UNIVERSITY OF IRINGA

FACULTY OF BUSINESS AND ECONOMICS (FABEC)

DEPARTMENT OF ACCOUNTING AND FINANCE

LECTURE NOTES
COURSE: ACC224 – MANAGERIAL ACCOUNTING
TOPIC: DECISION MAKING UNDER CERTAINTY AND UNCERTAINTY

INTRODUCTION
In day to day business life, managers are obliged to make decisions on the course of action
that has to be taken. Decisions are frequently classified as those made under certainty and
those made under uncertainty. Models, formulas and calculations shall form the basis of
argument and finally the decision making.

RELEVANT INFORMATION
Relevant information is that information which will affect the decision to be made.
Information that will not affect the decision to be made is obviously irrelevant information.
 Relevant costs or benefits are those costs or benefits that will differ between various
alternatives considered. In other words, only differential or incremental cash flows
should be taken into account
 Compare the following two alternatives: Manufacturing components and Buying the
components from outside.

Manufacture Buy
100 components 100 components
TZS TZS
Direct Labour 10,000 10,000
Direct Materials 30,000 -
Variable overheads 5,000 -
Fixed overheads 20,000 20,000
Supplier’s purchase price - 50,000
Total 65,000 80,000

 Costs which remain the same under both alternatives are irrelevant for decision
making
 Now, let us take the relevant costs only:

Manufacture Buy
100 components 100 components
TZS TZS

Direct materials 30,000 -

Variable overheads 5,000 -
Supplier’s purchase price - 50,000
Total 35,000 50,000

DECISION: MANUFACTURE THE COMPONENTS SINCE COSTS WILL BE

TZS15,000 LOWER

1
  For the purpose of decision making, the cost of manufacturing one component will be
TZS 350
 However, for the purpose of stock valuation, the cost of manufacturing one
component will be TZS 650, i.e. all costs will be included.

TYPICAL DECISION RULES

Typical decision rules include the following:
 Expected value (by using decision tables or decision trees)
 Maximin, maximax and regret criterion

Decisions under certainty

Conditions of decisions under certainty

 There is absolutely no doubt about the occurrence of an event – probability is
100%.
 There is a single outcome for each possible action

Decisions table (pay off table)

In decision making, a decision making table (payoff table) can be used.

Illustration 1:
Think of a situation whereby we have to make decision on whether to re-arrange our labour
force and incur a labour cost of TZS 1.8 per unit or do nothing and incur a labour cost of TZS
2 per unit. We want to produce 100,000 units. If we are quite certain of these outcomes, then
the probability is 100% or 1 for each outcome. We can show this in the following decision
table:

Event: Successful Implementation

Probability of event 1.0

ACTIONS: OUTOME:
(Labour cost)
Do nothing TZS 2 x 100,000 x 100% (labour cost) = TZS 200,000
Re-arrange TZS 1.80 x 100,000 x 100% = TZS 180,000

This is a one- column decision table.

The most desirable outcome should be the one giving the least labour cost, TZS 180,000

Decision under uncertainty

A decision under uncertainty is when there are many unknowns and no possibility of
knowing what could occur in the future to alter the outcome of a decision. We feel
uncertainty about a situation when we can’t predict with complete confidence what the
outcomes of our actions will be. We experience uncertainty about a specific question when
we can’t give a single answer with complete confidence.

2
Decision under uncertainty applies the following model:

Identify Search for Identify state List possible Measure Select

Objectives possible of outcomes pay offs course of
E.g. course of nature/events action-Do
Minimize action Success/fail- Monetary nothing/rea
labour cost Success/fail ure Value rrange

