QUANTITATIVE TECHNIQUES - PAPER 1

UNIT 23 - DECISION THEORY

 Chapter 23 Decision Making

Chapter 23
DECISION MAKING
23.1: Introduction
Decision making is a process of selecting from a set of alternative courses of action, the best option that is considered
to meet the objectives of the decision problem more satisfactorily than others as judged by the decision maker.

The best possible decision (or option) is arrived at through abstraction, observation, investigation, experimentation,
measurement and analysis of data in the relevant real world environment.

Therefore a good decision is always based on logic, uses data, applies techniques of quantitative analysis and considers
all possible alternatives.
A good decision may sometimes result in unexpected unfavourable outcomes but it still remains a good decision
whereas a bad decision may sometimes lead to favourable occurrences but still it remains a bad decision.

23.2: Decision making concepts

(i) Decision maker

This is an individual or group of persons charged with the responsibility of making the decision by selecting one best
alternative from a set of possible courses action.

(ii) Alternatives (or acts)

These are events or outcomes that are available to the decision maker and are within the control of the decision maker
e.g. construction of a large or small facility launch or do not launch the product.

(iii) States of nature (SN)

States of nature are the circumstances (events) that affect the outcome of the decision, but are beyond the control of
the decision maker. e.g. Strong market Vs poor market.

(iv) Outcome / payoff

For each combination of an act and event there will exist an outcome. It may be expressed in terms of profits, present
value or in terms of some non-monetary measure. We refer to the outcome as payoff.

23.3: The Decision making process

Before a decision can be taken, the decision maker must carry out the following steps;

(i) Define the problem at hand.

This involves making clear and concise statements about the problem.
(ii) Gather data on the problem.
This involves collection of data about the current awareness of the problem. Several techniques can be
employed to execute this process.
(iii) List alternative courses of action.
Having defined the problem and collected data about it, it is now possible to develop a list of all possible
alternatives to making the decision.
(iv) Identify possible outcomes.
Where possible, all outcomes should be identified for informed decision.

(v) Estimate the profits or payoffs of the alternatives.

Estimation of profits or payoffs will require one’s knowledge of the market behaviour.

(vi) Develop a mathematical model and use the model to generate alternative solutions to the problem.

(vii) Choose from the alternative solutions.

223
 Chapter 23 Decision Making

23.4: The decision making Environments

A decision making environment is simply a combination of all forces that may affect the best alternative choice.
Therefore, it is of significant importance for any decision maker to be aware of such forces governing the decision
environment.

There are basically three types of decision making environments and these include;
(i) Decision making under certainty
(ii) Decision making under risk.
(iii) Decision making under uncertainty

These are explained as below;

(i) Decision making under certainty.

This is a situation in which the decision maker knows for certain the consequences of every alternative taken. The
decision maker will always choose the best alternative to maximise his objective in case of profits or to minimize the
objective in case of loss or costs.

In other words, there would exist only one outcome for the decision maker. For example if an investor knows for sure
that there is better business in a service sector than in farming, then he will invest in the service sector which provides
better returns.

(ii) Decision making under Risk.

This is a situation in which the decision make knows the probability with which each outcome is likely to happen. If the
probability of an occurrence is known, then the decision maker will attempt to maximise his objective by maximizing the
expected monetary value i.e. the EMV or by minimising the expected loss.

(iii) Decision making under Uncertainty.

This is a situation where nothing is known about any choice that may be taken.
Under conditions of uncertainty, the decision maker is constrained because of the lack of knowledge about the
probabilities of occurrence of the outcome of possible courses of action.

For example, a potential investor may not know the political environment of the country in the next five year period to
necessitate him to take a proper investment decision.

23.5: Theorems under decision making Environments

23.5.1 Decision making under risk

There are two criteria that the decision maker may use under the conditions of risk and these include;
(i) Expected monetary value, EMV and
(ii) Opportunity loss

These are explained as below;

(i) Expected monetary value, EMV

Once the conditional values or payoffs and probabilities of each state of nature (SN) and for each alternative are given
(in a decision table), the EMV of each alternative is the product of the payoff of the states of nature by their respective
probabilities.

In other words;

EMV (alternative k) = (payoff of SN1) × Pr(SN1) + payoff of SN2 × Pr(SN2) + ……+

payoff last SN × Pr(last SN).

Expected monetary value is further discussed under two headings;

224
 Chapter 23 Decision Making

Expected value of perfect information (EVoPI)

If the decision maker knows what state of nature was going to prevail, he or she would simply choose the strategy giving
the best return for the state of nature.

The perfect information about the state of nature would help to remove all the uncertainties, and hence the decision
maker’s problems would be solved.
Unfortunately, this is not possible without incurring a cost for market research or some other appropriate method of
investigation about the state of nature.

By finding out how much better off the decision maker would be if he or she had perfect information helps in setting an
upper limit (bound) on what it is worth paying to seek real information.

Definition:
The expected value of perfect information refers to the maximum or minimum payment a decision maker may require
as payment for provision of any technical information that may be employed in order to make any decision environment
change from uncertainty into that of certainty.
In otherwords;

EVoPI = EVwPI - EMVmax

Where;
EVwPI is the Expected monetary value with perfect information.
EMVmax is the maximum expected monetary value.

