity for discount, par, and premium bonds.
160
Discount Bond
Par Bond
Premium Bond
140

120
Bond Price

100

80

60

40
0 2 4 6 8 10 12 14 16 18 20
Term to Maturity n

Figure 16.5 The price of discount, par, and premium bonds with
different terms to maturity.

Exercises for 16.2.2 Relationship between coupon rates and bond prices

Ex.16.3 What happens to the price at 4%, 5% and 6% interest in Exercise 16.1,
if you change the term to:
• 2 years, and
• 10 years.
What is the effect of term to maturity on the difference between the
prices at the different interest rates?
324 Chapter 16 Bonds

16.2.3 Inflation-linked bonds

One of the risks of buying a fixed-interest bond is that the cashflows are fixed as
certain rand amounts, i.e. they are fixed in nominal terms.

Example 16.8
Suppose that you are holding R100 000 nominal of a 20 year bond which pays
annual coupons at a rate of 8% p.a.
(a) What is the nominal value of each coupon payment?
(b) Find the value in today’s terms of the 16th coupon payment if
i. inflation is 5% per annum,
ii. inflation is 10% per annum.
(c) If you are relying on the payments from this bond to pay for living expenses,
which of the above two scenarios makes you worse off?
(d) If you are using the payments from this bond to pay premiums for a life
insurance policy with fixed premiums, which of the two scenarios will make
you worse off?

Solution:
(a) The nominal coupon payment is

8% × R100 000 = R8 000.

(b) i. The value of the 16th coupon payment in today’s terms is 8 000×(1.05)−16 =
R3 664.89.
ii. The value of the 16th coupon payment in today’s terms is 8 000(1.10)−16 =
R1 741.03.
(c) You would be worse off if inflation is 10% p.a., but in both scenarios, you would
have less available for day to day expenses than you would today.
(d) If you are paying a fixed life insurance premium, your expense will not go
up with inflation, and the amount from the coupon will always be enough to
cover the premium. So it does not matter what inflation is.
16.2 Types of bonds 325

CPI
110

100

90

80

70

60

50
Jul-02

Jul-03

Jul-04

Jul-05

Jul-06

Jul-07

Jul-08
May-02

Nov-02

May-03

Nov-03

May-04

Nov-04

May-05

Nov-05

May-06

Nov-06

May-07

Nov-07

May-08

Nov-08
Mar-02

Mar-03

Mar-05

Mar-07

Mar-08
Jan-02

Sep-02

Jan-03

Sep-03

Mar-04
Jan-04

Sep-04

Jan-05

Sep-05

Mar-06
Jan-06

Sep-06

Jan-07

Sep-07

Jan-08

Sep-08

Figure 16.6 CPI for 2002–2008. The base index value is 100 in mid 2008.

Inflation
Inflation is a rise in the general level of prices of goods and services in an economy
over a period of time. In South Africa, the rate of inflation is often determined based
on the Consumer Price Index, or CPI for short. The South African CPI shows the
change in prices of a standard package of goods and services which South African
households purchase for consumption. Figure 16.6 contains a plot of the historical
CPI values.
There are additional indices calculated by Statistics South Africa; for example, there
is an index for the producers of goods called the Producer Price Index. You can find
out more here: [Link]

Inflation-linked bonds were created to reduce inflation risk. Their cashflows

increase with a specified index, for example the Consumer Price Index (CPI).
Figure 16.7 illustrates the cashflows from an index linked bond. You can see that
all of the cashflows have been increased for inflation — the purple additions to
the coupons and the black addition to the redemption value.
326 Chapter 16 Bonds

Redemption

Adjustment
Amount
Inflation
Redemption
Unadjusted

Amount
Coupon Coupon
Coupon
Inflation Inflation
Inflation
Adjustment Adjustment
Issue Adjustment
Unadjusted Unadjusted Unadjusted Unadjusted Unadjusted
Date Coupon
Coupon Coupon Coupon Coupon
Maturity
Date

Initial
Price

Figure 16.7 Cashflows from an inflation-linked bond.

In the South African market, inflation-linked bonds are issued mainly by the
government.

Example 16.9
An inflation-linked bond is issued on 1 January 2013 and is redeemable at par in 15
years. A coupon of 4% per six month period is payable every 6 months. The coupon
and redemption payments are indexed using the CPI.
What are the cash amounts of the first two coupons for R10 000 nominal of the
bond? Use the table of CPI values below.

Date CPI
Jan 2013 101.5
Jul 2013 102.4
Jan 2014 103.9
Jul 2014 105.1
Jan 2015 106.2
Jul 2015 107.1

Solution:
The unadjusted coupon amount is

4% × 10 000 = R400.

The first coupon amount paid is

102.4
R400 × = R403.55.
101.5
16.2 Types of bonds 327

The second coupon amount paid is

103.9
R400 × = R409.46.
101.5

In general, for an inflation-linked bond the actual cashflow paid out at time t is
calculated as
Q(t )
unadjusted amount × ,
Q(0)
where Q(t ) is the index value at time t , and Q(0) is the index value at the issue date
of the bond.

In general we will not know the index values at future dates and we will have to " In practice it takes time for
an index to be calculated,
assume an inflation rate for the future periods. Let’s consider the above example
and we will not have the
again using an assumed inflation rate. current index value available
to calculate the cash paid
Example 16.10 out. As a result, there will be
a lag in the indices used to
An inflation-linked bond is issued on 1 January 2013 and is redeemable at par in 15 calculate the actual cashflow
years. A coupon of 4% per six month period is payable every 6 months. The coupon paid out, i.e.
and redemption payments are indexed using the CPI.
What are the cash amounts of the first two coupons for R10 000 nominal of the Q(t − L)
unadjusted amount× ,
bond? Inflation, as measured by CPI, is expected to be 3% p.a. for the next 15 years. Q(0 − L)

where L is the lag period.

Solution: The unadjusted coupon amount is
We will ignore this to keep
4% × 10 000 = R400. it simple. You will see much
more of this in the second
The first coupon amount paid is year course.

R400 × 1.031/2 = R405.96.

The second coupon amount paid is

R400 × 1.03 = R412.00.

In general when assuming a fixed level inflation rate, the actual cashflow paid out
at time t is calculated as

unadjusted amount × (1 + i )t .

Note that the future inflation rate does not have to be assumed to be level.

Exercises for 16.2.3 Inflation-linked bonds

Ex.16.4 Calculate the expected redemption payment received at time 6 from

R10 000 nominal of a 6-year inflation-linked bond with an annual
coupon rate of 10%, payable semi-annually. The bond is redeemed at
par. The inflation rate is expected to be 5% p.a. for the next 3 years, and
7% p.a. thereafter.
328 Chapter 16 Bonds

Pricing inflation-linked bonds

Lets consider an inflation-linked bond with the following terms of issue:

• a nominal value of N ,
• term to maturity of n years,
• a redemption rate of R,
• and an annual coupon rate D (compounded semi-annually) with coupons
paid semi-annually, which means the unadjusted coupon amount paid
every 6 months is C = D×N
2 .
Assume that the market anticipates future inflation at a rate of j % p.a. and
the current nominal interest rate is 2i per annum compounded semi-annually.
The price of the bond is given by:

2n
C (1 + j )t /2 × (1 + i )−t + R × N × (1 + j )n × (1 + i )−2n
X
(note that we have adjusted each Price =
payment using the assumed t =1
2n ¶−t ¶−2n
1+i 1+i
µ µ
inflation rate to find the actual X
= C +R ×N ×
cashflow paid) t =1 (1 + j )1/2 (1 + j )1/2
(1 + i )
(note that i 0 is now the now let i 0 = 1
−1
inflation-adjusted semi-annual (1 + j ) 2
2n
interest rate)
C (1 + i 0 )−t + R × N × (1 + i 0 )−2n
X
=
t =1
2n
v it0 + R × N v i2n
X
= C 0
t =1
= C a 2n i 0 + R × N × v i2n
0 .

All we have done is increase the cashflows for inflation and then discounted them
to find the present value.

Example 16.11
Calculate the price of an inflation-linked bond which pays a coupon of 10% per
annum, semi-annually in arrear, and which is redeemable at par for a nominal
amount of R100 000 after 5 years. You may assume a yield to maturity of 8% p.a.
compounded semi-annually, and an annual rate of inflation of 5% effective.

Solution: From the question we know that:

• par value N = 100 000,
• coupon rate = D = 10% p.a., i.e. semi-annually = 5% per half year, and an
unadjusted coupon amount of C = 5% × 100 000 = 5 000,
• n = 5 years, i.e. 10 coupon payments,
• yield to maturity = 8% compounded semi-annually = 4% per half year,
• redemption rate R = 1, and
16.3 Credit ratings 329

• j = 5%.
The price is then found as follows:
10
(1.05)t /2 (1.04)−t + 100 000(1.05)5 (1.04)−10
X
Price = 5 000
t =1
(1.04)
let i 0 = 1
− 1 = 0.01494
(1.05) 2
⇒ Price = 5 000a 10 i 0 + 100 000v i10
0
µ −10 ¶
1 − (1.01494)
= 50 00 + 100 000(1.01494)−10
0.01494
= 46 126.54 + 86 221.01
= R132 347.55.

Exercises for 16.2.3 Pricing inflation-linked bonds

Ex.16.5 Calculate the price of an inflation-linked bond which pays a coupon of

8% per annum, semi-annually in arrear, and which is redeemable at
par for a nominal amount of R150 000 after 10 years. You may assume
a yield to maturity of 7.5% p.a. compounded semi-annually, and an
annual rate of inflation of 4.5% effective.

16.3 Credit ratings

Bond issuers have found that they are more successfully able to raise funds if they
maintain credit ratings with a recognised ratings agency. A rating agency is an
independent body that measures the risk of default for different bond issuers.
Currently there are three major global ratings agencies: Fitch Ratings, Moody’s
Investors Services and Standard & Poor’s.
A credit rating gives an indication of the creditworthiness of a bond. The
credit rating process involves a thorough investigation of the financial condition
of the entity issuing the bond in order to assess the likelihood that the entity will
fail to the meet its future obligations.
A typical categorisation system, such as that used by Standard & Poor’s, assigns
a grade of AAA to the most creditworthy bonds, which are deemed to have almost
no chance of defaulting. AA is then the next best rating, then A, then BBB, then
BB and so on. The worse a company’s credit rating, the higher the yield will be on
the company’s bonds to compensate the investors for the additional credit risk.

Credit rating agencies and the 2008 financial crisis

The 2008 global financial crisis was partly attributed to rating agencies. The crisis
took place when many bond issues plummeted in value and millions of investors
lost large amounts of money. Credit rating agencies came under serious criticism
as the rating of those bonds was often very high, and critics blamed agencies for
330 Chapter 16 Bonds

giving these bonds high credit ratings and thus encouraging investors to buy these
assets at high prices. Part of the problem was that the bonds that were being issued
were new and very complex instruments, and that credit agencies did not assess
the risk of these instruments correctly. Read more at [Link]
Exercises 331

Exercises
Below are some further pricing examples. You may also want to revise the section
on bond pricing in Module 1.

Exercises for 16 Bonds

Ex.16.6 Calculate the price of a ZCB with a term to maturity of 15 years and a
face value (nominal value) of R100 000. Assume the following interest
rates:
• 4.5% per annum effective,
• 6.5% per annum effective,
• 8.5% per annum effective and
• 10.5% per annum effective.
Comment on your results.
Ex.16.7 A ZCB with a face value of R1 000 000 and a term to maturity of 5 years
is currently being sold in the market for R747 258.17. Determine the
gross redemption yield implied by the price of the bond, i.e. calculate
the annual effective interest rate.
Ex.16.8 Calculate the price of a bond which pays a coupon of 6% per annum,
semi-annually in arrear, and which is redeemable at par for a nominal
amount of R1 000 000 after 25 years. You may assume a yield to maturity
of 11.5% per annum compounded semi-annually.
Ex.16.9 A 30 year bond issued by the government pays a coupon of 8% p.a.
(semi-annually in arrear) and is redeemable at a premium of 15%, i.e.
R = 1.15. Calculate the price of the bond per R100 nominal (par) value,
assuming a yield to maturity of 18% p.a. compounded semi-annually.
Ex.16.10 Calculate the price of an inflation-linked bond which pays a coupon of
12.5% per annum, semi-annually in arrear and which is redeemable at
par for a nominal amount of R1 000 000 after 15 years. You may assume
a yield to maturity of 11% per annum compounded semi-annually, and
an annual rate of inflation of 2.5%.
Chapter 17

Equities

“Equities” refers to the shares of a company. The owners of the shares of a company
are referred to as shareholders, and own the company. They can participate in
important decisions made by the company. Each share in a company also entitles
the owner to share in the residual profits of the company. In practice, equity
investors receive their returns through dividend payments, which are a share of
profits which the directors of the company agree to pay out to shareholders each
year based on the company’s performance. They can also realise a return from
capital gains achieved when the share is sold, i.e. a capital gain is the difference
between the price at which they sell the share and the price at which they bought
the share for.

Example 17.1 Zali buys 200 shares of BigCompanyInc at R1 000 per share,
i.e. she invests R1 000×200 = R200 000 in BigCompanyInc shares. BigCompa-
nyInc has issued 100 000 shares, so Zali now owns 0.2% of BigCompanyInc.
During the year, BigComanyInc makes sales of R100 million, pays expenses
and costs of R80 million, and pays out interest of R5 million on bonds it has
issued. At the end of the year, BigCompanyInc needs to decide on dividends.
The directors of the company will consider issues such as:
R100m−R80m−R5m
• the profit per share is 100 000 = R150,
• the dividend paid last year was R80 per share, and
• there is an opportunity to acquire a small competing business for R8 mil-
lion next year.
The directors think it would be a good idea to keep back some of the profits
to fund the acquisition of the new business. They distribute a dividend of
R85 per share, i.e. R85 × 100 000 = R8.5 million. The remaining R6.5 million
is kept back to fund the acquisition of the small company (the remainder of
the acquisition cost will be borrowed).
Zali receives R85 × 200 = R17 000 as a dividend, deposited into her bank
account.
During the next year, the acquisition of the small company really pays off and

333
334 Chapter 17 Equities

600

Underlying value Share price

500

400

300

200

100

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14

Figure 17.1 The true underlying value of a share versus the market price.

the share price increases to R1 200 per share. Zali needs some money so she
decides to sell her 200 shares at this price. She gets R1 200 × 200 = R240 000
from the sale. Her total investment return is made up of R17 000 income
(from the dividend), and R40 000 capital gain from the sale of the shares.

17.1 Characteristics of equity

An investor could invest in equity by buying into a private business, but most
commonly when we talk of equities from the perspective of the institutional
investor, we will be referring to shares in public companies which are listed on
stock exchanges. In South Africa, listed shares are quoted on the Johannesburg
T Stock exchanges are market Stock Exchange (JSE).
places where listed shares
The returns on equity investments are driven, in the long run, entirely by
are traded.
the financial performance of the company. In the short run, however, returns
T Some large South African are driven by supply of and demand for the share which causes short term price
companies have their pri- movements; such price movements are heavily influenced by market sentiment,
mary listings overseas, and or investors’ views on the company’s prospects, as well as views on the prospects
many of these are on the
of equities in general relative to other asset classes. This can result in the share
London Stock Exchange.
price being different to its true underlying value. This is illustrated in Figure 17.1.
17.2 Types of equity 335

Example 17.2 BigCompanyInc is now trading at R1 200 per share. An arti-

cle is published indicating that one of the directors of BigCompanyInc has
tweeted "I’ve just about had enough of this job!". Investors react by thinking
that the director will leave; without this person it is expected that BigCom-
panyInc will be in trouble; profits will fall and the value of BigCompanyInc
shares will plummet. Investors respond by selling BigCompanyInc shares
immediately. The share price falls since everyone is trying to sell. By the
next day, BigCompanyInc shares are trading at R800 per share.
It is then discovered that the director meant to type “hob” not “job”, and
that the tweet related to an old stove that was giving him trouble. He has no
intention of leaving BigCompanyInc (he is, however, in the market for a new
cooker). Investors react immediately by buying back their shares . . . within a
few hours, the price is back to R1 200 per share.
Nothing intrinsic changed about the company over the last day, but the price
fluctuated wildly due to market sentiment. Moral of the story: be careful
what you tweet. . . and how you respond to market information!