 An objective is a target that the decision maker is hoping to achieve. For example to
maximize profits or to achieve a certain present value of cash flows. The
quantification of an objective is often called OBJECTIVE FUNCTION. An
objective function is used to evaluate the alternative courses of action and provides
the basis for choosing the best alternative.
 The search for alternative courses of action that will enable the objective to be
achieved
 Because decision problems exist in an uncertain environment, it is necessary to
consider those uncontrollable factors that are outside the decision maker’s control and
that may occur for each alternative course of action. These uncontrollable factors are
called EVENTS or STATES OF NATURE. For example, in a product launch
situation, possible states of nature could consist of events such as similar product
being launched by a competitor at a lower price or no similar product being launched.
 Each outcome (result) is conditionally dependent on a specific course of action and a
specific state of action.
 The value (payoff) of each possible outcome is measured in terms of the decision
maker’s objectives. Payoffs are normally expressed in monetary terms such as profits
or cash flows, but in some problems the decision maker may be interested in other pay
off such as time and market share etc.
 Selection of a course of action

Risk and Uncertainty

Although the terms ‘risk’ and ‘uncertainty’ may be used interchangeably, they differ in
the following aspect:

‘Risk’ is applied to a situation where there are several possible outcomes and there is
relevant past experience to enable statistical evidence to be produced for predicting the
possible outcomes.

‘Uncertainty’ exists where there are several possible outcomes, but there is little
previous statistical evidence to enable the possible outcomes to be predicted.

There are three risk views:

 A person who is a risk seeker is prepared to take risk; such a person will be
prepared to invest in a high risk project in anticipation of high returns.
 A person who is a risk averse is not prepared to take risk; such a person will not
be prepared to invest in high risk projects. S/he will go for low risk projects
although their returns are small.

3
  A person who is a risk neutral will be indifferent to all available alternatives if
they have the same expected value.

Important factors to note

 Frequently decision makers incorporate uncertainty.
 Probabilities are assigned to uncertain outcomes. The probability is the likelihood that
an event or state of nature will occur.
 Probabilities are expressed in decimal form with values between 0 and 1 or in
percentage form with values between 0% and 100%. That is to say, probabilities
should sum up to 1 or 100%
 Probabilities may be derived from mathematical proofs based on given information,
e.g. historical data. When probability is established from historical data, it is known as
‘objective probability’.
 However, it is unlikely that objective probabilities can be established for business
decision since many past observations or repeated experiments for particular decision
are not possible. Probabilities may then be established by managerial judgment.
Probabilities established in this way are known as ‘subjective probabilities’
 Sometimes probability assignment is just subjective (no mathematical proof)

Illustration 2:
Two proposals have been submitted to the Management of Asaad Company Ltd:

Proposal A:
Invest in textile manufacturing business.

Proposal B:
Invest in transportation business
The expected cash inflows (Revenues) per annum with their respective probabilities from
each proposal are shown in Table 1.

Table 1:
Investment Proposals
Proposal A Proposal B
(Textile Manufacturing) (Transportation)

Probability Cash Inflow Probability Cash Inflow

Million TZS Million TZS
0.10 3,000 0.10 2,000
0.20 3,500 0.25 3,000
0.40 4,000 0.30 4,000
0.20 4,500 0.25 6,000
0.10 5,000 0.10 7,000

CALCULATION
The question is: Which proposal should be accepted and why? In order to be able to answer
this question, we need to carry out calculations of expected values, standard deviation and
coefficient of variation.

 Expected value of proposal A is calculated as:

4
 n
A = ∑ AiPi
i=1

= 0.1 (3,000) + 0.2 (3,500) + 0.4 (4,000) + 0.2 (4,500) + 0.1 (5,000)

= TZS 4,000 million

 Expected value of proposal B is calculated as:

n
B = ∑ BiPi
i=1

0.1 (2,000) + 0.25 (3,000) + 0.3 (4,000) + 0.25 (6,000) + 0.1(7,000)

= TZS 4,350 million

Where:
Ai and Bi denote ith outcomes (Cash inflows) and Pi denotes probability of the ith
outcome.

 However, mere comparison of expected values is an oversimplification. We need to

go further to the calculation of standard deviation, SD (a measure of dispersion) and
coefficient of variation, CV (a measure of relative dispersion).

