BASICS OF RESERVING FOR

GENERAL INSURANCE

November 16th – 18th, 2020

Contents

INTRODUCTION TO INSURANCE
ACTUARIAL RESERVING IN
GENERAL INSURANCE
INTRODUCTION TO RESERVES

DATA RECONCILIATION

DATA CHECKS

TECHNICAL RESERVES

www.badriconsultancy.com 3
BASICS OF RESERVING FOR
GENERAL INSURANCE
Introduction to Insurance
Introduction to Insurance and Reserving
Risk
An insurance cover provides Insurer
the insured party protection Insurance
Policy
against future risks in
exchange of a small amount, Policyholder
i.e. premium.
Motor
Lines of
For a policyholder, the Medical
Business
purpose of buying insurance Property/Fire (LOBs)
is not to profit from the
Marine
arrangement, but to be
indemnified against potential
losses. Written
Premium Earned/Unearned

Paid/Outstanding
Reported Claims
Ultimate

www.badriconsultancy.com 5
The Underwriting Process

Underwriting is the process of an insurance company preparing and selling an

insurance contract to a customer, the policyholder. More importantly, it includes
assessing whether the risk is acceptable and, if so, determining the appropriate
premium together with the terms and conditions of the cover.

An underwriter’s goal is to protect the insurance company against:

• Adverse Selection (very poor risks) and
• Moral Hazard (parties who may have a fraudulent intent).

Using a Rating Manual

Evaluating information
and Underwriting Policy is then issued and
and resources to
Guidelines to calculate subsequently delivered
determine how an
the premium that the to the policyholder by
individual will be
applicant will be the insurance agent
classified
charged.

www.badriconsultancy.com 6
General Insurance Policy Life-Cycle

Business Policy Policy

Acquisition Commence Expires
Activities
Policy Year Post Expiry Period

Premium
Written
Premium Earned

Claims Incurred
Claims Reported
Claims Paid

www.badriconsultancy.com 7
Policy Life-Cycle & Valuation Date

Business Policy Policy

Acquisition Commence Expires
Activities
Policy Year Post Expiry Period

Premium
Written
Premium Earned

Claims Incurred
Claims Reported
Claims Paid

Valuation Date
e.g. Financial Period Closing

www.badriconsultancy.com 8
The Life of a Claim
A claim is reported

Inspection by the Claim

Department (to confirm the
extent of the damages and
check if the claim is covered).
Claim Denied (customer
Claim is accepted and a
can pursue a legal route
case reserve is established.
to make sure they are
indemnified if they think
the claim should be
Update case outstanding if covered
either new information is
available on loss or a partial
payment is made.
The claim is rejected, and
no payment is made
The final payment is
made, and the claim is
closed.

www.badriconsultancy.com 9
BASICS OF RESERVING FOR
GENERAL INSURANCE
Introduction to Reserves
What is an Actuarial Reserve?

An Actuarial Reserve, also known as Technical

Provisions in Europe, is the present value of the
liabilities that the insurance company has taken
on its book as a result of writing a block of
business, which gives rise to the obligation of
settling the valid claims that incur in a
prescribed period, irrespective of when it is
reported and settled. In addition, the Actuarial
Reserve includes the premium that has been
written but not yet earned including any
potential deficiency in the premium thereof!

www.badriconsultancy.com 11
Types of Technical Reserves

Outstanding Loss
Reserve (OSLR)
Includes claims that have been reported but not yet settled.

Unearned Premium
Reserve (UPR) Includes the premium amount that corresponds to the time period
remaining on an insurance policy.

Incurred But Not

Reported Reserve Refers to an estimate of the amount owed by an insurer to all valid
(IBNR) claimants who have had a covered loss but have not yet reported it.

Premium Deficiency
Reserve (PDR) Refers to the amount required by an insurer if the unearned premium
may not be sufficient to meet future claims and expenses.

Unallocated/Allocated
Loss Adjustment The costs incurred by the insurer in handling claims are included in
Expenses (ULAE/ALAE) these reserves.

www.badriconsultancy.com 12
Understanding your Data - Premium
Data Field Description Example
The name of the business
Class / Line of Business Medical
segment
A unique code assigned to each
Class Code 300
LOB.
Different sub-categories that exist
4140 is the sub class code
Sub Class within a line of business. This
for Medical Group.
example is for Medical Group only
A unique ID assigned to each
Policy Number P/300/4140/19/000771
policy
This ID is generated if there has
Endorsement Number E/300/4140/19/000771/6
been an endorsement.
Document Date / Issue The date the policy is issued by
12-Feb-19
Date the Company.
Policy From / Effective Policy coverage starts from this
01-May-19
Date date

Policy To / Expiry Date Policy coverage ends at this date 30-Apr-20

www.badriconsultancy.com 13
Understanding your Data - Premium
Data Field Description Example
This can be different from the
Member Start Date Policy Start Date if a member joins 09-May-19
a group later.
This date is always the same as
Member End Date 30-Apr-20
the Policy End Date.
Endorsement Type / New Business / Renewal /
The type of endorsement passed.
Transaction Type Cancelation
Insured Name Name of the policyholder ABC Company Ltd.
The total premium charged for
Gross Premium SAR 1,000
the policy.
Premium due to the reinsurer if
Reinsurer Premium SAR 500
the LOB is reinsured.