Equity investments are therefore by their nature inherently riskier than the
other asset classes we have dealt with so far. Investors require some compen-
sation for bearing this risk, with the result that equity returns have significantly
outperformed those of bonds and cash over any suitably long period in history.
Equities have also tended to outperform inflation consistently over the medium-
to long-term. They are thus regarded as an appropriate investment for long-term
real liabilities.

Example 17.3 The retirement benefits for a young defined benefit fund
member are a good example of a long-term real liability. This liability is
long-term in the sense that cash outflows may only be expected to occur
in 40 years’ time, and the liability is real in the sense that it is linked to
salary increases, which in turn are expected to be linked to inflation, over
the member’s working career. Equities are a good match for this liability.

17.2 Types of equity

The two main types of equity are ordinary shares and preference shares.

17.2.1 Ordinary shares

The shares we have been discussing until now have all been ordinary shares.
Ordinary shares are the most common way in which companies in many countries
are financed, i.e. they are the most common means by which companies raise
capital to finance their operations. They offer investors the potential for high
returns at a high risk — in particular, the risk of capital losses.
336 Chapter 17 Equities

Ordinary shares are the lowest ranking form of finance issued by companies.
Dividends are not a legal obligation of the company and are paid at the discretion
of the company.
T When we talk about the The expected overall future returns on ordinary shares should be higher than
ranking of equities, we mean
those for most other classes of security to compensate for the greater risk of
the ranking relative to other
means of financing which default and for the variability of returns. The return on ordinary shares is made
may have claims to the as- up of two components, the dividends received and any increase in the market
sets of the company. For price of the shares.
example, bonds are ranked Listed ordinary shares are generally very marketable. The shares of large com-
higher than equities, and panies are usually very actively traded on a stock exchange. The shares of smaller
this means that the coupons companies, however, may be traded less actively and so will be less marketable.
owed to the bondholders
Other limitations on marketability can arise in practice; for example, if a company
will be paid before dividends
are paid to shareholders.
only has a few shareholders who each hold large proportions of the share issue
Similarly, if a company is and do not trade very often, the share will not be very marketable. Ordinary
wound up, the bondhold- shares are not considered to be liquid as share values are very unpredictable; it is
ers will be paid what they difficult to predict what price you will get if you decide to sell a share.
are owed first, before the Ordinary shareholders get voting rights in proportion to the number of shares
shareholders are given any held, so shareholders have the ability to influence the decisions taken by the
money.
directors and managers of the company.
T In May 2013, the JSE reached
a new record number of 17.2.2 Preference shares
daily trades: 290 259 trades
were concluded, with 595.9 Preference shares are less common than ordinary shares. Assuming that the
million shares worth R34.3 company makes sufficient profits, preference shares offer a fixed stream of income.
billion changing hands on That is, they do pay dividends, but preference share dividends are set at the issue
a single day. Read more at: date of the share, and do not vary from year to year. The company may choose
[Link] not pay out the promised dividend in some years, but this dividend is then owed
to the preference shareholders and must be paid to them in the future. If a
company owes dividends to preference shareholders, they must pay the owed
dividends before any dividends are paid to ordinary shareholders. The investment
characteristics of preference shares are more like those of bonds than ordinary
shares.

Example 17.4 SmallCompanyInc has issued 5 000 preference shares, which

pay a dividend of R100 each, and 20 000 ordinary shares. In 2013, the com-
pany has a profit of R900 000. The first thing it must do is pay preference
share dividends: at R100 per share and 5 000 shares, that will cost R500 000.
The remaining R400 000 can be distributed to ordinary shareholders. If
SmallCompanyInc chooses to distribute all of the remaining profits, that
will be R20(= R400 000/20 000) per ordinary share.
In 2014, the profits are only R300 000. SmallCompanyInc cannot afford to
pay out all its preference shareholders. It pays the full profit to the preference
shareholders, and it still “owes” them R200 000. The ordinary shareholders
get nothing in 2014.
17.3 Classification of equities 337

In 2015, the profits are R700 000. The first call on profits is the R200 000
owed to the preference shareholders from the previous year. The next call
on profits is the R500 000 owed to the preference shareholders for 2015. After
that, there is nothing left for the ordinary shareholders, again.

The crucial difference between preference shares and ordinary shares is that
preference share dividends are limited to a set amount which is almost always
paid. Preference shareholders also have no voting rights, unless there is money
owed to them.
In a given company, the risk of preference shareholders not getting their
dividends is greater than the risk of loan stockholders not being paid, but less
than the risk of ordinary shareholders not being paid a dividend. As a result, the
expected return on preference shares is likely to be lower than on ordinary shares
because the risk of holding preference shares is lower. Marketability of preference
shares is likely to be similar to that of loan capital (i.e. bonds).

17.3 Classification of equities

It is often very useful to categorize equities in different ways, especially when
wanting to analyse and compare the performance of different companies. The
features that can be used to categorise individual shares include:
• industry grouping,
• size of the company, and
• exposure to the economic cycle.
We discuss each of these in turn.

17.3.1 Industry grouping

Shares are often grouped according to the industry in which they operate. This
makes sense as the factors affecting one company within an industry are likely to
be similar, or even the same, as those affecting other companies in the industry.
Furthermore, much of the information for companies in the same industry will
come from a common source and will be presented in a similar way.
Research has shown that, after overall market movements have been taken
into consideration, the share price movements of companies within industrial
groupings tend to correlate more closely with each other than with companies in
other industries. The share price movements often reflect the changes that have
occurred in the operating environment and such changes affect companies in the
same industry in similar ways.
Companies in the same sector will, for example,:
• use similar resources (e.g. labour, land, and raw materials), and will therefore
have similar input costs, and will
338 Chapter 17 Equities

• supply their goods and services to the same markets, and will, therefore, be
affected similarly by changes in demand.
Figure 17.2 presents a diagram illustrating the high-level classification system
used on the JSE. One of the most important sectors in South Africa is the resources
T Market capitalisation refers sector as the companies in this sector make up about 28% of the total JSE market
to the total value of the is-
capitalisation.
sued shares in a publicly
traded company,
Exercises for 17.3.1 Industry grouping
i.e. market cap. =
Ex.17.1 Where do you expect the following companies to be classified? Use the
no. of shares × share price.
classification system in Figure 17.2.
• MTN
• Old Mutual
• Anglo American
• Kulula
• Pick and Pay
• Woolworths
Ex.17.2 Are your parents or other family working for listed companies? Try and
figure out which sector their company would be listed in.
Ex.17.3 Did you experience any difficulties with classifying some of the above
companies? Why?
You can check the sub-sector classification of a company on the JSE
here [Link]

The main difficulties with categorising companies by sector are:

• Some companies operate across several sectors; for example, Woolworths is
a retailer of both food and clothing. This is becoming more of an issue as
T In fact, Woolworths’ sub-
companies increasingly diversify into new sectors.
sector classification, which
is the level below the sector • Companies in the same sector may still be very different to each other, e.g.
classification given in Fig- due to size, or because they operate within different niche markets of a
ure 17.2, is “Broadline Retail- larger market.
ers”. This classification re-
flects the difficulty of putting
Woolworths into a single sub- 17.3.2 Size of the company
sector.
Large companies, also referred to as blue chip stocks, have different investment
T The name “blue chip” came characteristics to smaller companies. Blue chip stocks are expected to have
from poker where blue chips market prices which are less volatile than those of smaller companies because
usually have the highest
they are better able to weather downturns in the economy, and are often able
value.
to operate profitably in such times. Investors may buy blue chip companies to
provide steady growth in their portfolios. Examples of blue chip stocks in South
? See if you can name some Africa are the shares of Old Mutual, or South African Breweries.
more South African blue
chip stocks.
17.3 Classification of equities 339

Cyclical Financial
Resources Basic Industries Cyclical services
consumer goods Institutions

Household
General
Mining Chemicals goods and Banks
retailers
textiles

Construction Automobiles Support

Oil&Gas and building Insurance
and Parts services
materials

Utilities Forestry and Non-cyclical Leisure and

Life Assurance
Paper consumer goods hotels

Electricity Steel and Specialty and

Beverages Transport
Metals other finance

Other General Medial and Investment

Health
Industries entertainment companies

Information Aerospace and Non-cyclical

Food producers Real Estate
systems defence services

Personal care &

Engineering and Food and drug Investment
Hardware household
Machinery retailers entities
products

Software and Diversified Drugs and

Telecoms
services Industrials Biotech

Electronic and
Electrical
Equipment

Figure 17.2 FTSE-JSE Global Classification System. For a more detailed description
of each sector visit [Link] Note that the diagram only shows the main
categories.
Source: JSE. URL [Link]
340 Chapter 17 Equities

17.3.3 Exposure to the economic cycle

The sector classification in Figure 17.2 includes sectors for companies which sell
goods or services which are labelled cyclical and non-cyclical. The performance
T The business cycle refers to of cyclical companies is heavily affected by the business cycle while non-cyclical
the up and down of a market.
companies are not affected as much.
These upswings and down-
swings tend to follow each Cyclical stocks are the shares of companies whose performance follows the
other and alternate over business cycle. Cyclical companies sell goods and services that are popular when
time. the economy is doing well, but that people don’t buy when the economy is doing
badly (and people don’t have a lot of money). For example, a person who needs
a new car may choose not to buy a new car for another year if they are worried
about losing their job, but in good times people will replace their cars as soon as
they need to.
Non-cyclical stocks are also called “defensive” stocks, because they are usually
unaffected, or at least less affected, by the business cycle. For example, even if
the economy is doing poorly, people still have to eat, and get medical care. So
companies which provide non-cyclical goods and services are less affected by the
economic cycle and provide more stable returns.

17.4 Equity valuation

There are a number of methods used by practitioners to value equities. You
will be exposed to many of these in the coming years, but for this course we
will concentrate on the (by now) familiar method of using the present value of
future cashflows (also known as the “discounted cash flow method”). Recall from
Module 1 that in order to value equities, we need to ask ourselves two questions:
what are the expected future cash flows from a share, and what is the appropriate
interest rate to discount these cash flows?
Equities are assets of indefinite term, and do not provide a terminal capital
payment as do bonds. In Module 1, you learnt to value equities as the sum of the
present values of an infinite stream of dividends growing at a rate of g per annum.
Recall that the value S 0 of a share immediately after the annual dividend payment
can be calculated as follows:
∞
D 0 (1 + g )t v it
X
S0 =
t =1
1+i
= D 0 a∞ j where j = −1
1+g
D0
= ,
j
where
• i is the risk discount rate = risk-free rate + risk premium,
• g is the expected future dividend growth rate, and
• D 0 is the value of the dividend just paid.
17.4 Equity valuation 341

Note that the underlying assumption is that g < i .

The alternative equation using D 1 , the dividend expected in one year’s time,
is:
∞
Dt v t
X
S0 =
t =1

Assume that D i = D i −1 (1 + g ) for i ≥ 2

∞
D 1 (1 + g )t −1 (1 + i )−t
X
⇒ S0 =
t =1
D1 1+i
= a ∞ j where j = −1
1+g 1+g
à !
D1 1
=
1 + g 1+i − 1
1+g
D1
= .
i −g
Note that, again, the underlying assumption is that g < i .
In order to use these formulae to value equities, we need to decide on the
appropriate discount rate to use. A useful starting point is the rate on long-term
government bonds. Clearly, if we can achieve this yield with certainty, as we can
with government bonds, we would not expect to earn less on an equity investment
whose cashflows are uncertain. In fact, investors generally would demand higher
expected returns to compensate them the additional risk of equity investments.
The big question is: how much higher should the discount rate on equities be
than the risk-free rate? This differential is known as the “equity risk premium”
and is a matter of debate; commonly a figure in the range from 2% to 6% would
be chosen. Later in this chapter, we examine the relative performance of bonds
and equities over long-term history in South Africa, and comment on the historic
equity risk premium.

Example 17.5
Let’s assume that the risk-free rate is currently at 10% p.a. and that we feel that an
equity risk premium of 4% p.a. is appropriate. We wish to value an equity holding
which has just paid a dividend of R12.50 per share, which we expect to grow at 6%
p.a. going forward. The dividends will be paid annually.

Solution: We can find a value for the share as follows:

D0
S0 =
j
12.50
= 1+0.1+0.04
1+0.06 − 1
12.50
=
0.07547
= R165.63.
342 Chapter 17 Equities

Therefore our valuation of the share indicates that the price should be R165.63.

Example 17.6
An industrial company, listed on the JSE, is trading at a price of R70 per share. The
company has just paid an annual dividend of R5 per share. Dividends (payable
annually) are expected to increase by 10% p.a. for the next ten years and thereafter
by 7% p.a. in perpetuity.
An investor is evaluating a possible investment in the company and uses a valuation
interest rate of 15% p.a. for the first ten years and 20% p.a. thereafter.
What value does the investor place on the company’s shares and explain whether
or not he will invest?

Solution:

1.12 1.110 1.110 1.07 1.072

· ¸ µ ¶· ¸
1.1
S0 = 5 + +...+ +5 + + ...
1.15 1.152 1.1510 1.1510 1.2 1.22
1.110
= 5a 10 j + 5 × a∞ k
1.1510
1.15 1.2
where j = − 1 = 0.04545, and k = − 1 = 0.12150
1.1 1.07
1 − v 10 5v 10
" #
j j
= 5 +
j k
= 39.48 + 26.39
= 65.86.

Therefore, the investor will not invest as his estimate of the value of the share is less
than the price of the share.

17.5 Evaluating equities

Investors have come up with many ways to compare and evaluate different eq-
uities. The price of an equity or the amount of dividend in themselves are not
very meaningful quantities, e.g. if share A is priced at R1 000 per share, and share
B is priced at R5 per share, that does not mean that share B is better value than
share A. The difference in price could be because share A listed many years ago
and has grown over time; or company A chose to issue fewer shares of larger size
than company B; or company A is much larger than company B; etc.
Some of the more commonly used methods for getting a feel for an equity
include:
17.5 Evaluating equities 343

17.5.1 Earnings per share

Earnings per share is a measure of the amount of profit made by the company for
the benefit of ordinary shareholders, and is defined as:

net income − preference share dividends

EPS = .
average number of ordinary shares

The net income is effectively the profit of the company net of tax. Preference
share dividends are also subtracted when calculating net income as they are seen
as more like a payment on a loan than a profit distribution. The remainder of
the profit is divided by the number of shares to get a feel for the level of income
that the company is making per ordinary share. EPS is not used by itself, but it is
generally used as a part of other ratios.