For a single variable:

SDA = √ ∑ (A – A)
n
i
2
Pi
i=1

= √ 300,000
= TZS 547.72 million

CVA = SDA
A
= 547.72
4,000

= 0.137

SDB = √ ∑n (B – B) i
2
Pi
i=1

5
 = √ 2,427,500
= TZS 1,558.04 million

CVB = SDB
B

= 1,558.04
4,350
= 0.358
WORKINGS:
For Proposal A:

i Pi Ai Ai – A (Ai – A)2 (Ai – A)2 Pi

1 0.10 3,000 - 1,000 1,000,000 100,000

2 0.20 3,500 - 500 250,000 50,000
3 0.40 4,000 0 0 0
4 0.20 4,500 500 250,000 50,000
5 0.10 5,000 1,000 1,000,000 100,000
300,000
For Proposal B:

i Pi Bi Bi – B (Bi – B)2 (Bi – B)2 Pi

1 0.10 2,000 - 2,350 5,522,500 552,250

2 0.25 3,000 - 1,350 1,822,500 455,625
3 0.30 4,000 - 350 122,500 36,750
4 0.25 6,000 1,650 2,722,500 680,625
4 0.10 7,000 2,650 7,022,500 702,250
2,427,500

 You can tabulate the results obtained from the illustration as follows:

PROPOSAL Expected value Standard deviation Coefficient of

TZS million TZS million variation
A 4,000 547.72 0.137

B 4,350 1,558.04 0.358

Argument and Decision:

Proposal B results into greater expected value than proposal A, but the cash inflows are more
dispersed and the proposal carries a greater risk (coefficient of variation being measure of
risk).

Implement proposal A.
It should be born in mind that predictions are more subjected to uncertainty than historical
events.
6
General approach to uncertainty
 Suppose in illustration 1, the manager of accompany is faced with equal likelihoods of
success or failure regarding doing nothing and re-arranging the labour force and the
shown cost data is given, the one-column decision table becomes a two column
decision table
EVENT Success Failure Expected value of
cost
Probability of event 0.5 0.5
Cost Cost
ACTIONS:
Do nothing TZS 200,000 TZS 200,000 TZS 200,000
Re-arrange TZS 100,000 TZS 260,000 TZS 180,000

 The manager would decide to re-arrange the labour force since re-arrangement has a
lower expected labour cost, TZS 180,000
 By selecting the re-arrangement option, the company expects to save TZS 20,000 i.e.
TZS 200,000 – 180,000. If the manager thinks that the amount of saving is
reasonable, then it will be wise to select the re-arrangement option.
 The accountant provides most of the data required in these decision models.

Buying perfect information

 Suppose a consultant can provide perfect information (information that gives absolute
certainty) to the company that the ‘do nothing’ option will fail and the re-arrangement
option will succeed, what would be the maximum amount that the company would be
willing to pay for the consultant’s wisdom?
 We can prepare a decision table with perfect information.

EVENT Success Failure Expected value of

cost
Probability of event 0.5 0.5
Cost Cost
ACTIONS:
Do nothing TZS 200,000 TZS 100,000
Re-arrange TZS 100,000 TZS 50,000
Total TZS 150,000

 Since the perfect information that the company is considering buying will contain one
of the two predictions, the expected value with perfect information will be:
(TZS 200,000 x 0.5) + (TZS 100,000 x 0.5) = TZS 150,000

 Value of perfect information:

Expected value with existing information 180,000
Expected value with perfected information 150,000
Value of perfect information 30,000

The maximum amount that can be paid t to a consultant in order to obtain perfect
information is TZS 30,000.

7
Buying imperfect information
 In reality, it may be difficult to obtain perfect information
 But, imperfect information may still be worth buying
 The consultant will have to consider each of the three possible reports:
o Neutral
o Optimistic
o Pessimistic
The optimistic and pessimistic reports would change the manager’s assessment of
probabilities of success or failure and hence change the decision table:

EFFECT OF CHANGE IN PROBABILITY

EVENT Success Failure Expected value of
costs
Optimistic Report:
Probability of event 0.8 0.2

Cost Cost
ACTIONS:
Do nothing TZS 200,000 TZS 200,000 TZS 200,000
Re-arrange TZS 100,000 TZS 260,000 TZS 132,000