Commission Commission due to other parties. SAR 100

Sum Insured/ Annual The maximum loss a policy will

SAR 200,000
Limit cover
www.badriconsultancy.com 14
Understanding your Data – Claims Paid/Outstanding
Data Field Description Example
The name of the business
Class Medical
segment
Different sub-categories that exist
Sub Class Group
within a line of business.
A unique ID assigned to each
Claim Number C/300/4140/20/000135
claim received
A unique ID assigned to each
Policy Number P/300/4140/19/000771
policy
Policy coverage starts from this
Policy Start Date 01-May-19
date

Policy End Date Policy coverage ends on this date 30-Apr-20

Reported Date The loss was reported on this date 05-Jan-20

Loss Date The loss occurred on this date 01-Jan-20

www.badriconsultancy.com 15
Understanding your Data – Claims Paid/Outstanding

Data Field Description Example

This is the date when the

Payment Date 06-Feb-20
payment was made to the insured
The amount payable to the
Claim Amount SAR 3600
policyholder for a claim
Claim amount recoverable from
RI Claim SAR 1800
the reinsurer
Outstanding Claim The amount of claim that is yet to
SAR 500
Amount be paid to the policyholder.
RI Outstanding Claim The amount of the claim that is to
SAR 250
Amount be paid by the reinsurer.
The date at which the total
As at Date 30-Sep-20
Outstanding claims are valued.

www.badriconsultancy.com 16
BASICS OF RESERVING FOR
GENERAL INSURANCE
Data Reconciliation
Why Perform Data Checks and Reconciliation?

Accuracy • The data should match with the financials.

• The data must meet the requirements for its intended

Relevancy use.

• The data should not have missing values or missing

Completeness data records.

Timeliness • The data should be up to date.

• The data should have the format as expected and can

Consistency be cross reference-able with the same results.

www.badriconsultancy.com 18
Data Reconciliation

How to Reconcile Data?

Premium Data Claims Paid Data Claims OS Data

Reconciled on the basis of Reconciled on the basis of Reconciled on the basis of
the date on which the policy the date on which the claim the date on which the claims
was issued known as the was paid known as the are being valuated known as
Document Date / Issue Date Payment Date / Settled Date. the Valuation Date or the As
/ Invoice Date. at Date.

www.badriconsultancy.com 19
BASICS OF RESERVING FOR
GENERAL INSURANCE
Premium Data Quality Checks
Premium Data Quality Checks

• Policy IDs are important for distinguishing

Blank Policy ID between each issued policy.

Blank Policy Effective •Policy effective dates are crucial for calculating the
date Unearned Premium Reserve (UPR).

Blank Policy Expiry •Policy Expiry dates are required to calculate the
date UPR.

•Document / Issue date is used for data

Blank Document date reconciliation, hence an important field.

Blank Gross Premium •It is an important KPI for reserving

www.badriconsultancy.com 21
Premium Data Quality Checks

Difference in start •A policy should have only one start date. If the data
has multiple start dates for each policy, it is likely
dates that the data is endorsement adjusted.

Here is a specific example of a The Document Date for this policy

policy where we can observe was changed when the policy was
differences in the start date. We do canceled. Several companies have
not have a policy start or end date a similar practice. Hence, this
for Marine Cargo policies. Hence, check is important to get accurate
the Document Date can be used as policy start dates.
a proxy for the effective date.

Premium Sub Policy Effective Policy Expiry Document Endorsement Gross

Policy ID
Class Date Date Date Type Premium
XYZ/1234/001254 Marine Cargo 05-May-2016 05-May-2016 Cancelation -163
-

XYZ/1234/001254 Marine Cargo 04-May-2016 04-May-2016 Policy 163

-

www.badriconsultancy.com 22
Premium Data Quality Checks

Difference in End •A policy should have only one end date, and if the
data has multiple end dates for each policy, it is
dates likely that the data is endorsement adjusted.

•A policy’s expiry date should be greater than its

effective date as a policy cannot expire before it
Policy Dates Check starts.
•This check is also useful for ensuring that the
dates’ format is as required.

Premium Weighted •The average number of days that policies in a

portfolio are effective for. This can be different for
Days each LOB.

•The policy is issued/recorded before its coverage

Effective Date Before starts hence the document date should be before
Document Date the effective date.

www.badriconsultancy.com 23
Premium Data Quality Checks
Premium Trend •Used to identify the trend of the business written
Analysis during the year.

www.badriconsultancy.com 24
Premium Data Quality Checks

•Logically, the policy shouldn’t have negative gross

Negative Premium premium.

A policy, effective from October If the premium data set received

1, 2014, has a premium of SAR is from the UWY 2015, it will
1000. contain the policy cancelation
endorsement but not the policy
On March 31, 2015, the policy is addition endorsement, which
canceled and the unearned will result in an overall negative
portion of the premium of SAR premium for this policy.
500 is refunded to the
policyholder.

GWP
Policy Start Date 01-Oct-14 1,000
Policy Cancelation Date 31-Mar-15 (500)

www.badriconsultancy.com 25
Premium Data Quality Checks

•The RI Premium and Retained Premium should

RI Premium Check sum up to the Gross Premium.

RI Premium •To check RI Distribution into different treaties i.e.

Quota Share, Surplus, Facultative, and Excess of
Distribution Loss.