17.5.2 Price/Earnings ratio

This is the most common ratio used in equity analysis. It is defined as:

share price
P/E ratio = .
EPS
The share price is the current market price, which is quoted on the stock exchange
and can be easily found. The EPS could either be historical, or prospective.
Historical EPS is the earnings per share over the last 12 months. Prospective EPS
is the expected earnings per share over the next 12 months. Prospective earnings
per share is more subjective, but it allows the investor to factor in actual future
events that are expected to take place in the company.
The simple interpretation of a P/E ratio is that it shows how much investors
are prepared to pay per rand of future earnings of the company. What it really
reflects is how optimistic investors are about a particular stock — how much they
anticipate earnings to grow. So technically speaking, for two similar companies,
the one with the higher P/E ratio is the one that investors are expecting more
growth from.

Example 17.7 Company A has a share price of R110 per share, and over
the last 12 months it has earned R10 per share. It has no preference shares.
The (historical, or “trailing”) P/E Ratio is 110/10 = 11. Company B has a
share price of R287 per share, and over the last 12 months it has earned
R12 per share. It has no preference shares either. The trailing P/E ratio is
287/12 = 23.92.
This may indicate that the market is expecting more future growth from
share B, since it is willing to pay more per unit of earnings for company
B’s shares. But which of the shares should you buy? You will not be able
to tell from the P/E ratio alone. A company with a low P/E ratio may be a
company that has been undervalued by the market (in which case this may
be a buying opportunity) or a company that genuinely has poor prospects
344 Chapter 17 Equities

(in which case it should not be bought!). The P/E ratio is a starting point for
further analysis to identify buying opportunities.

Note that the P/E ratio is a very simple measure and can be distorted by
several factors. These include:
• the way that companies calculate earnings,
• the industry in which a company operates, and
• once off events in a company’s life.
The ratio should be used with caution and should be used to compare companies
in the same sector. A wider analysis should be done to check why a P/E ratio may
be higher or lower than another.
U See [Link] for
a short tutorial on how the
P/E ratio works. 17.5.3 Dividend yield
Another common ratio is the dividend yield, defined as:

annual dividend
dividend yield = .
share price

The dividend yield is similar to the inverse of the P/E ratio, except that it uses
? Would a stock that the mar- dividends per share rather than earnings per share in its calculation. It can be
ket expects to have a lot of
used to identify high income stocks, and also can be used to get a feel for whether
dividend growth have a high
a stock is expecting a lot of growth.
or low dividend yield com-
pared to its peers? The same caution as for P/E ratio should be applied when using any single
ratio to make investment decisions.
U Have a look at [Link]
fEpIri for some other simple
ratios that investors can use 17.6 Equity indices
to evaluate shares.
An equity index is a statistical measure that represents the relative changes in the
share prices of the constituent companies which make up the index. This number
summarises the fluctuation of share prices on a given day. The primary purpose
of an index is to reflect the aggregate movement of the constituent companies (or
market) it represents. Hence, a single index value would be meaningless if not
compared to a previous/historical value.
The most important equity indices in South Africa are the FTSE/JSE Africa
Headline Indices which comprise of:
• FTSE/JSE All-Share Index (ALSI), which consistis of 99% of all listed compa-
nies; Figure 17.3 contains a plot of the ALSI value for the years 2002–2013;

T See the fact sheet at http:

//[Link]/zuEFXV for infor- • TOP40, which consists of the 40 largest stocks, constituting around 84% of
mation about the ALSI. the ALSI;
• Mid-cap Index, which is made up of stocks from position 41 to 100, and
covers around 14% of the ALSI;
17.6 Equity indices 345

ALSI Value
45 000.00

40 000.00

35 000.00

30 000.00

25 000.00

20 000.00

15 000.00

10 000.00

5 000.00

0.00
1/1/2002
4/1/2002
7/1/2002
10/1/2002
1/1/2003
4/1/2003
7/1/2003
10/1/2003
1/1/2004
4/1/2004
7/1/2004
10/1/2004
1/1/2005
4/1/2005
7/1/2005
10/1/2005
1/1/2006
4/1/2006
7/1/2006
10/1/2006
1/1/2007
4/1/2007
7/1/2007
10/1/2007
1/1/2008
4/1/2008
7/1/2008
10/1/2008
1/1/2009
4/1/2009
7/1/2009
10/1/2009
1/1/2010
4/1/2010
7/1/2010
10/1/2010
1/1/2011
4/1/2011
7/1/2011
10/1/2011
1/1/2012
4/1/2012
7/1/2012
10/1/2012
1/1/2013
4/1/2013
7/1/2013

Figure 17.3 ALSI values for 2002–2013.

• Small-cap Index, which is made up of stocks from position 101 and lower,
and covers about 2% of the ALSI;
• Fledgling Index consisting of the 1% of listed companies not included in the
ALSI.
The most important equity indices in overseas markets include:
• UK: FTSE 100 Index and FTSE All-Share Index;
• USA: The Dow Jones Industrial Average, the NASDAQ, and the Standard &
Poor’s Composite Index (often referred to as the S&P 500);
• Japan: the Nikkei Stock Average 225 and the Topix;
• Germany: the Deutscher Aktien IndeX (DAX);
• France: the CAC General Index.
" Note that in this section we
present arithmetic averages,
i.e. a sum of the observed
annual returns divided by
the number of observations,
rather than geometric av-
erages. They give a reliable
estimate of the average re-
turn in any one given year
but tend to overstate the
compound return that a no-
tional investor would have
received over the period —
can you see why?
346 Chapter 17 Equities

17.7 Historic asset class returns in South Africa1

Over the years from 1900 to 2007, the average annual increase in the Consumer
Price Index in South Africa was 5.0% p.a. Over that same period, money invested
in the money market would have earned 6.1%, or a little more than 1% p.a. in
real terms. Long-term government bonds returned an average of 7.1% p.a., while
equities were the best-performing of these asset classes, delivering 14.9% on
average.
U You will have learnt in Stats It’s equally interesting to consider the standard deviations of returns from
that standard deviation is a
these three asset classes. The money market has shown a standard deviation of
measure of volatility.
5.71% over the period, while the corresponding figures for bonds and equities
are 9.38% and 22.62%, respectively. So equities have delivered far higher average
returns than bonds and cash over the long-term history in South Africa, but
these returns have also been significantly more variable. This is in line with
our expectation: asset classes with highly variable returns will be perceived as
being more risky, and consequently investors will require higher returns to induce
investment in them.
Remember though that these measures of average return and standard devi-
ation are annual measures. In the context of long-term actuarial liabilities, we
ought to think about the relative risks and returns over suitably long periods
of time. As we have seen, equities have significantly outperformed bonds over
more than a century of investment market history in South Africa, but this is the
extreme long-term, even by actuarial standards. It is striking to note that bonds
fail to outperform equities in any rolling twenty-year period from 1920 onward, of
which there are 89 in the sample period! This is the primary reason that equities
(and to some extent property) are generally a major part of investment strategies
to back long-term real liabilities.
The above figures suggest that the historic equity risk premium has been a
little more than 7% p.a. You may question why our discussion of equity valuation
? Can you see how we calcu- above suggested the use of a margin of between 2% and 6% p.a., in the light of this
lated this from the returns
historic experience. A full answer to this question is multi-faceted and beyond
above?
the scope of this course; suffice to say for now that the relative returns on equities
and bonds are a function of the investment market history in South Africa and in
some periods have been distorted by artificial restrictions placed on the holdings
of institutional investors. In general, forward-looking expectations of the equity
risk premium would be considerably lower than this long-term average. If we look
at more recent history, we note that equities have outperformed bonds by only
3.3% p.a. on average in the years from 1994 to 2007, although it should be noted
that this was a period of strong bond performance, coinciding with a significant
fall in interest rates.
1 The sources of the figures used in this section are:

Firer, C. and Staunton, M. (2002), 102 years of South African financial market history, Investment
Analysts Journal, 56, 57–65. URL: [Link]
Firer, C. (2008), Spreadsheet provided with data for the 2002 paper updated to 2007.
17.8 Summary of equity characteristics 347

17.8 Summary of equity characteristics

Using SYSTEM T, we can summarise the investment and risk characteristics of
equities as follows:
S Security:
The security of the dividend income will depend on the company issuing
the shares. In particular, it will depend on the stability of the company’s
profits — the more stable the profits, the more stable the dividends. The
ratio of earnings to dividends (dividend cover) also matters — if a company
has historically only paid out a small proportion of its earnings, it has more
flexibility to keep dividends stable. A company that pays out most of its
earnings has very little room for smoothing its dividends. If the company is
wound up, the shareholders will receive the residual assets after all creditors
have been paid.
Y Yield (real vs. nominal):
Historically, equities have provided a real yield over the long term. This
means that they have provided a yield in excess of inflation. This is because
company profits tend to rise with inflation and economic growth and hence
so do dividends. However, the relationship is a loose one, i.e. there is no
guarantee of inflation protection.
Y Yield (expected return relative to other asset classes):
Equities are perceived to be more risky than bonds and would be expected
to give a higher return to compensate for this. The margin will depend on
the company issuing the shares/bonds.
S Spread:
Both equity prices and dividends can be volatile. The price of individual
equity shares is determined by the interaction of supply and demand. Some
investors in the market will buy or sell on the basis of short-term speculation,
but the most important basis for buyers and sellers when deciding on the
price for a share is an assessment of its value based on the present value of
future dividends.
T Term:
Equities can generally be held in perpetuity.
E Expenses:
The costs of dealing in equities are closely linked to marketability, i.e. the
more marketable the share, the lower the trading costs will be. Dealing
expenses are generally greater than for bonds, but this depends on the
relative marketability of the stocks being compared.
E Exchange rate:
Equities are available in many overseas countries. To this end, there will be
a currency risk for an investor who is investing in equities denominated in
one currency but who has liabilities denominated in another.
348 Chapter 17 Equities

M Marketability and Liquidity:

The marketability of equities varies enormously between companies. In
general terms, the larger the company, the better the marketability. This
relationship is not perfect, however. For example, where only a few investors
hold a large proportion of the shares in a particular company, the marketabil-
ity could be low since those investors may not be willing to trade very often.
The extreme in poor marketability will be shares that are not listed on any
recognised stock exchange. Such shares can only be sold by finding another
party who wishes to buy the shares.
Shares are not considered to be very liquid because the value realised on
sale is not predictable.
T Tax:
Income and capital gains from shares may be taxed differently. In addition,
different investors will often pay differing tax rates on these two elements.
Chapter 18

Property

Property, also commonly referred to as real estate, is an investment that gives the
investor legal rights to the use of land and/or buildings. Most investors purchase
property with the intention of leasing it out, and therefore collecting income in
the form of rent. Properties can also be sold, which can lead to capital gains and
losses.

18.1 Characteristics of property

18.1.1 Nature
Like equity, property is a real asset and is considered to be a good match for U A “real asset” here means an
asset that delivers positive
inflation-linked liabilities. The reason property is a real asset is that the property
real returns, i.e. nominal
owner should be able to increase rents in line with inflation. Because of its real returns greater than the in-
nature, property is often considered as an alternative to equity investment. flation rate.

18.1.2 Marketability ? Will rents always keep up

with inflation? Think about
The most pronounced difference between property and equity lies in the much how rents are set, and about
reduced marketability of property. Property is not marketable because of the how tenants behave.
following three characteristics:
U If you would like to invest in
• large unit sizes generally make direct investment in property unfeasible for a retail property, you would
most investors, i.e. one property can cost several million Rands which not have to buy a whole mall
all investors can afford; for millions of rands. If you
wanted to invest in a retail
• uniqueness of properties, i.e. a property will have unique features particular
share, you could buy a single
to it which make it less marketable than the identical shares in a company; Pick n Pay share for very
• valuation is much more difficult since properties are infrequently traded much less.
and there is no market quoting prices for them between trades, unlike listed
? Why is a unique item less
equities. marketable?
The fact that there is no exchange on which property is traded also makes ? How would the fact that the
property unmarketable. value of the asset is unknown
make it less marketable?

349
350 Chapter 18 Property

Exercises for 18.1.2 Marketability

Ex.18.1 We have discussed how property is unmarketable. Is it liquid?

Ex.18.2 Do you think a free-standing house in Woodstock is more or less mar-
ketable than a house that is part of a housing estate? Why?

18.1.3 Yield
The expected yield from property investments will normally be comparable with
that of ordinary shares. Property has advantages and disadvantages over equity.
The disadvantages include:
T Residential leases generally
renew annually, but com- 1. dividends usually increase annually, whereas rents are reviewed less often;
mercial leases — leases to 2. property is much less marketable;
businesses — are often fixed
for longer periods. 3. expenses associated with property investments are much larger;
4. large indivisible units of property are much less flexible than a portfolio of
? What type of expenses does shares.
a property investor have to
cover? But on the other hand:
1. property is tangible – it consists of real land, bricks, and mortar. Investors
find security in having a real asset they can see.
2. property is less volatile than equities. One of the reasons for this is that it is
not traded as often. In addition, property always has some residual value
(because it has real land), which means that property prices can’t fall to zero
as they can for a company that goes bankrupt.
3. property is desirable as it provides diversification away from equities.
4. some of the upkeep costs can be delegated to the tenants.
5. the owner can change the nature of the property.

Exercises for 18.1.3 Yield

Ex.18.3 How could you change the nature of a residential building on Long
Street that you have just bought? The building currently has 10 small
apartments.

Properties are generally expected to deliver returns of around inflation + 4%

p.a.
U The running yield is the The running yields on property are generally higher than those of equities.
ratio of income to value of
When estimating the expected returns on a particular property, one must
an asset. So for equities, that
is the dividend yield; for allow for the possibility of void periods. Void periods are periods when the
property, it’s rent/price. property has no tenants, and therefore no rental income will be received. In this
case, the only return on the property will be through capital growth. In addition,
U Can you think of reasons one should also account for the expenditure on maintaining the property over
why the running yield on time and the possibility that the building will need to be modernized in the future.
properties is generally high? These are additional costs which reduce the returns from property investments.
18.2 Property markets 351

18.2 Property markets

There are a number of factors that distinguish the property market from other
financial markets. These include:
• low marketability and liquidity – in both the commercial and residential
property markets, liquidity is very low in comparison with other financial
markets.
• cumbersome settlement and clearing processes – property generally re-
quires repeated exchanges of physical documents amongst legal specialists,
which can take a substantial amount of time and money.
• lack of pricing transparency – the most consistent and reliable prices come
from prior transactions on the property and other properties in the area.
These data may be weeks or months old by the time they are published, and
in many cases they may never be disclosed.
• asymmetric price movements – property market prices tend to move up
more readily than downwards. This is because property sellers who do not
receive offers in their desired price range tend to resist selling, and may even
withdraw their property from the market.
• significant holding costs – there are a number of significant costs associ-
ated with holding property including maintenance expenses, refurbishment
costs, rates and taxes.
Another risk with property is that property is susceptible to government interven-
tion through rent and planning controls (which may limit the supply of property).

18.3 Property types

The main sectors of the property market include:
• residential property;
• retail property, including shopping centers, retail warehouses, standard
shops, supermarkets, and department stores;
• offices, including standard offices and business parks;
T Note that “commercial prop-
• industrial property, including industrial estates, distribution warehousing, erty” includes all of retail, of-
and logistics facilities; fices, industrial and possibly
agricultural property.
• agricultural property, including fruit farms, wine estates, and wildlife parks;
• leisure property, including leisure parks, restaurants, pubs, and hotels;
• healthcare properties, including hospitals and clinics.
352 Chapter 18 Property

Exercises for 18.3 Property types

Prime Property
When investors compare various Ex.18.4 List some shopping malls in Cape Town which you think would be
properties, the properties which prime property. What made you choose them? What do you think
are deemed to be the most attrac- classifies a shopping mall as prime property?
tive to these investors are normally
called “prime” property. A number Ex.18.5 Do you think you would expect to get a higher yield on a prime or a
of factors would influence whether non-prime property? Why?
the property is classified as prime
property and these will differ by
the type of property. These factors
include location, age and condi- 18.4 Residential vs. non-residential property markets
tion, quality of the tenant, lease
structure, size, rental yields and
number of comparable properties Exercises for 18.4 Residential vs. non-residential property markets
available for valuation purposes.
Ex.18.6 Have you or your family ever rented property? If you are living in
residence, you are renting from UCT.