Pessimistic Report:
Probability of event 0.2 0.8

Cost Cost
ACTIONS:
Do nothing TZS 200,000 TZS 200,000 TZS 200,000
Re-arrange TZS 100,000 TZS 260,000 TZS 228,000

 Therefore, the expected values with imperfect information would be:

TZS
o If neutral (The existing situation), 180,000
Re-arrange
o If optimistic, re-arrange 132,000
o If pessimistic, do nothing 200,000

 Suppose the manager assesses a probability of 40% neutral, 30% optimistic and 30%
pessimistic, then we can proceed determining the maximum price of the imperfect
information by preparing the following decision table:

EVENT Neutral Optimistic Pessimistic Expected value

(TYPE OF of costs
REPORT)
Probability 0.4 0.3 0.3

ACTIONS:
Do nothing TZS 200,000 60,000
Re-arrange TZS 180,000 TZS 132,000 111,600
Total 171,600

8
  Value of imperfect information
TZS
Expected value with existing information 180,000
Expected value with imperfect information 171,600
Value of imperfect information 8,400

The maximum amount that can be paid to the consultant in order to obtain imperfect
information is TZS 8,400.

EXPECTED MONETARY VALUE AND UTILITY

 In most business cases, expected monetary value is a useful guide to action.

 However, there are instances where expected monetary value will not be governing
e.g. utility value and fear of bankruptcy will be governing.

Illustration 3:
Each of two managers has been given an opportunity to submit his proposal which costs TZS
10,000 to prepare. There is a 50 – 50 chance that the proposal will be accepted, in which case,
there would be a TZS 25,000 after deducting all associated costs (including the TZS 10,000
proposal cost). If you were one of the managers, would you submit the proposal?

In order to be able to answer this question, let us prepare the following decision table:

EVENT Proposal accepted Proposal not Expected value

accepted
Probability 0.5 0.5

Net Profit Net Profit

ACTION:
Submit proposal TZS 25,000 -TZS 10,000 TZS 7500
No proposal 0 0 0

The two managers may have different attitudes towards the situation:
(i) The manager who does not have adequate capital (funds) may decide to forego the
opportunity.

(ii) The managers who has adequate capital may decide to take the opportunity

(iii) The decision of each manager will also depend on their conceived utility value of
dollar.

(iv) The manager who is prepared to take risk may decide to take the risk of losing
TZS 10,000 and take the opportunity.

(v) The manager who is not prepared to take risk (risk averse manager) may decide to
avoid risk of losing TZS 10,000 and forego the opportunity.

9
COST OF PREDICTION ERROR
Cost of prediction error = Expected financial results under an alternative parameter value

LESS
Expected financial results under the original prediction
under the alternative parameter value

A parameter is a constant or the coefficient of some variable in a model or system of

equations. Examples of parameter are:
 Total fixed costs
 Quantity demanded
 Unit variable cost
 Unit selling price

Illustration 4:
You are given the following information about product M:
Per unit
TZS
Selling price 90
Variable cost 50

Total fixed costs of production are TZS 20,000

The sales manager predicted that 1,000 units of product M would be purchased and sold
during April, 2008. Unfortunately, due to the unforeseen competition in the market, only 600
units were sold. The Manager had a privilege of returning the unsold products at no cost and
getting back a refund for the price paid. What was the cost of prediction error- that is, the cost
of the Sales Manager’s failure to predict demand accurately?

SOLUTION:
(1) Initial predicted sales = 1,000 units
Optimal original decision: Purchase 1,000 units
Expected net income = (1,000 x TZS 40 contribution)
Less: TZS 20,000 fixed costs
= TZS 20,000

(2) Alternative parameter value (actual sales) = 600 units

Optimal decision: Purchase 600 units
Expected net income = (600 x TZS 40 contribution)
Less: TZS 20,000 fixed costs 4,000

(3) Results of original decision under alternative parameter value:

Expected net income = (600 x TZS 40 contribution)
Less: (TZS 20,000 fixed cost plus 400 units
returned at no cost) 4,000

(4) Cost of prediction error = (2) – (3) 0

10
 Suppose in the illustration, the Manager had no privilege of returning any unsold
products. Instead because of their souvenir marketing they became worthless. What
would be the cost of prediction error?