•To check whether the UW Year and FY Year

UY vs FY Premium correspond i.e. a policy with an UW year of 2018
must have a financial date of 2018 or beyond.

Gross Written Premium (SAR '000)

LOB UY/ Document Year 2015 2016 2017 2018 Total
2015 14,093 - - - 14,093
2016 218 14,821 - - 15,039
Fire
2017 - 158 14,794 - 14,952
2018 - - 123 9,775 9,898
Total 14,311 14,979 14,917 9,775 53,982

www.badriconsultancy.com 26
BASICS OF RESERVING FOR
GENERAL INSURANCE
Claims Data Quality Checks
Claims Data Quality Checks

Blank Claims Paid/OS •Crucial KPI for all kinds of analyses.

Blank Claim IDs •Important for identifying each claim in the data.

•It is a crucial field since reserving is done on a Loss

Blank Loss Dates Year basis.

•Reporting Date is a crucial field from the reserving

Blank Reporting Dates point of view
•Crucial for several analyses like reporting lags.

•Payment Date is a crucial field from the reserving

Blank Payment Dates point of view.
•Crucial for several analyses like payment lags.

www.badriconsultancy.com 28
Claims Data Quality Checks

Follow the Reporting

chronological Date
order! before
Loss Date

Payment
Loss Date Date More Data Payment
Date
Reporting Checks! before
Date Loss Date

Payment
Claim ID Loss Date Reporting Date Payment Date Date
C/300/4140/20/000045 27-Sep-17 27-Jun-17 15-Oct-17 before
Reporting
Date

www.badriconsultancy.com 29
Claims Data Quality Checks

Differences in Dates

Follow the
chronological
order!

Claim ID Loss Date Reporting Date Payment Date Claim Paid As at Date Gross Claims OS
XYZ/1234/000001 03-Aug-16 06-Aug-16 - 30-Sep-16 1,048
XYZ/1234/000001 03-Aug-16 06-Aug-16 - 31-Dec-16 1,048
XYZ/1234/000001 03-Aug-16 06-Aug-16 - 31-Mar-17 1,048
XYZ/1234/000001 27-Jun-17 28-Jun-17 28-Jun-17 1,898 -

Difference in Difference in
Loss Dates Reporting Dates

www.badriconsultancy.com 30
Claims Data Quality Checks
Comparing claims against Accident Year (AY) and Financial Year (FY)
Claims reported in a certain Accident Year will be recorded in the financials
Why AY vs. FY? of the same year.
Treatment for
Claims Paid Compare Loss Year with the Payment Year.
Treatment for
Claims Outstanding Compare Loss Year with Valuation Date.

Gross Claims Paid (SAR '000)

LOB Loss/Payment Year 2015 2016 2017 2018 Total
2015 316 3,980 6,294 10,591
2016 212 3,476 3,688
Fire
2017 208 2,744 2,952
2018 138 138
Total 316 4,192 9,978 2,882 17,368

Gross Claims Outstanding (SAR '000)

LOB Loss Year/As at Date 31-Dec-15 30-Jun-16 30-Jun-17 30-Jun-18 Total
2015 15,853 13,322 7,074 -393 35,856
2016 289 2,223 91 2,604
Fire
2017 841 689 1,530
2018 444 444
Total 15,853 13,611 10,138 832 40,433

www.badriconsultancy.com 31
Claims Data Quality Checks

Some Logical Checks

Reinsurance
Claims > Gross
Claims
Claims without
Policy Data
Negative Claims •Since it is always a portion
of the Gross claims that is
Reported
due to the reinsurer, the
•These are also known as Reinsurance Claims amount
Orphan Claims. They are
can never be greater than
•Claims reported cannot be claims that cannot be linked
the Gross Claims amount.
negative by logic. However, to the respective policy.
recoveries made by an They can be caused due to:
insurance company might •An absence of a
appear as negative in some common key between
data sets. premium and claims
data.
•Data entry issues

www.badriconsultancy.com 32
Claims Data Quality Checks

Loss Date

Policy Start Date Policy End Date

Loss Date before Risk

Start Date
Loss Date after Risk End
• Not covered by insurance Date

• Not covered by insurance

www.badriconsultancy.com 33
BASICS OF RESERVING FOR
GENERAL INSURANCE
Technical Reserves
Outstanding Loss Reserves (OSLR)
Out of the claim amount that has been reported as at the valuation date, OSLR
is the amount of claim the insurer has not yet paid and thus expects to pay.

The OSLR is also known as the Case Reserve.

OSLR is calculated on a claim level and is summed across LOBs or segments.

Note that the amount of claim that has not been reported as at the valuation
date, will not be a part of the OSLR.

Outstanding claims are estimated by the insurance companies. It is likely that

the estimation is being made by the claim adjuster.

This estimation can be made using various OSLR philosophies.