Table 18.1 Definition of prime 1. When renting residential property, how long is the contract usually
property. for?
2. How frequently does the rent go up?
3. When something breaks, who must pay for the repairs?
While you are probably most familiar with residential property, the majority of
the property bought for investment purposes is non-residential. This is property
for business use, like offices, factories, and shops. There are a number of critical
differences between these property markets and residential property markets,
including:
• lease period – residential tenants typically commit to relatively short renew-
able leases (normally 1 year), while commercial ones usually sign long-term
contracts (periods of 10 years or more are not uncommon);
• property repairs – commercial tenants are normally liable for repairing the
property while landlords of residential property are usually responsible for
repairs;
• return profile – the returns on residential property come mainly from in-
creases in capital value, whereas a large part of the commercial property
return is income;
• capital values – commercial properties usually cost significantly more than
residential properties. For example, in the case of shopping centres or large
office buildings these can cost hundreds of millions of Rands.

18.5 Direct vs. indirect property investments

18.5.1 Direct property investments
Direct property investments may be made in offices, shops, factories and indus-
trial units, warehouses, shopping centres, retail warehouses, agricultural, and
18.5 Direct vs. indirect property investments 353

residential property. Direct investment means that the investor buys and owns
the property.
There are a number of disadvantages with direct property investment. These
include:
• size of investment – most properties require a significant upfront invest-
ment and this is generally too large for most individual investors. In addi-
T By investing in multiple prop-
tion, in order to create a well-diversified portfolio, a number of properties
erties, with different charac-
need to be purchased which may be impractical for smaller funds and indi- teristics, a portfolio becomes
vidual investors due to the size of the investment. more diversified and this re-
• lack of marketability – there are a number of costs associated with buying duces the investment risk.
and selling property which reduce the marketability of property. In addition,
it can take several months from initial marketing of a property to completion
of the sale and transfer of monies. This time lapse clearly represents an
? What is the opportunity cost
opportunity cost for investors.
referred to here?
• valuation of property – obtaining estimates of the values for different prop-
erties can be expensive and also very subjective. The true value of a property
will only be known once a sale has gone through.
• expertise required – most investors do not have the specialist expertise in
the local market or the time required to manage property investments so as
to ensure that they are successful investments.

18.5.2 Indirect property investment

Some of the difficulties with direct property investments can be minimised by
investing in property indirectly. Indirect property investments include investing
in property companies and trusts. Property companies and trusts purchase and
manage many properties through the legal structures of companies and trusts.
An investor invests in a property company by buying shares in that company,
and they invest in a property unit trust by buying units in the unit trust. Individual
shares and units will have a much smaller value than the property portfolio held
by the company or trust. This enables smaller investors to obtain exposure to
the property market without being limited by large unit size, and allows diversifi-
cation of property holdings. It also allows investors to obtain access to property
investment expertise without having to acquire the expertise themselves. In addi-
tion, these investments are more liquid than direct property investments because
the units/shares can be sold more easily at relatively stable prices.

Pooled property funds

Pooled property funds are collective investment schemes (either open-ended

unitised funds or closed-ended investment trusts) which enable exposure to
the property market by selling units to investors. The money raised from selling
these units is then used to purchase various properties according to the objectives
of the fund.
354 Chapter 18 Property

In an open-ended unitised fund, managers can create or cancel units in the

fund as new money is invested in or disinvested from the fund. In a closed ended
investment trust the number of units in the trust is fixed and after the launch of
the fund, the only way of investing in the fund is to purchase units from a willing
seller.
T Gearing is an important con- One of the disadvantages of pooled property funds is the restrictions that may
cept in investing. It means
be placed on them. These restrictions vary from country to country. Possible
borrowing money to invest.
For example, a company restrictions include limitations on the categories of properties that can be held by
may invest R100 000 in a the fund, the management expenses that can be charged, the maximum level of
property, use the property gearing, and the tax reliefs available.
as security for a loan of
R100 000, and invest the
Property company shares
R100 000 in another prop-
erty. Their final position is An alternative to investing in pooled property funds is to invest in the shares of
said to be geared. Gearing property companies, such as companies which develop properties or companies
increases the potential gains,
which invest in properties themselves. These companies have the advantage that
but also increases the poten-
they do not have restrictions placed on the investments that they can make or
tial losses.
on the management expenses that they can charge. This means that the investor
can gain exposure to properties that they may not have been able to with pooled
property funds. However, the investor is now buying shares rather than property
and this may have disadvantages for the investor. In particular, the investor would
be exposed to equity market movements, i.e. their returns would not only reflect
property market movements, and they would also be affected by any constraints
placed on their equity investments.
Exercises 355

Exercises

Exercises for 18 Property

Ex.18.7 Use the SYSTEM T mnemonic to analyse the investment and risk char-
acteristics of property investments.
Ex.18.8 Discuss the factors that would determine whether the following prop-
erty types would be classified as prime property:
• offices,
• factories,
• shopping centres, and
• residential property.
Ex.18.9 Discuss the advantages and disadvantages of investing in property
directly compared with investing in property companies.
Chapter 19

Derivatives

Derivatives are securities whose price is dependent on or “derived” from the price U Variables include indices, ex-
change rates, interest rates,
of other assets or the value of other variables. Derivatives are effectively contracts
and other market indicators.
where two parties agree to make a specified transaction at a specified future date
(or dates).

Example 19.1
Rob, a wheat farmer, has just planted a field of wheat. Wheat is currently
trading at R3 450 per ton, but Rob will only harvest the wheat in 6 months. If
he can still sell it at R3 450 per ton, he will be fine, but if the wheat price in
6 months falls, his business will suffer. Rob can enter an agreement with a
buyer to sell his wheat in 6 months time at R3 450 per ton regardless of what
the market price is at that time. This agreement is an example of a simple
derivative contract.

An important feature of a derivative is that it is always a contract between two U Derivatives may seem like
very innovative instruments,
parties. Any profits made by one party will be the losses of the other, i.e. no value
but they have been around
is created by derivatives. for hundreds of years. For ex-
ample, the Yodoya rice mar-
Example 19.2 ket in Osaka, Japan, offered
something substantively
Rob enters into a derivative contract with Katlego, who agrees to buy Rob’s wheat in
similar to modern futures
6 months time for R3 450 per ton. There are now 3 possible outcomes:
contracts as far back as 1650!
(a) wheat prices in 6 months time are less than R3 450 per ton;
(b) wheat prices in 6 months are exactly R3 450 per ton; and
(c) wheat prices in 6 months are more than R3 450 per ton.
U This type of agreement is
The agreement is binding on both parties (that means that no matter what market
called a forward contract or
prices are, Katlego and Rob will trade with each other at R3 450 per ton of wheat).
a future if it is traded on an
What will be the outcome for each party in each of the 3 scenarios above?
exchange.
Solution: If wheat is now trading at, say, R3 000 per ton, Katlego has to buy it at
R3 450 per ton. He could have bought it for R450 less per ton in the market, and is
making a loss of R450 per ton. Rob, on the other hand, is happy. He is selling his

357
358 Chapter 19 Derivatives

wheat at R3 450 per ton when the best price he would have got otherwise is R3 000
per ton. Rob has made a profit of R450 per ton on this deal compared to what he
would have got in the market.
If wheat is trading at R3 450 per ton exactly, then the deal doesn’t matter; Katlego and
Rob could have bought and sold the wheat from/to anyone for the same amount as
they are trading with each other at. So neither makes any profit or loss on the deal.
If wheat is trading at, say, R3 600 per ton, Katlego buys it from Rob for a bargain; he
makes a profit from the deal of R150 per ton. Rob is losing out on the great market
T Katlego could buy the wheat
price though, so he makes a loss of R150 per ton on the deal compared to what he
from Rob and immediately
could have got in the market.
sell in the market for R150
In all cases, the profit that one party makes is equal to the loss the other party
more!
makes — derivatives are a zero-sum game!

19.1 Derivative Markets

Derivatives are traded either over-the-counter (OTC) or on organised exchanges.

19.1.1 Over-the-counter Markets

OTC markets generally have one (or a few) central dealer who tailor-makes deals
for the investors. The dealer finds out what one party is looking for, and then finds
T Think of OTC as a “shop” for another party, the “counter-party”, who is willing to take the opposite position.
derivatives where products
The parties directly transact with each other — the dealer is only assisting
are tailor-made for each
investor. them. There is no supervisory body to ensure that the contract commitments
are fulfilled. So OTC transactions have some amount of credit risk depending on
who your counter-party is.

Example 19.3 Investor A contacts a dealer because they want to buy a

specific derivative. The dealer phones around and finds a counter-party,
Investor B, who is willing to sell this derivative. The dealer helps the two
parties make the transaction, but their contract is with each other. If one of
A or B goes bankrupt or cannot fulfil the terms of the deal, the other party
will suffer.

19.1.2 Exchange Traded Markets

U You have already come Derivative exchanges are marketplaces where derivatives can be centrally bought
across exchanges where eq- and sold.
uities are traded. Think of
The key features of exchange traded derivatives is that the contracts are stan-
exchanges as big supermar-
dardised. In addition, clients deal with the clearing house rather than directly
kets where you can buy mass
produced goods! with a counter-party, and the clearing house acts as a counter-party to all trades.
When a trade is executed, a margin payment is required as security to reduce
U A margin payment is a de- credit risk to the clearing house. Trading is usually carried out electronically by
posit paid to the clearing
house.
19.2 Forwards and futures 359

entering the trades on a computer, however some exchanges still use the open
outcry system, where trades are carried out through dealers at the exchange, in a
“pit”, using specified hand signals.

19.2 Forwards and futures

Terminology
The person who sells a derivative is called the writer of the contract, or the short
party. The term “writer” is used because that person sets the terms of the contract,
i.e. writes the contract. The term “short” party is used because they are selling the
contract.
The person who buys a derivative is called the long party. ? Why is selling called “short-
ing”? Maybe because the
person is going to be short of
Forwards and futures are contracts that obligate both parties to trade a speci- the stuff they are selling!
fied asset (“the underlying”) at a specified future date for a specified price (the
“strike price”). The party that agrees to buy the asset is called the long party, and
the party that agrees to sell the asset is called the short party.

Exercises for 19.2 Forwards and futures

Ex.19.1 Investor A buys a forward contract for 1 000 MTN shares from Investor
B through a dealer, C. Investor A agrees to buy the MTN shares in 3
months time at R195 per share, and MTN is currently trading at R193
per share.
1. Who is the long party?
2. Who is the short party?
3. What is the strike price?
4. If the share price in 3 months is R200 per share, what profit or loss
did the long party and the short party each make?

The only difference between futures and forwards is that forwards are traded T Forward contracts are gen-
on OTC markets, and futures are traded on exchanges. Forwards are therefore not erally much larger than indi-
vidual futures contracts as a
standardised and can be designed to specifically fit an investor’s needs. Futures
result.
are standardised and generic.

Example 19.4
A futures contract is a standardized forward contract. Discuss some of the factors
that would need to be standardized in a futures contract.

Solution: A futures contract should specify:

• the underlying asset — what is the exact asset the parties are agreeing to
exchange?
360 Chapter 19 Derivatives

• the contract size, e.g. number of shares, barrels or weight of commodity to be

exchanged.
• delivery month, i.e. which month and when in the month does the contract
need to be settled.
• delivery arrangements; for financial futures, this is mostly cash, but for futures
on commodities such as coal or coffee, the contract will state acceptable
locations and methods of delivery.

T Are you the long or the short Say you buy a 6 month future on an asset at a strike price of R100. You have
party? therefore entered into a contract to purchase the asset for R100 in six months’
time. Ignoring taxes and transaction costs, if in six months’ time the asset is worth
more than R100 you will have made a profit equal to the difference between the
price at the time and R100. Conversely, if the asset is worth less than R100 you will
have made a corresponding loss. In practice, the asset is usually not physically
delivered, but the contract is rather settled in cash for the difference between the
asset price and the strike price.
T If you had bought the actual The profit profile to you (the long party) in six months’ time is:
asset and held it, what would
your profit profile be?
100

80

60

40

20
Profit

0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
-20

-40

-60

-80

-100
Price

Figure 19.1 Profit profile to the long party on a futures contract.

The profit profile to the short party, i.e. the seller of the futures contract, is the
mirror image of that to the long party:
19.2 Forwards and futures 361

100

80

60

40

20
Profit

0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
-20

-40

-60

-80

-100
Price

Figure 19.2 Profit profile to the short party on a futures contract.

Example 19.5
Say, for example, Bob has just entered into a long forward position to purchase
T “Entered a long forward po-
1 000 Anglo Gold shares for R145 per share in one years’ time. What are the possible
sition” means “bought a for-
outcomes?
ward”.
Solution: The forward contract obligates Bob to buy 1 000 Anglo Gold shares for a
total cost of R145 000. If the share price at the end of the 1 year rose to, say, R157.58,
the forward contract would be worth R12 580 (= 1000 × (R157.58 − R145) to Bob.
It would enable Bob to purchase the 1 000 Anglo Gold shares for R145 000 and
immediately sell them in the market for R157 580. Similarly, if the share price fell to,
say, R139.24 at the end the 1 year, the forward contract would have a negative value
to Bob of R5 760 (= 1000×(R139.24 - R145)) because it would lead to Bob paying
R5.76 more per share than the market price for Anglo Gold shares.

In general, given that the long party will pay a specified price for a given T Notice that the combined
quantity of the asset at maturity, the pay-off to the long party of a forward or payoff from both a short
and long forward position
future is S T − K , where S T is the price of the underlying at maturity (i.e. time T )
is (S T − K ) + (K − S T ) = 0, as
and K is the strike price. In contrast, the pay-off to the short party of the forward expected.
or future is K − S T . The graphs of the pay-off plotted against the final value of the
underlying asset are shown in Figure 19.3 for both the long and short position.
362 Chapter 19 Derivatives

Payoff Long Forward Short Forward

Payoff
Payoff = ST – K Payoff = K - ST
K

0
K ST 0 K ST

-K

Figure 19.3 Payoffs from a long and short forward.

Buying and selling a future or forward has the same net profit profile as buying
or selling the underlying asset. The difference is that you don’t need to buy the
T If the derivative was a future real asset, and the amount of money exchanged at the outset of the transaction is
rather than a forward, you
zero with forward contracts and very small with futures.
would have to pay the clear-
ing house an initial margin
on the day of purchase, and Example 19.6 If you purchase an actual asset for R100, you pay R100 on the
possibly further margin pay- day of purchase. After 6 months, if the asset is now trading at R120, you can
ments over the term of the
sell it and make R20 profit. If it is trading at R85, you would make a R15 loss
future contract if needed.
on the sale.
However, these payments
are much smaller amounts If you purchase a 6 month forward on the same asset with a strike of R100,
than the full price of the un- you pay nothing on the day of purchase. After 6 months, if the asset is now
derlying asset. trading at R120, you receive your profit of R20. If the asset is trading at R85,
you are required to pay R15 to the counter-party, i.e. you make a R15 loss.
So effectively, you have made a profit or loss without investing anything.