SOLUTION:

(1) Initial predicted sales = 1,000 units

Optimal original decision: Purchase 1,000 units
Expected net income = (1,000 x TZS 40 contribution)
Less: TZS 20,000 fixed costs
= TZS 20,000

(2) Alternative parameter value (actual sales) = 600 units

Optimal decision: Purchase 600 units
Expected net income = (600 x TZS 40 contribution)
Less: TZS 20,000 fixed costs 4,000

(3) Results of original decision under alternative parameter value:

Expected net income = (600 x TZS 90 revenue)
Less: costs (1,000 units x TZS 50 + TZS 20,000
fixed costs) -16,000

(4) Cost of prediction error = (2) – (3) 20,000

DECISION TREES
 A decision tree is a diagram showing several possible courses of action and possible
events (i.e. state of nature) and the potential outcomes for each course of action.

 Each alternative course of action or event is represented by a branch, which leads to

subsidiary branches for further courses of action or possible events

 Decision trees are useful for decision making conditions of uncertainty

Illustration 5:
A company is considering whether to develop and market a new product. Development costs
are estimated to be TZS 180,000, and there is a 0.75 probability that the development effort
will be successful and a 0.25 probability that the development effort will be unsuccessful.
If the development is successful, the product will be marketed, and it is estimated that:

If the product is very successful, profit will be TZS 540,000

If the product is moderately successful, profit will be TZS 100,000
If the product is a failure, there will be a loss 0f TZS 400,000

Each of the above profit and loss calculations is after taking into account the development
cost of TZS 180,000. The estimated probabilities of each of the above events are as follows:

11
Very successful 0.4
Moderately successful 0.3
Failure 0.3

REQUIRED:
Present the problem in a decision tree, calculate the expected values and select the course of
action to be taken.

Maximin, Maximax and Regret Criteria

In some situations it might not be possible to assign meaningful estimates of probabilities to
possible outcomes. Where these situations occur managers might use any of the following
criteria to make decisions: maximin, maximax or the criterion of regret

Maximin criterion is that the worst possible outcome will always occur and the decision
maker should therefore select the largest payoff under this assumption. In the illustration 6
below, the worst outcomes are TZS 100,000 for machine A and TZS 10,000 for machine B.
So machine A should be purchased using the maximin decision rule.

Maximax criterion is based on the assumption that the best payoff will occur. The highest
payoffs in the illustration 6 are TZS 160,000 for machine A and TZS 200,000 for machine B.
So machine B will be selected under the maximax criterion.

Regret criterion is based on the fact that, having selected an alternative that does not turn out
to be the best, the decision maker will regret not having chosen another alternative when he
or she had he opportunity.

Illustration 6:
You are provided with the following profits data for two machine options in two possible
demand conditions:
Low demand(TZS) High demand(TZS)
Machine A 100,000 160,000
Machine B 10,000 200,000

What will the decision maker decide under maximin, maximax and regret critetia?

If probability of low demand and high demand is currently estimated at 50-50 what amount
of money will the company be willing to pay for perfect prediction of demand?

12
DECISION MAKING UNDER CERTAINTY AND UNCERTAINTY
ILLUSTRATIONS

Illustration 1
Compare the following two alternatives: Manufacturing components and Buying the
components from outside.
Manufacture Buy
100 components 100 components
TZS TZS
Direct Labour 10,000 10,000
Direct Materials 30,000 -
Variable overheads 5,000 -
Fixed overheads 20,000 20,000
Supplier’s purchase price - 50,000
Total 65,000 80,000

What are the relevant costs? And what the decision the management will take?

Illustration 2
Think of a situation whereby we have to make decision on whether to re-arrange our labour
force and incur a labour cost of TZS 1.8 per unit or do nothing and incur a labour cost of TZS
2 per unit. We want to produce 100,000 units. If we are quite certain of these outcomes, then
the probability is 100% or 1 for each outcome. What decision the m managers will take?

Illustration 3
Two proposals have been submitted to the Management of Asaad Company Ltd:

Proposal A:
Invest in textile manufacturing business.