Most Likely Expected

Maximum
Settlement Settlement
Settlement
Value Value
Value, Etc.
(Mode) (Mean)

www.badriconsultancy.com 35
Outstanding Loss Reserves (OSLR) Sufficiency

Gross Claims OS Gross Claims OS Over / (Under)

Reserving LOB Gross Claims Paid
as of 30-Jun-2020 as of 30-Sep-2020 Reserved
Medical 1,689,180 42,161 1,545,530 101,489

Gross Claims OS Gross Claims OS Over / (Under)

Reserving LOB Claim ID Gross Claims Paid
as of 30-Jun-2020 as of 30-Sep-2020 Reserved
Total 1,689,180 42,161 1,545,530 101,489
C/300/4140/19/000001 200,000 0 200,000 0
C/300/4140/18/000001 1,256,585 0 1,256,585 0
C/300/4140/20/000052 12,598 0 0 12,598
C/300/4140/18/000012 3,000 0 0 3,000
C/300/4140/18/000011 3,000 0 0 3,000
C/300/4140/18/000010 3,000 0 0 3,000
Medical C/300/4140/18/000008 3,000 0 0 3,000
C/300/4140/16/000203 100,380 0 0 100,380
C/300/4140/15/000055 88,045 0 88,045 0
C/300/4140/20/000056 199 85 0 114
C/300/4140/20/000055 1,373 452 0 921
C/300/4140/20/000054 18,000 0 900 17,100
C/300/4140/20/000053 0 41,624 0 -41,624

www.badriconsultancy.com 36
Unearned Premium Reserves (UPR)

Determine how insured risk is spread over the policy period

Is it spread uniformly across the policy Does the risk increase later into the policy
period? (This is the usual case) period? (e.g. Engineering (CAR & EAR)

For our training purpose, we will assume that insured risk is spread uniformly over the
policy period.
Hence written premium is also earned uniformly over the policy period i.e. the same
proportion of premium is earned everyday of the policy period

Policy Policy
Commence Expires Post Expiry Period
Business
Acquisition
Activities
UNIFORM PREMIUM EARNING PATTERN

Written
Premium Valuation Date

www.badriconsultancy.com 37
Unearned Premium Reserves (UPR)
Earned Premium = share of written premium received associated with the proportion of policy
period that has lapsed as at the valuation date

Unearned Premium = share of written premium received associated with the proportion of
policy period that is yet to lapse as at the valuation date

Written Premium = Unearned Premium + Earned Premium

UPR is calculated on a policy level and then summed across LOBs or segments

Policy Policy
Commence Expires

UNIFORM PREMIUM EARNING PATTERN

Written
Premium Valuation Date Unearned
Earned Premium
Premium
www.badriconsultancy.com 38
UPR Example 1

An insurance policy starts on January 1,

2020 and expires on December 31, 2020.

The total written premium is SAR 2,400.

At June 30, 2020, what are the values of

Earned Premium and Unearned Premium
Reserve?

Value of Earned Premium: SAR____ 1200___

Value of Unearned Premium Reserve: SAR____ 1200 ___

www.badriconsultancy.com 39
UPR Example 2

If an insurance policy starts on January 1,

2020, for 1000 SAR. Policy has a two-year
life span, and no claim has been filed in nine
months.

At September 30, 2020, what are the values

of Written premium, Earned Premium and
Unearned Premium Reserve?

Value of Written Premium: SAR __ 1000___

Value of Earned Premium: 375 __

SAR_____

Value of Unearned Premium Reserve: SAR____625 __

www.badriconsultancy.com 40
UPR Checks

UPR to Premium •To see if this ratio is in line with the premium
ratio trend.

UPR to Premium Ratio

UW Year Document Year 2017 2018 2019 2020
2017 2017 0% 0%
2018 2018 0% 0%
2019 2019 5% 23%
2020 2020 67%
Total 0% 0% 4% 64%

Gross UPR greater •Gross UPR should not be greater than Gross
than Gross Premium Premium.

•Logically Gross UPR should not be negative, because

Negative Gross UPR the premium can’t be negative.

www.badriconsultancy.com 41
UPR Checks

RI UPR greater than

•RI UPR should not be greater than Gross UPR.
Gross UPR

•the retention ratio should be aligned with the

Net to Gross Ratio reinsurance treaties.

UW Year 2016 2017 2018 2019 2020

2016
Net to
2017
Gross
2018
UPR
2019 60% 60%
Ratio
2020 60% 60%
Total 60% 60%
2016 61% 61%
Net to
2017 60% 60% 60%
Gross
2018 60% 60% 60% 60%
Premium
2019 60% 60% 60%
Ratio
2020 60% 60%
Total 61% 60% 60% 60% 60%

www.badriconsultancy.com 42
Claim Reserves
Policy Policy
Commence Expires

Post Expiry Period

Claims Incurred Paid Claims

Claims Reported
Claims Paid

Claims Incurred
Claims Reported

Claims Paid

Claims Incurred
Claims Reported
Claims Paid CLAIM RESERVES
Claims Incurred
Claims Reported
Claims Paid

Valuation Date

www.badriconsultancy.com 43
Claim Reserves
CLAIM EVENT HAPPENED

CLAIM INITIALLY REPORTED TO CLAIM FULLY REPORTED TO THE

INSURANCE COMPANY INSURANCE COMPANY

CLAIM PROCESSED AND SETTLED

Various claims permutation are possible, such as:

Claim Incurred Claim Incurred Claim Incurred Claim Incurred

+ + +
Claim Initially Reported Claim Initially Reported Claim Initially Reported

+ +
Claim Fully Reported Claim Fully Reported IBNR = IBNER + IBNYR

+
Claim Settled OUTSTANDING CLAIMS INCURRED BUT NOT INCURRED BUT NOT
RESERVE ENOUGH REPORTED REPORTED
(OSLR) (IBNER) (IBNYR)

www.badriconsultancy.com 44
BASICS OF RESERVING FOR
GENERAL INSURANCE
Loss Development Methods
Basic Loss Development Methods