U The much-publicised col- This example illustrates one of the potential dangers of derivative trading: for
lapse of Barings Bank sev- very little initial investment, investors can take large positions in an asset and be
eral years ago, as well as the
exposed to enormous profits or crippling losses. This type of derivative trading is
more recent losses sustained
by Société Générale, both
called speculation.
resulted from traders elud- The other, less risky use for futures is as a hedging strategy.
ing their respective organ-
isations’ risk management Example 19.7 A pension fund is expecting a R1 million cash inflow in 9
protocols in order to take
months’ time, which it plans to invest in the equity market. The fund is
futures positions which they
concerned that prices in the equity market are going to increase over the
hoped would translate into
massive profits, but which next 9 months, and that it will have to buy equities at inflated prices when it
turned against them badly. receives the cash inflow. This risk could be hedged by purchasing a 9 month
futures contract on an appropriate equity index, allowing the fund to secure
U Hedging is defined as a re- the price at which it will buy into the market in nine months’ time. This
duction in the probability of strategy of course removes the opportunity to buy into the market at lower
losses as a result of adverse
prices if there is a drop in the market over the nine-month period.
movements in the market.
19.3 Options 363

19.3 Options
Whereas a forward or futures contract creates an obligation to trade an asset at a
specified future date for a specified amount, an option creates the right to do so
without the obligation for the holder of the option. This means that if conditions
are unfavourable at the maturity of the option, the holder can simply let the
option expire without exercising it, losing only the initial option premium that
was paid to the writer of the option. The premium is compensation for the right
to choose to make the transaction or not.
In this course we will focus only on very simple options. The kinds of options U Very simple options are also
referred to as “vanilla” op-
we will be talking about are called European options. European options can
tions, because vanilla is the
only be exercised at maturity. American options, in contrast, can be exercised at plainest icecream flavour.
any time prior to maturity. Options are available in both the OTC and exchange There are, unfortunately, no
markets. chocolate options!

Terminology
A call option gives the holder the right to buy the underlying asset at a future
date/s for the specified strike price.
A put option gives the holder the right to sell the underlying asset at a future date/s
for the specified strike price.

19.3.1 European call options

A European call option gives the purchaser the right to buy an asset at a specified U “European” is just the name
future date for an agreed price called the strike price. The holder of the option, i.e. of the option, it does not re-
fer to where they are traded.
the long party, does not have to exercise this right.
Other types include Amer-
ican, Asian, Barrier, Look-
Example 19.8 back, Rainbow and Quanto
What is the difference between holding a future and holding a call option? Consider options, and many more,
Previn, who thinks that the equity market is going to rise. He wants to benefit from all with their own charac-
this rise. He could buy some equities, but he is a bit short on cash right now, so he teristics. Have a look at
wants to use derivatives. He is considering either buying a future on the FTSE100 [Link] and
index, or buying a call on the FTSE100 index, at the same strike price. What are the [Link] for
differences between these two strategies? more information about
exotic options.
Solution: Consider 2 possible future scenarios:
(a) The FTSE100 is above the strike price at maturity.
• If Previn bought the future, he must now buy the underlying for the strike
price. He gets a bargain and realises a profit of

market price − strike price.

• If Previn bought the call option, he can now choose whether to buy the
underlying at the strike price or not. Given that the strike price is below
364 Chapter 19 Derivatives

the market price, it is in his interest to do so. He also makes a profit, but
he did pay a premium for the option, so his profit is

market price − strike price − premium.

(b) The FTSE100 is below the strike price at maturity.

• If Previn bought the future, he must now buy the underlying for the strike
price. He is buying the index at above market price, so he is making a loss
of
strike price − market price.
• If Previn bought the call option, he can now decide whether to execute
the option. Given that he will make a loss if he buys the stock at the strike
price, he should choose not to exercise the option. So the only loss he
makes is the premium he paid.

Notice that the main difference between a long future and a long call option is
that the holder of the long future is obligated to buy the underlying asset, whereas
the holder of the long call has the right but no obligation. As a result, there is a
cost for acquiring an option whereas it costs nothing to enter into a future.
So the buyer of a call has unlimited potential for profit, but the biggest loss he
will make is the premium.
? What would happen if no This profit profile for a long call position is shown below:
premium was paid on the
option?
100

80

60

40

20
Profit

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
-20

-40

-60

-80

-100
Price

Figure 19.4 Profit profile for the purchaser of a European call option.
In this case the strike price is equal to R100 and the premium is R10,
and so the profit to the long party is max(S T − 100, 0) − 10, where S T is
the market price at the maturity date.
The seller, or writer of the call option is in a different situation. They have no
choice about whether the option will be executed or not.
19.3 Options 365

Example 19.9 In the previous example, Previn would have had a counter-
party who would have been dealing with him. The counter-party’s losses
and profits would mirror Previn’s profits and losses. For example:
• If the market had gone up above the strike price, and Previn had bought
a call, Previn would want to exercise his right to buy the underlying asset
at the strike price. The writer of the call would have to sell it (he has
no choice). The writer of the option would however earn the premium
Previn paid for the option. So the writer’s profit is

premium − (market price − strike price).

• If the market had ended below the strike price, Previn would not exer-
cise his call option. So the writer would not be making any trades, but
he would earn the premium, so his profit would be just the premium.

The profit profile to the option writer is therefore again the mirror image of
the profit profile of the option purchaser:

100

80

60

40

20
Profit

0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
-20

-40

-60

-80

-100
Price

Figure 19.5 Profit profile for the writer of a call option. In this case
the strike price is equal to R100 and the premium is R10, and so the
profit to the short party is 10 − max(S T − 100, 0), where S T is the market
price at the maturity date.
In general, given that the long party holding a European call has the right to
buy an asset in the future, they will not exercise this right unless the current asset
price exceeds the strike price (as any rational investor would not choose to make
a loss). Therefore, the pay-off for the long party is max(S T − K , 0), and the profit at
maturity is max(S T − K , 0) − P , where S T is the price of the underlying at maturity
(i.e. time T ), K is the strike price, and P is the premium paid for the call option. In
366 Chapter 19 Derivatives

contrast, the short party will have the reverse position, i.e. the pay-off for the short
party is − max(S T − K , 0), and the profit at maturity will be P − max(S T − K , 0).

Exercises for 19.3 Options

Ex.19.2 Call options on Stock A are trading on an exchange for a premium of

R7. The call options have a strike price of R124 and a 3 month term.
For each of the cases below, state the amount of profit or loss made by
the relevant party:
1. Jack writes a call on Stock A. After 3 months, Stock A is trading at a
price of R90.
2. Gogo buys a call on Stock A. After 3 months, Stock A is trading at a
price of R90.
3. Sani writes a call on Stock A. After 3 months, Stock A is trading at a
price of R150.
4. Anton buys a call on Stock A. After 3 months, Stock A is trading at
a price of R127.
Clearly writing options is significantly riskier than purchasing an option. For
? Does the writer of an option the writer of a call option, the profit is limited to the option premium, but the
have a right or an obligation? potential loss is theoretically infinite. The purchaser of the call option faces the
reverse: their loss is limited to the option premium, but they have a potentially
unlimited upside.

Example 19.10
Consider the pension fund referred to in Example 19.7, which is expecting a large
cash inflow in nine months’ time. How could they use call options in their hedging
strategy instead of futures?

Solution: They could buy a call on the equity index with a term of nine months and
an appropriate strike price. The fund would be able to buy into the equity market at
the strike price, and would thus be protected against unexpectedly sharp increases
in equity prices. The important distinction from futures hedging is that the fund
would have a limited downside as it can allow the option to expire worthless and
buy into the market at prevailing prices in the event of a market fall. In effect, the
fund is purchasing insurance against market price increases and the cost of this
insurance is the option premium.

19.4 European put options

A European put option offers the holder the right, but not the obligation to sell
an asset at a future date for a given price. As with the call option, the holder of
the option (i.e. the long party) does not have to exercise this right. Therefore, the
buyer pays a premium to purchase the option.
19.4 European put options 367

Example 19.11 Consider an option to sell an asset for R100 in six months’
time. The premium for the option is R10. If the price of the asset is below
R100 in six months’ time, the option holder will exercise and receive the
difference between R100 and the asset price; clearly an asset price of less
than R90 will be required for a net profit, ignoring interest, trading costs etc.
If the price ends above R100, the option will expire worthless.

The profit profiles to purchasers and writers of put options are shown below.
Once again, the option purchaser’s loss is limited to the option premium, as is the
writer’s profit, and the purchaser has the chance of a significant gain while the
writer has the chance of an equally significant loss.
The profit profile from the put option purchaser’s perspective looks as follows:

? What is the maximum profit

the long party can make?
100
Was there a maximum profit
80 for the purchaser of a call?
60

40

20
Profit

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
-20

-40

-60

-80

-100
Price

Figure 19.6 Profit profile for the purchaser of a put option. In this
case the strike price is equal to R100 and the premium is R10, and so
the profit to the long party is max(100 − S T , 0) − 10, where S T is the
market price at the maturity date.
368 Chapter 19 Derivatives

And from the writer’s perspective the picture is the reverse:

U Note that the potential loss
to the writer of a put option
is limited to the strike price 100

less the premium. How does

80
that compare to the poten-
tial loss to the writer of a call 60

option? 40

20
Profit

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
-20

-40

-60

-80

-100
Price

Figure 19.7 Profit profile for the writer of a put option. In this case
the strike price is equal to R100 and the premium is R10, and so the
profit to the short party is 10 − max(100 − S T , 0), where S T is the market
price at the maturity date.
In general, given that the long party holding a European put has the right
to sell an asset in the future, they will not exercise this right unless the current
asset price is below the strike price (as any rational investor would not choose to
make a loss). Therefore, the payoff for the long party is max(K − S T , 0) and the
profit at maturity will be max(K − S T , 0) − P . In contrast, the short party will have
the reverse position, i.e. the payoff for the short party is − max(K − S T , 0) and the
profit at maturity will be − max(K − S T , 0) + P

Exercises for 19.4 European put options

Ex.19.3 Put options on Stock A are trading on an exchange. The put options
have a strike price of R124 and a 3 month term. The current premium
for the puts is R7. For each of the cases below, state the amount of
profit or loss made by the relevant party:
1. Jack writes a put on Stock A. After 3 months, Stock A is trading at
R90.
2. Gogo buys a put on Stock A. After 3 months, Stock A is trading at
R90.
3. Sani writes a put on Stock A. After 3 months, Stock A is trading at
R150.
4. Anton buys a put on Stock A. After 3 months, Stock A is trading at
R127.
19.5 Uses of derivatives 369

19.5 Uses of derivatives

Derivatives can be used to manage investment risk. They can be used to
1. remove or reduce investment risk, this is known as hedging;
2. increase investment risk in order to enhance returns, this is a form of specu-
lation.

Exercises for 19.5 Uses of derivatives

Ex.19.4 Recall Example 19.1 where Rob used futures to ensure he could sell his
wheat at a specific price.
1. What was Rob using derivatives for: to speculate or hedge?
2. What option strategy could Rob use to achieve a similar result? (I.e.
should he buy or sell a put or call?)
3. If derivatives can be used to remove investment risk, why doesn’t
everyone do that all the time?
4. What is the downside of using derivatives to enhance investment
returns?

Derivatives can also be used to change the notional composition of a portfolio of

assets without the investor having to actually buy or sell any assets. Consider the
following example.

Example 19.12 Wesley thinks that equity markets are about to fall, but
bonds will perform fantastically over the next 2 weeks. He has a portfolio of
R100 000 in equities, and no bonds. He could sell his equities, and buy some
bonds, but he would incur trading costs, and would need to do research to
decide which bonds to buy, and this would take time. He can instead sell
a 2 week futures contract on R100 000 worth of equities and buy a 2 week
futures contract on bonds.
In 2 weeks’ time, the futures contracts mature. If he is right and the equity
markets have fallen, he will make a profit from the equities futures contract
which will offset the losses on his equities portfolio, and he would not make
a loss on the fall in the equity prices. In addition, if the bond markets have
done very well, he will make a profit from holding the long futures contract
on the bonds.
In effect, by selling the futures contract on the equities and by buying the
futures contract on the bonds, he has reduced his exposure to the equities
market and gained exposure to the bond market. He has done this with-
out incurring trading costs, and while being able to change his portfolio’s
exposure immediately.
370 Chapter 19 Derivatives

19.6 Derivative Traders

Individuals and organizations trading in derivatives can be broadly divided into
three categories: hedgers, speculators, and arbitrageurs.

19.6.1 Hedgers
A hedger aims to reduce the risk from potential future market movements. Con-
sider the following examples.

Example 19.13 Using forwards/futures to hedge

John, a grape farmer in the Western Cape, is worried about the proceeds
he will obtain from his next harvest in 5 months’ time. To guarantee the
price received for the grapes, he can enter into a short forward agreement,
whereby he agrees to sell a certain amount of grapes to a vineyard for a
fixed price in 5 months’ time. By entering into the short forward, John has
reduced the market risk from the changes in grape prices.

Example 19.14 Using options to hedge

Siphumelele, an investor who owns SAB shares, is concerned that the SAB
share price might drop significantly in the next year due to rising barley
prices. She could enter into a short forward as in Example 19.13 and elim-
inate the market risk, however, Siphumelele would then lose the chance
to gain from any increase in the SAB share price. A suitable hedge here
U Carrying out such a strategy
might be to purchase an appropriate number of put options with a strike
is often referred to as buying
a protective put. price corresponding to Siphumelele’s desired protection level. This would
protect her from a fall in market prices, but still allow her to benefit from
any increases.

From these two examples, it is evident that hedgers use forwards and futures
to neutralize risk by fixing the price that the hedger will pay or receive in the
future, whereas options provide a means of insurance to protect against adverse
price movements.

19.6.2 Speculators
Speculation is the opposite of hedging. Speculators take bets on future market
movements. They aim to make profit by taking a view on the direction and/or the
extent to which they anticipate the market/asset prices will move. Speculators
will adopt a risky investment strategy that will be profitable if prices move in the
direction and/or to the extent they expect. The strategy will make losses if the
prices do not move as expected.
19.6 Derivative Traders 371

Futures and options both provide the investor with leverage. Leverage is
similar to gearing, and involves the investor gaining increased exposure to the
market for the same or less investment than an un-leveraged investor, e.g. by
buying a long future we obtain exposure to the underlying asset price with no
actual cost other than the margin payments. Due to this leverage, speculators
using futures contracts can make extremely large profits and losses for a very
small investment amount. Options provide similar leverage, but their downside
is capped for the long parties to the premium paid.

Example 19.15 Using futures to speculate Suppose that gold futures are
currently trading at $1 300 per ounce and that Suzan, a speculator, thinks
gold will be at least $1 350 per ounce at the maturity date of the future. Suzan
can buy a 1 000 ounce gold future at no cost and if the gold price is $1 350
per ounce at maturity, she would make a profit of $50 000. If however, she
gets it wrong and the price of gold drops to $1 200 per ounce, say, she would
lose $100 000.

So speculation is a risky business, particularly due to the leverage involved,

i.e. exposure to large gains or losses is possible for relatively small amounts of
capital investment.