Proposal B:
Invest in transportation business

The expected cash inflows (Revenues) per annum with their respective probabilities from
each proposal are shown in Table 1.

Table 1:
Investment Proposals
Proposal A Proposal B
(Textile Manufacturing) (Transportation)

Probability Cash Inflow Probability Cash Inflow

Million TZS Million TZS
0.10 3,000 0.10 2,000
0.20 3,500 0.25 3,000
0.40 4,000 0.30 4,000
0.20 4,500 0.25 6,000
0.10 5,000 0.10 7,000

The question is: Which proposal should be accepted and why?

13
Illustration 4
Suppose in illustration 1, the manager of accompany is faced with equal likelihoods of
success or failure regarding doing nothing and re-arranging the labour force and the shown
cost data is given in the table below. What will the manager decide?

EVENT Success Failure

Probability of event 0.5 0.5
Cost Cost
ACTIONS:
Do nothing TZS 200,000 TZS 200,000
Re-arrange TZS 100,000 TZS 260,000

Illustration 5

Suppose in the illustration 4 a consultant can provide perfect information (information that
gives absolute certainty) to the company that the ‘do nothing’ option will fail and the re-
arrangement option will succeed, what would be the maximum amount that the company
would be willing to pay for the consultant’s wisdom?

Illustration 6
The optimistic and pessimistic data is given in the table below:

EVENT Success Failure

Optimistic Report:
Probability of event 0.8 0.2

Cost Cost
ACTIONS:
Do nothing TZS 200,000 TZS 200,000
Re-arrange TZS 100,000 TZS 260,000

Pessimistic Report:
Probability of event 0.2 0.8

Cost Cost
ACTIONS:
Do nothing TZS 200,000 TZS 200,000
Re-arrange TZS 100,000 TZS 260,000

Suppose the manager assesses a probability of 40% neutral, 30% optimistic and 30%
pessimistic, what is the maximum price of the imperfect information?

Illustration 7

Each of two managers has been given an opportunity to submit his proposal at a cost TZS
10,000. There is a 50 – 50 chance that the proposal will be accepted, in which case, there
would be a TZS 25,000 after deducting all associated costs (including the TZS 10,000
proposal cost). If you were one of the managers, would you submit the proposal?

14
Illustration 8
You are given the following information about product M:
Per unit (TZS)
Selling price 90
Variable cost 50

Total fixed costs of production are TZS 20,000

The sales manager predicted that 1,000 units of product M would be purchased and sold
during April, 2008. Unfortunately, due to the unforeseen competition in the market, only 600
units were sold. The Manager had a privilege of returning the unsold products at no cost and
getting back a refund for the price paid. What was the cost of prediction error- that is, the cost
of the Sales Manager’s failure to predict demand accurately?

Illustration 9
Suppose in the illustration 8, the Manager had no privilege of returning any unsold products.
Instead because of their souvenir marketing s, they became worthless. What would be the
cost of prediction error?

Illustration 10
A company is considering whether to develop and market a new product. Development costs
are estimated to be TZS 180,000, and there is a 0.75 probability that the development effort
will be successful and a 0.25 probability that the development effort will be unsuccessful.
If the development is successful, the product will be marketed, and it is estimated that:

If the product is very successful, profit will be TZS 540,000

If the product is moderately successful, profit will be TZS 100,000
If the product is a failure, there will be a loss 0f TZS 400,000

Each of the above profit and loss calculations is after taking into account the development
cost of TZS 180,000. The estimated probabilities of each of the above events are as follows:

Very successful 0.4

Moderately successful 0.3
Failure 0.3

REQUIRED:
Present the problem in a decision tree, calculate the expected values and select the course of
action to be taken.

Illustration 11:
You are provided with the following profits data for two machine options in two possible
demand conditions:
Low demand(TZS) High demand(TZS)
Machine A 100,000 160,000
Machine B 10,000 200,000

What will the decision maker decide under maximin, maximax and regret critetia?
If probability of low demand and high demand is currently estimated at 50-50 what amount
of money will the company be willing to pay for perfect prediction of demand?

15