There are several methods to estimate the unpaid claims, the most common ones
being:

Expected
Loss Chain-Ladder Bornhuetter-
(CL) Ferguson (BF)
Ratio (ELR)

There are other methods such as the Cape Cod method and the Frequency and
Severity method, also known as the Collective Risk Method (CRM).

www.badriconsultancy.com 46
The Problem
1. A block of policies sold throughout the year in 2012
2. Claims for this block of policies aren’t fully paid until the end of 2016
3. How does the company know if its business is profitable or not from 2012 till
2016? SAR 60

SAR 50

SAR 40
Thousands

SAR 30

SAR 20

SAR 10

SAR 0
12/31/12 12/31/13 12/31/14 12/31/15 12/31/16

Paid Case Reserve IBNR Reserve

Interested Parties in this Problem:

• Company Management – which business segments are the most/least profitable
• Investors – how profitable is the company?
• Regulators – does the company have sufficient funds available to meet its obligations
to policyholders

www.badriconsultancy.com 47
ELR Method

We are given the following information:

1. Valuation Date = June 30, 2020
2. Calculating IBNR for Accident Quarter 2020-Q2
3. Premium Earned During 2020-Q2 = SAR 25,000
4. Expected loss ratio for 2020-Q2 = 75%

Expected ultimate losses for 2020-Q2 = SAR 25,000 * 75%

= SAR 18,750

Reported Claims for 2020-Q2 at June 30, 2020 = SAR 1,406

IBNR for 2020-Q2 as at June 30, 2020

= Expected Ultimate Losses – Reported Claims
= SAR 18,750 – SAR 1,406
= SAR 17,344

www.badriconsultancy.com 48
ELR Method – Multiple Years

(1) (2) (3) (4) = (2) x (3) (5) (6) = (4) – (5)

Accident Earned Expected Expected Reported Loss as IBNR

Quarter Premium Loss Ratio Ultimate Loss at June 30, 2020

2018-Q3 15,000 75% 11,250 10,852 398
2018-Q4 16,500 75% 12,375 15,045 -2,670
2019-Q1 17,000 75% 12,750 15,878 -3,128
2019-Q2 19,000 75% 14,250 14,967 -717
2019-Q3 20,000 75% 15,000 15,425 -425
2019-Q4 22,000 75% 16,500 11,836 4,664
2020-Q1 24,000 75% 18,000 5,609 12,391
2020-Q2 25,000 75% 18,750 1,406 17,344
Total 158,500 118,875 91,018 27,857

www.badriconsultancy.com 49
ELR Method – Pros and Cons

– Simple calculation , easy to explain – Not responsive to data

– Minimal data requirement fluctuations

– Suitable to segments where there – Not responsive to changes in risk

is no or limited data available, e.g. environment
new LOBs, product expansion into a – No accounting for inherent
new region uncertainty in the expected loss
– Not sensitive to data fluctuations ratio – the key parameter

www.badriconsultancy.com 50
Reported Loss Development Data
Accounting Configuration

Accident Quarter Reported Losses (in SAR 000s)

Cumulative Totals as of:
Accident Quarter
2018-Q3 2018-Q4 2019-Q1 2019-Q2 2019-Q3 2019-Q4 2020-Q1 2020-Q2

2018-Q3 696 2,785 5,262 8,178 9,522 10,604 10,803 10,852

2018-Q4 776 3,907 8,383 12,748 14,161 14,805 15,045
2019-Q1 1,058 4,344 8,501 11,912 15,148 15,878
2019-Q2 1,106 4,589 7,929 12,618 14,967
2019-Q3 1,230 4,829 10,355 15,425
2019-Q4 1,281 5,696 11,836
2020-Q1 1,217 5,609
2020-Q2 1,406

www.badriconsultancy.com 51
Reported Loss Development Data
Actuarial Configuration

Accident Quarter Reported Losses (in SAR 000s)

Accident
Cumulative Totals by Development Age in Months
Quarter

3 6 9 12 15 18 21 24
2018-Q3 696 2,785 5,262 8,178 9,522 10,604 10,803 10,852
2018-Q4 776 3,907 8,383 12,748 14,161 14,805 15,045
2019-Q1 1,058 4,344 8,501 11,912 15,148 15,878
2019-Q2 1,106 4,589 7,929 12,618 14,967
2019-Q3 1,230 4,829 10,355 15,425
2019-Q4 1,281 5,696 11,836
2020-Q1 1,217 5,609
2020-Q2 1,406

www.badriconsultancy.com 52
Loss Development Factor (LDF)
From 3 to 6 Months

Accident Cumulative Reported Cumulative Reported Loss Development

Quarter at 3 Months at 6 Months Factor

2018-Q3 696 2,785 4.002 = 2,785 / 696
2018-Q4 776 3,907 5.032 = 3,907 / 776
2019-Q1 1,058 4,344 4.107 = 4,344 / 1,058
2019-Q2 1,106 4,589 4.151 = 4,589 / 1,106
2019-Q3 1,230 4,829 3.926 = 4,829 / 1,230
2019-Q4 1,281 5,696 4.445 = 5,696 / 1,281
2020-Q1 1,217 5,609 4.611 = 5,609 / 1,217
Total 7,364 31,759 4.313 = 31,759 / 7,364