19.6.3 Arbitrageurs
Arbitrageurs take offsetting positions in two or more instruments in order to make U Arbitrage refers to the op-
a certain profit from mis-pricing, i.e. they want to make a risk-free trading profit. portunity to make risk-free
profits. For example, an arbi-
These assets might be the same, but traded on different exchanges, or different,
trage opportunity may arise
but based on the same underlying asset.
if two assets with identical
Any arbitrage opportunities are likely to be small, especially after accounting future cashflows are trading
for transaction costs. However, when arbitrage opportunities do arise, they are at different prices. A trader
likely to be traded away very quickly by arbitrageurs who are looking for them. can make money from this
Hence, arbitrageurs help to keep market prices in line with each other, basically by arbitrage opportunity by
buying and selling the assets in different markets, they cause supply and demand buying at the cheaper price
in each market to vary until equilibrium is achieved. and simultaneously selling
at the more expensive price.
These days, markets are becoming more and more computerised, with the
result that arbitrage opportunities are becoming very rare. For example, it is
possible to build automated trading programs that find arbitrage opportunities
and immediately act on them.
372 Chapter 19 Derivatives

Exercises

Exercises for 19 Derivatives

Ex.19.5 Discuss the advantages and disadvantages of using futures rather than
forwards.
Ex.19.6 1. What is the difference between a long forward position and a short
forward position?
2. What is the difference between entering into a long forward con-
tract when the forward price is R50 and taking a long position in a
call option with a strike of R50?
T Being bullish means to be
optimistic, i.e. to expect good Ex.19.7 A speculator is bullish on BHP Billiton. Would he/she be prepared to
performance. The opposite is 1. buy a call option on BHP Billiton?
to be bearish.
2. write a call option of BHP Billiton?
3. buy a put option of BHP Billiton?
4. write a put option on BHP Billiton?
Give your reasoning.
Ex.19.8 A trader enters into a short forward contract on ¥100 million. The 3
month forward exchange rate is R18.77 per ¥. How much does the
trader gain or lose if the exchange rate at the end of the contract is
1. R19.02 per ¥?
2. R17.76 per ¥?
Ex.19.9 A South African company knows that it will need to buy equipment
valued at £1 million in 6 months time. The current exchange is R15 to £1.
The company has R15 million available now. How can it use a forward
to protect itself from the movement in exchange rate? How can it use a
call option to do the same? Compare the advantages/disadvantages of
each method.
Ex.19.10 1. Define what a European put option is and state the pay-off func-
tion.
2. Describe the difference between a long European put option and
a short forward.
Chapter 20

Investors and Investment

Strategy

Now that we are familiar with the main asset classes, it is time to consider how
an investor would go about deciding where to invest their money. In this section,
we discuss the factors that affect investment decision-making. The overall
“plan” that the investor has for their investment decisions is called the investment U Note that there is a differ-
ence between investment
strategy. This is different from the day-to-day investment decisions such as which
strategy and implementa-
stocks to buy and sell; it is rather the overall “big picture” of what the investor tion. The strategy is the plan
plans to do. and the implementation
The ultimate aim of the investment strategy is to ensure that the investor’s involves selecting (or hav-
objectives are met. ing someone else select on
your behalf) and trading the
actual assets.
Example 20.1
(a) List 4 different institutional investors, and consider what their investment
objectives may be.
(b) What potential investment objectives may an individual investor have?

Solution:
(a) The main institutional investors we generally talk about are life insurance
companies, general insurers, pension funds, and medical aids. But any large
institution with money to invest is an institutional investor — large corpora-
tions and banks fall into this category.
The main objective of insurance companies is to be able to pay out their
policyholders when claims are due. Similarly, the main objective of a pension
fund is to provide members with an adequate benefit on retirement.
(b) Individuals can have many different investment goals: someone may be saving
for a house, for a car, for university education for their children, for an overseas
holiday, for renovations, or for retirement.
T Note how these different goals
will require different types of
The investment objectives of an investor are strongly related to their liabilities. strategy — can you think what
type of assets may be suitable
for each investor?
T Actuarial risk is defined as
373
the risk of failing to meet
liabilities.
374 Chapter 20 Investors and Investment Strategy

20.1 Investment strategy considerations

Formulating an investment strategy means making the decisions as to how an
investor’s assets should best be invested in order to meet the investor’s objectives.
The following factors are generally considered when deciding on an investment
strategy:

1. The investor’s liabilities, especially their:

• nature,
• term,
• currency, and
• degree of uncertainty.
2. The investor’s risk appetite;
3. Diversification;
4. Expected returns of various asset classes;
5. Volatility of returns of these asset classes;
6. The level of free assets, i.e. the amount by which assets exceed liabilities;
7. Any restrictions on investment in various asset classes; and
8. Tax and transaction costs.

We now consider each of these factors in turn.

20.1.1 Liabilities
When considering liabilities, it is useful to consider their properties under four
separate headings:

Nature
T Remember that nominal li- Liabilities can be real or nominal in nature.
abilities are fixed in Rand
terms, and real liabilities in-
crease with inflation or some Example 20.2
other measure of economic Consider the following liabilities:
growth.
(a) a life insurance policy paying a fixed R1 million on death of the policyholder;
(b) a defined benefit pension fund which guarantees to pay a member a retirement
benefit of 2% of her pre-retirement salary for every year of service.
What is the nature of the liabilities in each case?

Solution:
(a) The policy is nominal in nature since the payment is not linked to inflation in
any explicit or implicit way.
(b) The pension fund has real liabilities since we would expect salaries to grow at
some rate equal to or exceeding inflation.
20.1 Investment strategy considerations 375

In general, we would use nominal assets, e.g. cash and fixed interest bonds, to ? So the life insurer in the ex-
back nominal liabilities, and we would use real assets, e.g. equities, property, and ample above would need to
set a strategy of investing in
index-linked bonds, to back real liabilities.
a combination of fixed in-
terest bonds and cash. Can
Term you think of any reasons why
the investments may devi-
The term of the liabilities refers to the expected length of time before they become
ate from this strategy and
due to be paid. include some equities?

Example 20.3
Are the following liabilities short- or long-term?
(a) A retirement benefit for a 25 year-old pension fund member.
(b) An endowment policy benefit with a 12 month term.
(c) The death benefit under a whole-life policy.

Solution:
(a) This is a long-term liability.
(b) This is clearly a short-term liability.
(c) This is uncertain as it depends on the policyholder’s age and health. For a
young healthy policyholder, the liability would be treated as a long-term liabil-
ity since our expectation would usually be that the policyholder’s death would
occur in many years’ time. For a very old policyholder, e.g. a policyholder aged
90 years old, the term of the liability may be considered to be a short.

The investment strategy should be consistent with the liabilities by term.

Long-term liabilities would tend to be supported by, for example, equities and
long-dated bonds, while short-term liabilities are more likely to be backed by ? What if you have a short-
term real liability — how can
cash and short-dated bonds.
you back that?

Currency

Investors doing business in a number of countries may well have liabilities de-
nominated in several different currencies. A multinational life insurer may, for
example, write life business with benefits (and premiums) expressed in Pounds
sterling, Euros, and US Dollars.
Typically such investors would aim to ensure that the liabilities in each cur-
rency are backed by appropriate assets in the same currency.

Degree of uncertainty

Investment strategy is also informed by how uncertain the amount and timing of T Recall that uncertainty can
liability payments are. be related to whether some-
thing is paid at all, when it is
paid, and how much will be
paid!
376 Chapter 20 Investors and Investment Strategy

Exercises for 20.1.1 Liabilities

Ex.20.1 Consider the following liabilities. Describe whether there is uncertainty

about whether there will be a payment, the timing of the payment, and
the amount of the payment:
1. An employer’s liability policy sold by a general insurer.
2. A term life assurance policy with a 5-year term and a R1 million
sum assured which increases with inflation.
3. A retirement benefit from a defined benefit plan with a 2% accrual
rate and a normal retirement age of 65.
4. The payments made from a medical aid scheme.

Uncertainty can be reduced by risk pooling, so large books of policies with

many policyholders tend to have less uncertainty than small books of policies.
As a general rule, the more uncertain the short-term liabilities, the greater the
liquidity needed in the asset portfolio. A medical scheme would therefore need to
have a larger proportion of their investments in cash due to the short-term nature
and uncertainty of their liabilities.

20.1.2 Risk appetite

Liabilities are not the only factor in setting investment strategy. Investors have
different levels of tolerance for risk, and that is also reflected in how they set their
investment strategy. The risk appetite of an investor refers to their tolerance for
risk in the pursuit of higher returns.
For individual investors, risk appetite is to some extent driven by their person-
ality — some people are simply not prepared to accept investment risks. Other
factors that can contribute to a person’s risk tolerance can be their level of finan-
cial ability, their level of wealth, and the time and effort they are prepared to invest
in their portfolio.

Exercises for 20.1.2 Risk appetite

Ex.20.2 For each of the investors below, think about what their risk appetite
with respect to their own investments may be:
1. a bank clerk,
U While it is easy to judge peo-
ple by their appearance, ap- 2. the CEO of a large company,
pearances can be deceptive — 3. a professional poker player, and
the true risk tolerance of a 4. your mom.
person is something you have
to find out from them person- For an institution, the risk appetite will generally depend on the type of institution,
ally! any legal or statutory controls under which it operates, and any self-imposed
constraints.
20.1 Investment strategy considerations 377

20.1.3 Diversification
Diversifying the investment portfolio across asset classes generally gives rise to
a far more acceptable risk-return relationship than an “all eggs in one basket”
approach. Diversification is similar to risk pooling — by avoiding a concentration
of investments in a particular asset or asset class, the risk is reduced. That is
because some of the random variations in returns cancel out between different
investments.
Diversification can be at the asset class level or at the individual holding level.
At the asset class level, diversification means holding a mix of equities, bonds,
property, cash, etc. At the individual holding level, diversification means avoiding U Diversification could also
be avoiding a concentration
concentration in one share or bond. Regulations often include a requirement to
in any particular industry,
diversify in order to reduce risk. geographical area, currency,
and many other things.
20.1.4 Expected returns and volatility of returns of asset classes
In order to make informed investment strategy decisions, it is obviously important
to understand the expected returns of the asset classes under consideration, as
well as the variability of these returns. A study of historic performance as well
as interpretation of the current economic environment with a view to projecting U A person who analyses the
performance of securities
expected future performance is therefore essential.
is called an investment an-
A very important point here is that the period over which one examines the alyst. They generally spe-
variability of returns should be consistent with the term of the liabilities under cialise in a certain sector
consideration. For example, over the short-term, equities can have extremely of the market; for example,
volatile returns compared to bonds and cash, and a short-term investor may not mining shares, or long-term
be willing to accept this huge variability for the additional expected return which bonds.
equities deliver. Over longer periods, however, it may in fact be far riskier not to
be invested in equities. So a long-term investor would assess the variability of
equities differently from a short-term investor. T Recall for example that
bonds have never outper-
formed equities over any 20-
20.1.5 Level of free assets year period in South African
investment market history.
The term free assets is used loosely to refer to the amount by which the value of a
company’s assets exceeds the value of their liabilities.

Example 20.4 Life insurance company A has completed a reserving exercise

and has calculated that it needs to hold reserves of R155.8 million to back its
policies. The insurer’s total assets are equal to R165.9 million. The insurer’s
free assets are therefore R10.1 million.

Insurers and other institutional investors are required to calculate their liabili- T This is called reserving in
life insurance and actuarial
ties regularly. They will therefore be aware of their level of free assets.
valuation in pension funds.

Example 20.5
378 Chapter 20 Investors and Investment Strategy

Insurance company B has calculated its free assets as R8.5 million. B’s liabilities are
R45 million. Compared to insurance company A in the previous example, which
company has a higher level of free assets?

Solution:
B’s free assets are 8.5/45 = 18.9% of its liabilities. A’s free assets are 10.1/155.8 = 6.5%
of its liabilities. B is comparatively better funded than A.

The lower the level of free assets, the more constrained the investment strategy
of the institution is likely to be. When there are no spare assets, it is necessary to
T Minimum solvency levels limit investment risk in order to minimise the risk of failing to meet minimum
are set by the regulator to
solvency levels in the future. By contrast, an institution which has significant
protect policyholders. They
set out the minimum re- free assets (i.e. where the value of the assets significantly exceeds the value of
serves an insurer must hold. the liabilities) has much greater freedom in setting its investment strategy, and
greater freedom to mismatch the liabilities. This is because the free assets provide
a buffer which can absorb losses that might arise from the additional risk of
mismatching.

Example 20.6
Both of insurers A and B have liabilities that are mainly nominal as they have both
sold a lot of policies with fixed benefits. What is the ideal investment strategy for
such liabilities? Why might these insurers want to deviate from such a matched
strategy? Which insurer is more likely to be able to deviate in such a way?

Solution:
Both of the insurers should invest in a combination of fixed-interest bonds and
cash to match their liabilities. However, over the long-term, it is likely that they
will achieve higher returns from investing in real assets, such as equities. However,
the risk is that equities will not behave in the same way as the liabilities over time
because of their different nature. For example, the liability values are likely to be
quite stable from year to year, while equity values can swing wildly. This could
potentially leave the insurer underfunded.
Insurer B has more free assets, so they can afford to take more risk when setting
their investment strategy. For example, say insurer B invests 25% of their assets in
equities, and the remainder in bonds and cash. B’s total assets are 45 + 8.5 = 53.5
million Rand. So 0.25 × 53.5 = 13.4 million Rand would be invested in equities. If
the equity market falls by 50% (and nothing happens to the bond market), B’s total
assets will fall to 53.5 − 0.5 × 13.4 = 46.8 million Rand. In such a case, insurer B
would still be more than 100% funded. In contrast, if A had done the same thing, it
would have invested 0.25 × 165.9 = 41.5 million Rand in equities. The same market
fall would have resulted in a fall in A’s total assets to 165.9−0.5×41.5 = 145.2 million
Rand. But A’s liabilities are R155.8 million, and A’s funding level would have reduced
to unacceptable levels as a result of this investment and the equity market fall.
20.2 Institutional investors 379

20.1.6 Investment restrictions

Many institutions face regulatory restrictions on the assets in which they can
invest.
U In South Africa, Regula-
tion 28 of the Pension
Example 20.7 Funds Act sets out in de-
Pension funds are generally prohibited from investing in the shares of the tail what a retirement fund
sponsoring employer. The reason for this is that it creates a concentration may and may not invest in.
of risk for the employees who are also members of the fund. For example, Read more about it here:
[Link]
if the employer runs into financial difficulties, this will affect not only the
security of their jobs but also the value of their pension benefits.

Apart from regulatory restrictions, many institutions impose limits on them-

selves for reasons of prudence and or ethical considerations. This may take the
form of, for example, a stipulated maximum of 75% of assets in domestic equities,
or a restriction on investing in shares in weapons or tobacco manufacturers.

20.1.7 Tax and transaction costs

Finally, although we have ignored costs in our discussion of the investment world
so far, in practice it is important to consider the costs that might be attracted by
a particular strategy. For example, if income is taxed more heavily than capital
gains, this may lead investors (all else being equal) to prefer investments which
deliver more of their return in the form of capital gains rather than income.
The transaction costs of trading also dictate that institutions should set their
investment strategies for the long-term and not revise their asset allocations too
frequently, which would give rise to significant trading costs.

20.2 Institutional investors

In Module 3, we introduced the major institutional investors, i.e. the different
types of insurers and pension funds. In this section, we will consider how these
institutional investors may go about setting their investment strategies.

20.2.1 Life insurers

Life insurers sell policies which pay out on death, on survival, or both. They also
sell annuity type products.

Exercises for 20.2.1 Life insurers

Ex.20.3 Referring to Module 3, list the different types of policies that a life
insurer may sell.

The nature of the liabilities of a life insurer will depend on the types of prod-
ucts it sells and the mix of different products.
380 Chapter 20 Investors and Investment Strategy

Nature

Whether the liabilities of a life insurance company are real or nominal depends
on how the sum assureds of its policies escalate.

Exercises for 20.2.1 Life insurers

Ex.20.4 List the different ways that a sum assured can increase over time. For
each increase method, state whether that makes the liabilities real or
nominal, and if they are real, suggest what the increase may be linked
to.

Term

The liabilities of a life insurer will generally tend to be long-term, because most of
the policies will have terms of many years. However, as policies move closer to
maturity, their term shortens, so at any time there will be some policies which
have very short terms. Another example of a short-term liability is an old book
of policies that are no longer sold. Over time, this book will have a shorter and
shorter term, until all of the policies have matured.

Currency

The currency of the liabilities depends on the regions in which the insurer oper-
ates.

Degree of uncertainty

Life insurance policies have sum assureds which are either fixed, or increase
either at a fixed rate or with a pre-defined index or asset return. The size of
the sum assured is therefore quite predictable, especially compared to general
insurance or medical aid companies, where claims are highly variable. The timing
of the claim, however, is still unknown. Death claims do follow fairly stable
? What might cause total an- distributions and when an insurer has a sufficient number of policyholders for
nual claims from deaths to
risk pooling, the total annual claims can be predicted with reasonable accuracy.
be unexpectedly large?
Some insurance policies are certain to pay out and others may or may not pay
? Which types of life insurance out.
policies may never pay out a
claim? Exercises for 20.2.1 Life insurers

Ex.20.5 A life insurer has two types of policies: term assurance policies and
endowment assurance policies. Which type of policy has more uncer-
tainty in their claims and why?

Below, we consider 3 case studies of different types of policies and how they
might be matched by suitable investments.
20.2 Institutional investors 381

Case study 1: Whole-life cover with fixed sum assured

Consider a portfolio of whole-life policies, each with a fixed benefit of R1 million,
written on a group of 30 year-old policyholders.
In order to determine the investment strategy, the actuary would project the future
outgoes of this portfolio of liabilities based on assumptions about future interest
rates and mortality rates. Typically, we would expect relatively few deaths from this
group of lives in the early years, with the bulk of the liability outgo being concen-
trated in the much longer term, e.g. after 20 or more years. A suitable investment
strategy might therefore target a portfolio with the following characteristics:
• A small portion invested in cash to meet short-term expected liability outgoes
and to provide a liquidity buffer for unexpected short-term payments.
• The majority of the assets are likely to be invested in fixed-interest bonds of
suitable term, concentrated toward the long-term given the expected liability
outgo profile; these bonds will clearly provide the best match for the bulk of
the liabilities.
• There may be some investment in equities, both to support the very long-term
liabilities (there are very few bonds with terms in excess of 30 years) and to
provide diversification. In addition, if there are significant levels of free assets,
more may be invested in equities to pursue higher returns in the knowledge
that the free assets provide a cushion against short-term adverse experience.

Case study 2: Unit-linked endowment assurance

Recall that unit-linked cover is distinguished from conventional cover in that it
provides a benefit linked to the value of a pool of assets. In many of these policies,
the broad investment strategy can be selected by the policyholder from a range
of different options offered by the insurer. For unit-linked policies, the insurer’s
investment strategy would be constrained by the broad investment strategy associ-
ated with the units. However, the investment strategy for the unit-linked product
may be more free than the strategy an insurer would have otherwise adopted.

Example 20.8 Consider a unit-linked endowment policy with three invest-

ment options:
(a) 100% equities,
(b) 50% bonds and 50% equities, or
(c) 100% bonds.
Darren buys one of these policies and chooses the 100% equity investment
option. The insurer invests the assets backing this policy in a 100% equity
portfolio, even though it may not have chosen such a risky investment for
its own assets. The reason for this is because the insurer has transferred the
risk to Darren, e.g. should equity prices crash, Darren’s policy will simply
pay a lower benefit value.
382 Chapter 20 Investors and Investment Strategy

Case study 2 cont.:

In theory, the insurer could invest all of the assets associated with a unit-linked
product in Venezuelan tin futures, but in practice the investment strategy is often
constrained by a concept known as “policyholders’ reasonable benefit expecta-
tions”. This term describes a standard to which life insurers are expected to aspire,
and requires them to be conscious of the level of benefits which a policyholder
entering into the contract (who may have very limited investment knowledge and
experience) would reasonably expect. As such, insurers are likely to limit the in-
vestment options available to reduce the chances of extreme performance. All the
same, it is likely that the level of investment in equities in unit-linked policies will
be greater than for equivalent conventional policies.

Case study 3: Inflation-linked annuity policies

Consider a portfolio of annuities each of which pays R5 000 per month to a 65
year-old policyholder, with the annuity payments increasing annually in line with
the increase in the Consumer Price Index.
The closest match to the portfolio of annuity benefits would be a portfolio of assets
invested mostly in index-linked bonds. The index-linked bonds would provide a
series of regular payments guaranteed in real terms and these would correspond
to the profile of the liabilities. The portfolio may also consist of some short-dated
? Why would the portfolio also conventional fixed interest bonds and cash, as well as a small holding of equities to
contain short-term nominal back the longest-dated liabilities.
assets?

20.2.2 General (short-term) insurers

As far as investment needs are concerned, the two major differences between
general insurance and life insurance are that general insurance liabilities are
both shorter in term and more uncertain in nature than the liabilities of a life
insurance company. As a consequence, general insurers will tend to hold a greater
proportion of their assets in cash and short-dated bonds.

Exercises for 20.2.2 General (short-term) insurers

Ex.20.6 Describe the features of the following general insurance liabilities tak-
ing into account their nature, term, currency, and degree of uncer-
tainty:
1. employer’s liability insurance, and
2. comprehensive car insurance.

20.2.3 Retirement funds

T Recall that retirement funds Pension funds are governed by a board of trustees who decide on important
are trusts, i.e. they are not-
matters relating to the operation of the fund, including the investment strategy.
for-profit legal entities sepa-
rate from the business of the
sponsor.
20.2 Institutional investors 383

Defined benefit and defined contribution funds tend to have different investment
strategies because of who carries the investment risk.

Exercises for 20.2.3 Retirement funds

Ex.20.7 Who carries the investment risk in a DC fund? Who carries the invest-
ment risk in a DB fund?
Ex.20.8 What is the nature of the liabilities of a DB fund? What about a DC
fund?

Defined benefit pension liabilities are long-term and real in nature. The invest-
ment strategy of such funds therefore tends to include relatively high proportions
of equities, typically 50% to 75% depending on the age profile of the fund and the
consequent term of the liabilities.
Defined contribution funds, on the other hand, are similar to the unit-linked
life insurance liabilities discussed above, in that the trustees and sponsoring
employer do not face the direct adverse consequences of poor investment perfor-
mance. Nevertheless, in the same way that the managers of life offices have to
consider policyholders’ reasonable benefit expectations, pension fund trustees
have a moral and fiduciary duty to ensure to the greatest degree possible that the
investment strategy of the fund will maximise the expected level of retirement
benefits subject to an acceptable level of risk. Trustees do this by, for example,
choosing an investment strategy with an appropriate level of investment in real
long-term assets such as equities, or allowing members to choose the investment
strategy for their own retirement assets, possibly with some default strategy which
the trustees believe to be optimal.

Exercises for 20.2.3 Retirement funds

Ex.20.9 A member of a DC pension fund is 30 years old now and plans to retire
at age 65. They are currently working as a junior consultant, but in the
long-term, they plan to rise into upper management of their company.
What is a suitable investment strategy for this member? What, other
than matching the liabilities, may affect the investment strategy for
this member?

20.2.4 Medical schemes

About 8 million South Africans are covered for private healthcare costs by mem-
bership of what are commonly known as medical aids. In legal terms, they are
members of a medical scheme, which like a pension fund is a separate not-for-
profit legal entity, where the assets of the fund effectively belong to the mem-
bers. In return for a monthly premium, medical schemes cover certain private
healthcare costs, ranging from day-to-day costs through to major hospitalisation
expenses.
384 Chapter 20 Investors and Investment Strategy

Since medical scheme premiums are paid on a monthly basis providing cover
for that month, the liabilities are very short-term in nature and also highly un-
certain. Medical scheme asset portfolios would therefore tend to be far more
concentrated in cash and short-dated bonds than, for example, those of a pension
fund.

20.3 Conclusion
The above discussion is intended to give a brief overview of the generic consider-
ations when formulating an investment strategy, as well as the main institutional
investors and the specific investment considerations for each of them. You should
now be able to put together your understanding of the main institutional investors
and the different asset classes to be able to design a suitable investment strategy
for any investor.
Designing the strategy is only one step of the investment process. The imple-
mentation of the strategy is the next step: the selection of suitable securities to
carry out the strategy. This is discussed in the next chapter on asset management.
Chapter 21

Asset and Portfolio Management

In the previous chapter, we discussed how an investor will decide on a suitable

investment strategy. Once the strategy is set, the investor still faces a number of
important decisions, for example:
• What specific securities should they buy, to start with?
• When should they trade securities, i.e. when should they sell and buy differ-
ent shares, bonds, and other investments?
• To what extent may they temporarily depart from the investment strategy?
• How do they evaluate if the strategy and implementation are working for
them?
Portfolio management comprises of all the processes involved in the creation
and maintenance of an investment portfolio. Portfolio management can be
performed by the investor for themselves, or by an asset manager on behalf of the
investor. The purpose of this chapter is to look at some of the principal techniques
in portfolio management.

21.1 Asset allocation

Asset allocation focuses on determining the combination asset classes that best U This process is related to
developing the investment
meets the investor’s objectives and provides the optimal risk and expected return
strategy.
profile for the investor. There are two types of asset allocation decisions, namely
strategic and tactical asset allocation.

21.1.1 Strategic asset allocation

The aim of the strategic asset allocation decision is to derive the long-term asset T E.g. an investor may specify
allocation weights used in the investor’s portfolio — these weights may be fixed that they want to hold 75%
in equities, or that they can
or specified as ranges. This involves identifying the proportions of the portfolio
hold 70%–80% in equities.
that will be invested in each asset class. Generally, these strategic weights are not
changed over time, unless there is a change in the investment mandate or the
investor’s risk profile or objectives. However, when there is a change in the market

385
386 Chapter 21 Asset and Portfolio Management

value of the underlying assets, the investor may have to adjust the portfolio so as
to maintain the desired fixed-percentage allocation.

Example 21.1 Reginald has set his investment strategy as 50% equities and
50% bonds. He decides on a portfolio of equities and bonds to purchase,
and invests R1 million in the market — R500 000 in each asset class.
After one week, the market value of his equities has moved up to R530 000
and bonds have dropped a little to R495 000. He now holds 51.7% in equities,
which is more than he strategically intended. If he wanted to realign his
portfolio, how much of the equities would he have to sell now to buy bonds?
Reginald is likely to be flexible about his strategy. For example, he could
choose to only rebalance once a year, or only rebalance if the proportion in
one asset class moved more than 10% away from the agreed proportion.

Investors using a strategic asset allocation approach would allocate a fixed

percentage of the portfolio value to each of the different asset classes. For example,
the asset allocation in the portfolio might be:

Asset class Allocation

Domestic assets 80%
Equities 40%
Money market instruments 5%
Government bonds 12.5%
Corporate bonds 7.5%
Property 15%
Foreign assets 20%
Equities 17.5%
Money market instruments 2.5%

Once this allocation has been decided, the investor, or the asset manager, will
analyse each of the sectors selected to decide on which securities to buy for the
portfolio.

21.1.2 Tactical asset allocation

The aim of tactical asset allocation is to allow the investor the ability to temporarily
move away from the strategic asset allocation. This decision will be driven by
short-term market movements. For example, the investor may make short-term
tactical deviations away from the long-term strategic asset allocation in order to
take advantage of the temporary under- or over-pricing of particular assets.

Exercises for 21.1.2 Tactical asset allocation

Ex.21.1 An investor has the strategic goal of investing 100% in South African
equities. However, they may deviate from this allocation from time to
21.1 Asset allocation 387

time. In which of the following situations would the investor want to

temporarily reduce their equity exposure? Why?
• Bond prices are at a historical low.
• International equity markets are crashing.
• Money markets are delivering very high interest while equities are
expected to have an average year.
• Equity markets have just crashed.
• Property is priced at record highs.

When an investor uses tactical asset allocation, the investor’s risk-return pref-
erences and goals remain unchanged. The short-term changes in asset allocation
are just used to improve performance or protect from a fall in the market.

Example 21.2
Asset managers offer funds with different levels of tactical asset allocation.
For example, the Allan Gray Absolute fund ([Link]/e3xi3v) is completely
tactical and allows the managers of the fund to have any asset allocation,
from 0% in equities up to 75% in equities, which is the maximum allowed by
Regulation 28 of the Pension Funds Act.
U The portfolio complies with
Another fund aimed at retirement fund investors is the Investec Balanced Regulation 28 because it is
Fund ([Link] This fund allows investment of between 50% designed as an investment
and 75% of the portfolio in equities, and so it is somewhat flexible, but not for pension funds which have
as flexible as the Allan Gray Absolute fund. to invest in Reg 28 compliant
portfolios.

There are various factors to consider before making a tactical asset switch.
Firstly, the additional expected returns to be made by doing the switch must
be weighed up against the additional risk of doing so, and also the additional
expenses of making the switch must be considered. If the additional risk and
expenses outweigh the potential gain, then it would not be in the investor’s
best interests to make the switch. Furthermore, one needs to consider the
investment mandate and legal restrictions on the fund to determine if there are T Regulation 28 is an exam-
ple of such a restriction for
any constraints on the changes that can be made to the portfolio. Finally, one
pension funds.
needs to consider the problems of switching a large portfolio of assets.

Example 21.3
Discuss some of the problems involved in making a large tactical switch to the
underlying assets in a portfolio.

Solution: Some of the main problems when making large changes to the asset
allocation are:
• the dealing costs involved, which could outweigh the profits made on the
switch;
388 Chapter 21 Asset and Portfolio Management

• the possibility of market prices shifting adversely (both on sale of assets cur-
rently owned, and on the purchase of new assets) — some investors are so
large that when they sell a certain share, the price of the share will actually fall
due to the large number of shares they are selling, for example;
• the time needed to make the change and the difficulty of making sure that the
timing of deals is advantageous; and
• the possibility of the crystallisation of capital gains leading to a tax liability.
T Capital gains tax is only
payable when a gain is crys-
tallised, i.e. when an asset is One way to reduce the problems involved in tactical switches is to use deriva-
sold for more than what was tives.
bought for. So a change in the
asset allocation in a portfolio
Exercises for 21.1.2 Tactical asset allocation
can lead to a CGT liability.
Ex.21.2 An investor has a strategic allocation of 80% equities, and 20% bonds
and cash. She wants to make a tactical switch to reduce her equity allo-
cation and increase exposure to bonds for 3 months. What derivative
strategies could she use to achieve this? What are the advantages and
disadvantages of each strategy?

21.2 Portfolio construction approaches

Two major approaches to portfolio construction and management are the top-
down and bottom-up approaches. In brief, the top-down approach makes use
of macroeconomic or overall market factors to drive portfolio construction. The
bottom-up approach starts from the micro end by looking at key items related to
individual investments.