2020-Q2 1,406 ??? ???

www.badriconsultancy.com 53
Loss Development Factor (LDF) (Cont’d)
From 3 to 6 Months

Accident Cumulative Reported Cumulative Reported Loss Development

Quarter at 3 Months at 6 Months Factor

2018-Q3 696 2,785 4.002 = 2,785 / 696
2018-Q4 776 3,907 5.032 = 3,907 / 776
2019-Q1 1,058 4,344 4.107 = 4,344 / 1,058
2019-Q2 1,106 4,589 4.151 = 4,589 / 1,106
2019-Q3 1,230 4,829 3.926 = 4,829 / 1,230
2019-Q4 1,281 5,696 4.445 = 5,696 / 1,281
2020-Q1 1,217 5,609 4.611 = 5,609 / 1,217
Total 7,364 31,759 4.313 = 31,759 / 7,364

2020-Q2 1,406 1,406 x 4.300 = 6,046 Selected = 4.300

www.badriconsultancy.com 54
Reported Loss Development Data
Actuarial Configuration

Accident Quarter Reported Losses (in SAR 000s)

Accident
Cumulative Totals by Development Age in Months
Quarter

3 6 9 12 15 18 21 24
2018-Q3 696 2,785 5,262 8,178 9,522 10,604 10,803 10,852
2018-Q4 776 3,907 8,383 12,748 14,161 14,805 15,045
2019-Q1 1,058 4,344 8,501 11,912 15,148 15,878
2019-Q2 1,106 4,589 7,929 12,618 14,967
2019-Q3 1,230 4,829 10,355 15,425
2019-Q4 1,281 5,696 11,836
2020-Q1 1,217 5,609
2020-Q2 1,406 6,046

www.badriconsultancy.com 55
Reported Loss Development Data
Loss Development Factor (LDF) Selection

Accident
Quarter 3-6 6-9 9-12 12-15 15-18 18-21 21-24 24-Ult
2018-Q3 4.002 1.889 1.554 1.164 1.114 1.019 1.005
2018-Q4 5.032 2.146 1.521 1.111 1.045 1.016
2019-Q1 4.107 1.957 1.401 1.272 1.048
2019-Q2 4.151 1.728 1.591 1.186
2019-Q3 3.926 2.144 1.49
2019-Q4 4.445 2.078
2020-Q1 4.611
2020-Q2

Wtd Avg 4.313 1.999 1.506 1.184 1.063 1.017 1.005

Selected 4.300 2.000 1.500 1.185 1.065 1.017 1.005 ???

www.badriconsultancy.com 56
Reported Loss Development Data
Loss Development Factor (LDF) Projection

Accident
Quarter 3-6 6-9 9-12 12-15 15-18 18-21 21-24 24-Ult
2018-Q3 4.002 1.889 1.554 1.164 1.114 1.019 1.005 ???
2018-Q4 5.032 2.146 1.521 1.111 1.045 1.016 1.005 ???
2019-Q1 4.107 1.957 1.401 1.272 1.048 1.017 1.005 ???
2019-Q2 4.151 1.728 1.591 1.186 1.065 1.017 1.005 ???
2019-Q3 3.926 2.144 1.490 1.185 1.065 1.017 1.005 ???
2019-Q4 4.445 2.078 1.500 1.185 1.065 1.017 1.005 ???
2020-Q1 4.611 2.000 1.500 1.185 1.065 1.017 1.005 ???
2020-Q2 4.300 2.000 1.500 1.185 1.065 1.017 1.005 ???

Selected 4.300 2.000 1.500 1.185 1.065 1.017 1.005 ???

www.badriconsultancy.com 57
Reported Loss Development Data
“Squaring the Triangle”

Accident Quarter Reported Losses (in SAR 000s)

Accident
Cumulative Totals by Development Age in Months
Quarter

3 6 9 12 15 18 21 24
2018-Q3 696 2,785 5,262 8,178 9,522 10,604 10,803 10,852
2018-Q4 776 3,907 8,383 12,748 14,161 14,805 15,045 15,121
2019-Q1 1,058 4,344 8,501 11,912 15,148 15,878 16,148 16,229
2019-Q2 1,106 4,589 7,929 12,618 14,967 15,940 16,211 16,292
2019-Q3 1,230 4,829 10,355 15,425 18,278 19,466 19,797 19,896
2019-Q4 1,281 5,696 11,836 17,754 21,038 22,405 22,786 22,900
2020-Q1 1,217 5,609 11,218 16,827 19,940 21,236 21,597 21,705
2020-Q2 1,406 6,046 12,090 18,135 21,490 22,887 23,276 23,392

www.badriconsultancy.com 58
Loss Development Factors (LDF)
The Tail Factor

1. Loss development beyond the oldest age observed in the

historical data
2. Several approaches:

Other
Insurance
Curve fitting / mathematical
industry
extrapolation / statistical
benchmarks
models?

www.badriconsultancy.com 59
Loss Development Factors
Cumulative Development Factor (CDF)

Accident Selected Selected

Age (Months)
Quarter Age-Age LDF Cumulative Development Factor (CDF)

2018-Q3 24 Tail Factor = 1.010

2018-Q4 21 1.005 1.015 = 1.005 x 1.010

2019-Q1 18 1.017 1.032 = 1.017 x 1.015

2019-Q2 15 1.065 1.099 = 1.065 x 1.032

2019-Q3 12 1.185 1.303 = 1.185 x 1.099

2019-Q4 9 1.500 1.954 = 1.500 x 1.303

2020-Q1 6 2.000 3.908 = 2.000 x 1.954

2020-Q2 3 4.300 16.806 = 4.300 x 3.908

www.badriconsultancy.com 60
Chain Ladder (CL) Method
Reported Losses
(1) (2) (3) (4) (5) = (3) x (4) ‘(6) = (5) – (3)
Selected
Cumulative
Accident Age Cumulative Projected Incurred But Not
Reported Loss as
Quarter (Months) Development Ultimate Loss Reported (IBNR)
at 31/12/17
Factor (CDF)
2018-Q3 24 10,852 1.010 10,961 109
2018-Q4 21 15,045 1.015 15,271 226
2019-Q1 18 15,878 1.032 16,386 508
2019-Q2 15 14,967 1.099 16,449 1,482
2019-Q3 12 15,425 1.303 20,098 4,673
2019-Q4 9 11,836 1.954 23,128 11,292
2020-Q1 6 5,609 3.908 21,921 16,312
2020-Q2 3 1,406 16.806 23,627 22,221
Total 91,019 147,841 56,823

1 𝑅𝑒𝑝𝑜𝑟𝑡𝑒𝑑 𝐿𝑜𝑠𝑠
𝑅𝑒𝑝𝑜𝑟𝑡𝑒𝑑 𝐿𝑜𝑠𝑠 𝑈𝑙𝑡𝑖𝑚𝑎𝑡𝑒 𝐿𝑜𝑠𝑠 =
𝐶𝐷𝐹 = 𝐶𝐷𝐹 𝑈𝑙𝑡𝑖𝑚𝑎𝑡𝑒 𝐿𝑜𝑠𝑠
𝑈𝑙𝑡𝑖𝑚𝑎𝑡𝑒 𝐿𝑜𝑠𝑠 = = % 𝑅𝑒𝑝𝑜𝑟𝑡𝑒𝑑
𝐶𝐷𝐹 𝑅𝑒𝑝𝑜𝑟𝑡𝑒𝑑 𝐿𝑜𝑠𝑠
= 𝑜𝑟 % 𝑡𝑜 𝑈𝑙𝑡𝑖𝑚𝑎𝑡𝑒

www.badriconsultancy.com 61
Chain Ladder (CL) Method
Implicit Assumptions

1. Volume of historical loss data is large enough to be credible

2. Future reporting patterns will be similar to historically
observed patterns:

Changes to
judicial/legal
environment
Changes to New types of
insurers claims not
operations seen before

www.badriconsultancy.com 62
Comparison of Methods

ELR Method CL Method

a. Exposure based i.e. uses a. Ignores exposures i.e. no

Earned Premium as the usage of Earned Premium
exposure base b. Based on actual loss
b. Ignores actual loss experience experience
c. Provides stable estimates c. Estimates change in response
to new information, hence is
more volatile

www.badriconsultancy.com 63
Bornhuetter Ferguson (BF) Method

1. It is a mixture of the ELR and CL methods

2. Accounts for the % Reported for each accident quarter
▪ For accident quarters where % Reported is high, BF method relies
more on the CL method, thereby giving more weight to the data
▪ For accident quarters where % Reported is low, BF method relies
more on the ELR method, thereby giving less weight to the data
3. Therefore, the BF method is a credibility weighed method
▪ Credibility weight = % Reported = z

BF = 𝑧 × 𝐶𝐿 + ((1 − 𝑧) × 𝐸𝐿𝑅)

www.badriconsultancy.com 64
Bornhuetter Ferguson (BF) Method
(1) (2) (3) (4) (5) = 1 / (4)
Selected
Cumulative
Accident Age Cumulative
Reported Loss as % Reported
Quarter (Months) Development Factor
at June 30, 2020
(CDF)
2018-Q3 24 10,852 1.01 99% OLDER
2018-Q4 21 15,045 1.015 99%
2019-Q1 18 15,878 1.032 97%
2019-Q2 15 14,967 1.099 91%
2019-Q3 12 15,425 1.303 77%
2019-Q4 9 11,836 1.954 51%
2020-Q1 6 5,609 3.908 26%
2020-Q2 3 1,406 16.806 6%
Total 91,019 RECENT

1. For older accident quarters there is sufficient data for CL method to be credible
▪ Majority claims have been reported i.e. % Reported ~ 100%
2. For recent accident quarters there is limited data requiring usage of ELR method
▪ Majority claims are unreported i.e. % Reported ~ 0%
www.badriconsultancy.com 65
Bornhuetter Ferguson (BF) Method

(1) (2) (3) (4) (5) (6) (7) = 1 / (6) (4) (6) = (4) – (5)
Expected Expected Reported Loss Expected
Accident Earned % Reported IBNR
Ultimate Loss Ultimate Loss as Ultimate Loss