21.2.1 Top-down approach

The top-down approach to constructing and managing a portfolio starts by decid-
ing on the “big picture” first and then filling in more and more of the detail. So
the first step is a decision on the overall strategic asset allocation. Then, within
each asset class, an analysis is made of how to distribute the available funds
between the different sectors, e.g. how to invest in the different industrial sectors
for equities. Finally, individual assets are selected for each sector and purchased.
The process followed in a top-down investment strategy will generally involve
the following steps:
1. Determine the strategic allocation of assets between countries and between
the main asset categories.
2. Decide on the short-term tactical split of investments, again between coun-
tries and between the main asset categories based on a shorter-term view of
global economic and investment issues.
3. Given the chosen tactical asset allocation, decide upon the sector split within
each asset category.
21.2 Portfolio construction approaches 389

4. Finally, within each sector decide which particular stocks are “best value”.

Example 21.4 An asset manager is developing a new product aimed at

investors with a moderate risk tolerance and a 5–10 year time horizon. Fol-
lowing a top-down approach, the manager proceeds as follows:
(a) The strategic asset allocation is determined to be as follows
Asset Class Allocation
Domestic Assets 70%
Equities 30%
Money Market Instruments 5%
Bonds 10%
Property 25%
Foreign Assets 30%
Equities 20%
Bonds 10%
(b) The asset management team then considers the current market situa-
tion. The outlook they agree on is that property markets are currently
weak and not expected to improve for the next year or so, while equities
are expected to deliver great performance. The international markets
are also expected to outperform local markets in the near future. A tacti-
cal decision is made to reduce exposure to property and replace it with
more equity in the short term, and to increase exposure to international
markets as well. The tactical allocation is then:
Asset Class Allocation
Domestic Assets 60%
Equities 35%
Money Market Instruments 5%
Bonds 10%
Property 10%
Foreign Assets 40%
Equities 30%
Bonds 10%
(c) The asset manager can now delve deeper into each of the above asset
classes and decide on the allocation between different sectors of the
market. For example, for international equities, the country and the
industry sector are both important decisions. So the international
equity allocation may then be split up as follows:
390 Chapter 21 Asset and Portfolio Management

International equity sector Allocation

International equities 30%
US equities 10%
US Industrials 2.5%
US Retail 1%
US Financials 5%
US Manufacturing 1.5%
Asian equities 5%
Asia - Financials 3%
Asia - Manufacturing 2%
...
(d) The final step is to now find suitable stocks for each category. So if 3%
of the full portfolio needs to be invested in Asian financial companies,
the asset manager must now analyse Asian financial stocks to uncover
some which are well priced and have good potential for increase in
? Just one Asian financial stock
value over time.
will probably not be enough –
can you think why?

Exercises for 21.2.1 Top-down approach

Ex.21.3 Discuss the type of information that would be important to consider

when determining the strategic asset allocation.
Ex.21.4 Discuss the type of information that would be important to consider
when determining the tactical asset allocation.
Ex.21.5 Explain briefly the distinction between strategic and tactical invest-
ment decisions.

21.2.2 Bottom-up approach

The bottom-up approach to constructing and managing a portfolio is based upon
identifying the best value individual investments, irrespective of their geographic
or sectoral spread. These investments are then combined to form the investment
portfolio. The bottom-up approach follows the reverse approach to that of the
top-down approach.
A number of methods can be used for asset allocation and individual stock
selection. These include:
• fundamental analysis is the analysis of a company’s share (or investment
asset’s) value and potential for future profit and dividends, based on ac-
counting and economic information.
• quantitative techniques are mathematical techniques that can be used to
aid stock and sector selection. Essentially these incorporate asset pricing
U Read more on the CAPM here:
models such as the Capital Asset Pricing Model (CAPM), which can be used
[Link]
Capital_asset_pricing_model. to identify mispriced assets and hence trading opportunities.
21.3 Active and passive portfolio management 391

• technical analysis aims to predict future market movements by analysing

past market data. The methods used include chartism, mechanical trading
rules, and relative strength analysis.
Asset managers and investors following a purely bottom-up approach have
little control over the strategic asset allocation of their portfolio. Their portfolio
is a random, and changing, combination of what they consider the best value
securities in the market. However, it is difficult to use such a portfolio to match
a specific investor’s investment strategy. One way to overcome this is to create
bottom-up portfolios that specialise in specific asset classes, e.g. equities or prop-
erty. The investor can then choose to purchase a combination of such portfolios
aligned with their strategy.
Some asset managers also follow a combination top-down/bottom-up ap-
proach, where they follow, say, steps 1 and 2 of the top-down process, but then
use bottom-up methods to find the shares for each asset class.

21.3 Active and passive portfolio management

When deciding on the share selection within each asset class, investors can choose
to select shares actively (using top-down or bottom-up methods), or they can
choose to manage the portfolio passively.

21.3.1 Active portfolio management

Active managers believe that they can outperform the market. They basically T By “the market” they mean
believe that the prices of some securities do not accurately allow for the true the total of all the shares
or bonds traded on an ex-
future performance of those securities, and that they can therefore pick shares
change. The market is often
and bonds which are underpriced.
represented by an index.

Example 21.5
A corporate bond of FigWit Corporation is trading at R98.50 per R100 nominal.
The bond is a 10-year bond paying a 7% coupon semi-annually in arrear, and is
redeemable at par.
(a) At what interest rate is the market valuing the bond?
(b) You have assessed Figwit Corporation and you believe that their 10-year bond
should be valued at an interest rate of 6.5% p.a. What price would you put on
this bond? Would you buy it or not based on this assessment?
(c) How would a passive investor view this bond?

Solution:
(a) The equation of value is

98.50 = 3.5a 20 j + 100v 20

j ,

where j is the semi-annual interest rate that the market is using to value the
bond and R3.50 is the semi-annual coupon paid on R100 of the bond.
392 Chapter 21 Asset and Portfolio Management

By using linear interpolation, or Newton’s method ([Link] or

Goal Seek in Excel, we can find that j = 3.607% and therefore the annual
effective interest rate i = 7.343%.
(b) 6.5% per annum is equivalent to 3.20% per half year. The price you put on the
bond is therefore
20
Price = 3.5a 20 3.20% + 100v 3.20% = R104.40.

Based on your assessment, the bond is underpriced by the market and if you
buy it now, you will get it at a bargain. You should buy.
(c) A passive investor believes that the market is efficient, and therefore, that the
interest rate implied by the market price is the correct one. The investor would
therefore not value this bond at 6.5% interest, but rather assume the bond is
fairly priced.

Active managers spend a lot of time and resources analysing securities to

discover opportunities for buying and selling. As a result, active management
U For example, the Coro- tends to require a large team of analysts and experts. To be able to finance this,
nation Balanced Fund
active managers charge fairly high fees for their services.
([Link]/6BifRk) charges in-
vestors 1.25% of the fund There are many different types of active managers, for example:
value per annum for the ac-
tive management it provides. Growth and value
This charge is fairly represen-
tative of the charges levied Growth managers focus on growth shares. These are shares of generally well-
by active managers in South regarded companies which perform well, tend to reinvest some or all of their
Africa. profits in the company, and are expected to experience a lot of capital growth.
Value managers, on the other hand, buy shares which are for some reason
U Allan Gray takes a very
strong value stance in their
undervalued by the market. Investors who purchase value shares are counting
investments, although on on these shares gaining value, not because the underlying company will grow
the whole, value is a more or change, but because the market will eventually realise that the shares were
popular style than growth underpriced and correct this view.
in South Africa — can you Some managers try and use both approaches. For example, rotational port-
think why? folio managers move between value and growth depending on which style is
believed to be attractive at any particular point in time. In general, such investors
will be in growth stocks when the market is rising and in value stocks when the
market is falling.

Momentum managers

Momentum portfolio managers purchase those stocks which have recently risen
significantly in price on the belief that they will continue to rise owing to an
upward shift in the demand for them. They also sell stocks which have recently
fallen sharply because they believe they will fall more. These managers aim to
take advantage of momentum effects in investment markets.
21.3 Active and passive portfolio management 393

Contrarian managers

Contrarian portfolio managers do the opposite to what most other investors are
doing in the market in the belief that investors tend to overreact to news. The
rationale for this approach is that whilst over the long-term most shares will give
an average performance, in the short-term markets tend to overreact to good and
bad news.

21.3.2 Passive portfolio management

Passive portfolio management investors believe that investment markets are T An efficient market is one in
relatively efficient. A passive manager therefore sees no need to buy a specific which asset prices accurately
reflect all available and rele-
share over any other.
vant information at all times.

Example 21.6 Say company A has excellent growth prospects. It is well

managed, efficient, and has a good strategy for expanding its market share.
Company B, meanwhile, has been struggling for a while and has made very
little profit for the last few years. They are selling unpopular products and
generally not doing well.
A passive manager would be equally happy buying shares in either company.
That is because he believes that the market knows company A has great
prospects, and therefore investors would have bought the shares of A, which
would have increased the share price of A to the point where the future
profitability is already reflected in the price. Similarly, for company B, the
passive investor believes that the share price of B should be very low, because
the market has priced in all the problems. So the price compensates for the
poor prospects and there is no need to prefer the one share over the other.

Exercises for 21.3.2 Passive portfolio management

Ex.21.6 Can you tell which of the two shares in the example above an active
investor would have bought?

So passive investors will generally buy all the shares in the market, or rather, ? Why is it not practical to try
because of practicality, a portfolio that is very similar to the market, i.e. a proxy and hold a perfect represen-
for the market. A good way to proxy the market is to buy the shares that make up tation of the market?
a large proportion of the market, e.g. buying shares in the companies with the
largest market capitalisations.
Passive investment has the following characteristics: T Remember, the market cap-
italisation of a company is
• it is easy, because there are fewer decisions to make and much less research; the total value of the com-
• it involves limited trading as the investor tends to hold shares for a while; pany’s shares.

• it is cheap compared to active investment, because there is less expertise T What would happen if all
required and much less trading; investors decided to become
passive investors?
394 Chapter 21 Asset and Portfolio Management

• it has built-in diversification, because the investor attempts to hold a portfo-

lio that represents the market.
However, passive investment aims to keep up with the market rather than outper-
form it, while active investment aims to deliver returns higher than the market.
There are supporters for both styles of investment, and research is not conclusive
as to which is more effective in the long run.

21.4 Choosing an investment style and implementation

The various styles we have been discussing can often be used in combination,
and make up an investor’s or asset manager’s “philosophy” or approach.
For example, an active manager can follow a top-down approach with a
growth focus. Or a bottom-up manager can have a value bias, and so forth.
U Individual investors can Investors who do not want to do their own investing but instead choose to
select a unit trust, while in-
appoint an asset manager to work for them will choose a manager whose philoso-
stitutional investors may
appoint an investment man- phy reflects their own beliefs about the market. So an investor who believes in
ager to create a specific port- the efficient market is likely to choose a passive manager, etc.
folio for them. Small institu-
tional investors who cannot
afford such bespoke man- 21.5 Monitoring and revision of the portfolio
agement can just use one of
the existing products of an It is essential to continually monitor the performance of a portfolio and if needed,
asset manager, such as a unit perform some portfolio revisions by selling certain assets in the portfolio and
trust. replacing them by purchasing new ones. The primary reasons for revising invest-
ment portfolios include:
• Changes in the economy - certain industries and companies become either
less or more attractive as investments over time;
? Which of these reasons are
applicable to a passive portfo- • Changes in investment objectives - over time there may be a change in the
lio? investor’s objectives and as a result the current portfolio may no longer be
optimal;
• Asset mix - over time some asset classes will outperform others, with the
result that the portfolio is more heavily weighted towards those classes.
• Diversification - over time there will be changes in the risk-return character-
istics of an investment and this may affect the current level of diversification
in the portfolio;
• Asset manager performance - if an investor is using an asset manager, the
investor will also review that manager’s performance to evaluate whether
the asset manager continues to deliver expert skills and be worth the fee
they are paid.

21.5.1 Monitoring the market

The need to monitor changes in the market is obvious. Investment decisions are
made in dynamic investment environment, where changes occur all the time.
21.5 Monitoring and revision of the portfolio 395

The key macroeconomic indicators (such as GDP growth, inflation rate, interest
rates), as well as the new information about industries and companies should be
observed by investors on a regular basis, because these changes can influence the
returns and risk of the investments in their portfolios. It is important to identify
the major changes in the investment environment and to assess whether these
changes will negatively influence the investor’s currently held portfolio. If it so,
the investor must take actions to rebalance his/her portfolio. An analysis of the
market can also reveal new opportunities, such as shares which are undervalued
or asset classes which are expected to do well.

21.5.2 Monitoring the investor’s objectives

When monitoring the changes in the investor’s circumstances, the following
aspects must be taken into account:
• changes in wealth;
• changes in the investment time horizon;
• changes in liquidity requirements;
• changes in tax circumstances;
• changes in legal considerations; and
• changes in other circumstances and the investor’s needs.
This is done less frequently than a market analysis. An investor might consider
these factors annually, for example, and if there are any changes that are identified,
the investor will see how the changes affect their investment strategy and therefore
investment implementation.

21.5.3 Monitoring the asset mix

The asset mix of a portfolio will change over time as assets grow at different rates.
Depending on the strategy of the investor, their tolerance for tactical deviations
from the strategic asset allocation, and the continued market outlook for the asset
classes, the investor may choose to rebalance the portfolio to be in line with their
strategic asset allocation from time to time.

Exercises for 21.5.3 Monitoring the asset mix

Ex.21.7 An investor has the strategic objective to have a 60%–70% exposure to

equities. He monitors his portfolio monthly, and realises that equities
have grown more than the other assets, and that he is now 75% invested
in equities. He considers whether he should rebalance immediately
or wait for another month. How will his decision be affected if he is
bullish (i.e. has a positive outlook) on equities? How will his decision
be different if his outlook is bearish (i.e. he expects equities to perform
poorly)?
396 Chapter 21 Asset and Portfolio Management

21.5.4 Monitoring diversification

This is best illustrated by way of an example:

Example 21.7 Say an investor has chosen to invest in the shares of three
companies, one involved in manufacturing, one in retail, and one in health-
care. Over time, the manufacturing company acquires a major chemical
plant which supplies ingredients for pharmaceuticals, while the retail com-
pany opens a chain of drugstores. What was originally a well-diversified
equity portfolio is now much more correlated, and the investor may consider
selling one or two of his shares and finding more diversified alternatives
instead.

21.5.5 Monitoring the asset manager

An investor chooses the asset manager for their skill in implementing the in-
vestor’s strategy according to the investment philosophy they share. So an investor
should check that the manager continues to:
1. have the requisite skills - watch out for staff turnover and loss of key personel;
2. be consistent about implementing their professed philosophy - for example,
a growth manager who continues to buy value shares seems to be conflicted
about their process;
3. deliver long-term performance in line with the stipulated return objectives
- but bear in mind that short-term performance may vary a lot, and that
performance by itself is not a good reason for changing managers. Man-
agers, like assets, go through performance cycles, and investors who always
leave a manager at the bottom of their performance cycle tend to severely
underperform the market over the long term.
Solutions 397

Selected Numerical Solutions for Part 3

Chapter 15 Solutions
Ex. 15.4 R6 967.32
Ex. 15.5 R10 261.23
Ex. 15.6 1. 10 day: R998.66
2. 1 month: R995.16
3. 6 months: R980.58
Ex. 15.8 R246 875
Ex. 15.9 Using 20% simple interest, R9 671.23

Chapter 16 Solutions
Ex. 16.1 • 4%: 108.11
• 5%: 100
• 6%: 92.64
Ex. 16.2 • Coupon of 4%: 78.68
• Coupon of 7%: 100
• Coupon of 10%: 121.32
Ex. 16.4 R14 181.40
Ex. 16.5 R214 653.29
Ex. 16.6 • 4.5%: R51 672.04
• 6.5%: R38 882.65
• 8.5%: R29 413.99
• 10.5%: R22 364.84
Ex. 16.7 6%
Ex. 16.8 R550 956.83
Ex. 16.9 R44.86
Ex. 16.10 R1 344 884.79 (11% is the nominal yield)

Chapter 19 Solutions
Ex. 19.2 • Profit of R7
• Loss of R7
• Loss of R19
• Loss of R4
Ex. 19.8 1. Loss of R25 million
2. Profit of R101 million