Quarter Premium ELR Method CL Method at Dec 31,2017 CDF z BF Method BF Method

2018-Q3 15,000 11,250 10,961 10,852 1.010 99% 10,964 112

2018-Q4 16,500 12,375 15,271 15,045 1.015 99% 15,228 183
2019-Q1 17,000 12,750 16,386 15,878 1.032 97% 16,273 395
2019-Q2 19,000 14,250 16,449 14,967 1.099 91% 16,251 1,284
2019-Q3 20,000 15,000 20,098 15,425 1.303 77% 18,913 3,488
2019-Q4 22,000 16,500 23,128 11,836 1.954 51% 19,892 8,056
2020-Q1 24,000 18,000 21,921 5,609 3.908 26% 19,003 13,394

2020-Q2 25,000 18,750 23,627 1,406 16.806 6% 19,040 17,634

Total 158,500 118,875 147,841 91,018 135,564 44,546

www.badriconsultancy.com 66
Premium Deficiency Reserve (PDR)
Policy Valuation Date
Policy
Business Commence Expires
Acquisition
Activities Post Expiry Period
EARNED PREMIUM UNEARNED PREMIUM RESERVE
Loss Ratio: 90% Loss Ratio: 90%
Expense Ratio: 30% Expense Ratio: 30%
Combined Ratio: 120% Combined Ratio: 120%

Usual method of projection:

If the combined ratio is more
At the valuation date, total - Loss Ratio for total claims than 100%, then the amount
claims and expenses are
- Expense Ratio for total of ‘deficiency’ in the
projected in respect of the
expenses premium must be set aside
group of policies which have
as Premium Deficiency
unearned premium. - Combined Ratio = Loss Reserve (PDR)
Ratio + Expense Ratio

The unearned premium will likely be deficient to pay the claims and expenses
associated with the unexpired policy period, and thus, a liability must be recognized.
PDR = Amount of Deficiency (20% in this case) * UPR

www.badriconsultancy.com 67
PDR Example 1

An insurance policy starts on January 1,

2020 and expires on December 31, 2020.

The total written premium is SAR 2,400.

The loss ratio on this policy is 50% and the
expense ratio is 25%, which results in the
combined ratio being 75%.

What is the value of the Premium

Deficiency Reserve as at June 30, 2020?

Value of Premium Deficiency Reserve: 0

SAR __________

www.badriconsultancy.com 68
PDR Example 2

An insurance policy starts on January 1,

2020 and expires on December 31, 2020.

The total written premium is SAR 2,400. The

loss ratio on this policy is 75% and the
expense ratio is 25%, which results in the
combined ratio being 100%.

What is the value of the Premium Deficiency

Reserve as at June 30, 2020?

Value of Premium Deficiency Reserve: 0

SAR __________

www.badriconsultancy.com 69
PDR Example 3

An insurance policy starts on January 1,

2020 and expires on December 31, 2020.

The total written premium is SAR 2,400. The

loss ratio on this policy is 100% and the
expense ratio is 25%, which results in the
combined ratio being 125%.

What is the value of the Premium Deficiency

Reserve as at June 30, 2020?

Value of Premium Deficiency Reserve: 300

SAR __________

www.badriconsultancy.com 70
Loss Adjustment Expenses (LAE)
The costs for an insurer in handling claims are known as the Loss Adjustment Expenses
(LAE). They can be divided into two categories:

Loss Adjustment Expenses

Unallocated Loss Adjustment

Allocated Loss Adjustment Expenses
Expenses (ULAE): Loss Adjustment
(ALAE): Loss Adjustment Expenses that
Expenses that cannot be directly
can be directly attributed to a specific
attributed to a specific claim. An
claim. An example might be the cost of
example would be the time it will take
hiring an independent adjuster to
the employees to process new and
handle a specific claim.
existing claims.

www.badriconsultancy.com 71
BASICS OF RESERVING FOR
GENERAL INSURANCE
Other Reserves
Other Reserves

Other Typical Reserves in a General Insurer’s Balance Sheet:

Unearned
Deferred Data
Reinsurance
Acquisition Deficiency
Commission
Cost Reserve
Income

www.badriconsultancy.com 73
Impact of COVID-19
Two major lines of business that were impacted by COVID-19 were Motor and
Medical.

Motor

▪ Due to a lockdown because of COVID-19, there was a significant decrease in the

number of accidents, and consequently in the number of claims received by a
company.
▪ Such events have an impact on the regular claims development and need to be
adjusted for during reserving.
▪ An example of making such adjustments was using the frequency and severity
technique to estimate the impact on reserves.

𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝐶𝑙𝑎𝑖𝑚𝑠 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝐿𝑜𝑠𝑠𝑒𝑠

𝐹𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 = 𝑆𝑒𝑣𝑒𝑟𝑖𝑡𝑦 =
𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝐸𝑥𝑝𝑜𝑠𝑢𝑟𝑒𝑠 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝐶𝑙𝑎𝑖𝑚𝑠

▪ The Najm Liability Determination report helps in the calculation of frequency

for each company.

www.badriconsultancy.com 74
Excerpt of the Najm Liability Determination Report

5608
4080

1400 1460

4080 5608
1400 1460

www.badriconsultancy.com 75
Impact of COVID-19
Medical

Medical is another line of business where people

might have avoided or delayed treatments because of
COVID-19. Thus, it is important to consider the impact
of this delay in the occurrence of claims on reserving.

A certain percentage of both outpatient and inpatient

claims were assumed as deferred claims. This was
than used to compute a final percentage of claims
that were deferred during the reserving period.

www.badriconsultancy.com 76