Expected value of perfect information (EVoPI)

This is the upper price limit (bound) paid for any perfect information used in making the decision.
It is computed using the formula below;

EVwPI = (Best outcome of SN1) × (Probability of SN1) +

(Best outcome of SN2) × (Probability of SN2) + ……. +
(Best outcome of SNn) × (Probability of SNn)

Example 23.1
XYZ Ltd wishes to set up a processing plant in order to launch its newly developed production Uganda. The managers
have come up with the following information about the possible courses of action and the corresponding payoffs.

Alternatives Strong Market Poor Market

Large plant 200,000 -180,000
Small Plant 100,000 -20,000
Do nothing 0 0

Probabilities 0.6 0.4

XYZ Ltd has contacted Kasiita, a consultancy firm, which is willing to charge Shs. 72,000 for the cost of provision of
perfect information about the states of nature.
Required;
(a) Compute
(i) The Expected monetary value (EMV) for each alternative and show the
(ii) The Expected value with perfect information(EVwPI)
(iii) The Expected value of perfect information(EVoPI)

(b) Should XYZ obtain the perfect information from Kasiita? Give reason why

225
 Chapter 23 Decision Making

Solution

(a)
(i) EMV for each alternative; from the formula;

EMV (alternative k) = (payoff of SN1) × Pr(SN1) + payoff of SN2 × Pr(SN2) + ……+

payoff last SN × Pr(last SN).

Where SN1 = Strong market and SN2 = poor market

200,000 0.6 180,000 0.4

120,000 72,000 , !!!

"# 100,000 0.6 20,000 0.4

60,000 8,000 $%, !!! ∗∗∗ '()(*+

,- - ./ 0 0.6 0 0.4 !

(ii) EVwPI ; from the formula

EVwPI = (Best outcome of SN1) × (Probability of SN1) +

(Best outcome of SN2) × (Probability of SN2) + ……. +
(Best outcome of SNn) × (Probability of SNn

Where
• the best outcome of SN1(Strong Market) = 120,000
• the best outcome of SN2(Poor market) = 0
We shall then multiply these with the corresponding probabilities and sum the results to give the EVwPI as shown below;
Hence;

EVwPI = (120,000 × 0.6) + (0 × 0.4)

= Shs. 120.000

(iii) EVoPI ; from the formula

EVoPI = 120,000 – 52,000

= Shs. 68,000/-

(b) From the above computations, XYZ can only pay a maximum of Shs. 68,000, meaning that there is no need
to look for perfect information about the strong markets and poor markets in Uganda from Kasiita since their
charge of Shs. 72,000 is significantly higher than the EVoPI.

(ii) Opportunity loss

For a given event, the difference between the payoff of the most favourable act and some other act may be termed as
loss due to losing the opportunity of choosing the best alternative.

To use the opportunity loss approach to make a decision, one must calculate the Expected opportunity loss (EOL), and
pick the alternative that gives the minimum EOL.

The opportunity loss table is computed and from which the EOL for each alternative is then obtained.

Steps in computing the Expected Opportunity Loss

226
 Chapter 23 Decision Making

There are basically two steps in computing the EOL;

(i) For each SN, subtract each outcome in the column from the best outcome in that column.
Draw the opportunity loss table.
(ii) Compute the EOL for each alternative by multiplying the probability of each SN by the appropriate
opportunity loss value, then select the alternative that gives the minimum EOL.

Example 23.2
Rework example 23.1 and determine;
(i) The EOL for each alternative
(ii) Advise XYZ on the best alternative using the EOL.

Solution

States of Nature
Alternatives Strong Market Poor Market
Large plant 200,000-200,000 0-(-180,000)
Small Plant 200,000-100,000 0-(-20,000)
Do nothing 200,000-0 0-(0)

• Now construct the opportunity loss table;

States of Nature
Alternatives Strong Market Poor Market
Large plant 0 180,000
Small Plant 100,000 20,000
Do nothing 200,000 0

Probabilities 0.6 0.4

• The EOL Workings;

01 0.6 0 0.4 180,000 2%, !!!

01 "# 0.6 100,000 0.4 20,000 3 , !!!*** minimum EOL

01 ,- - ./ 0.6 200,000 0.4 0 4%!, !!!

The minimum EOL is Shs. 68,000/- corresponding to the alternative of building a small plant. Hence XYZ can launch
their new product in Uganda by constructing a small production facility.

Point to note

Decision making under the EOL technique is always the same as that from the EMV approach and that the EVoPI is the
same as the minimum EOL value.

Decision making under uncertainty

Conditions of uncertainty as described before, imply that the decision maker does not know what conditions will prevail
for the working out of his chosen strategy.
For example, a pricing decision on a new product would be eased if the intentions of competitors were known.

The major criterion usually adapted while making decisions under conditions of uncertainty include but not limited to;

Maximin criteria
227
 Chapter 23 Decision Making

This is a pessimistic criterion based upon the conservative approach to assume that the worst possible is going to
happen.

Using this method, we identify the minimum of the payoffs for every alternative and then select the maximum amongst
them; this will be the alternative that maximises the minimum outcome / payoff for every alternative.

Therefore, the decision maker considers each strategy and locates the minimum pay for each and then selects that
alternative which maximizes the minimum payoff.

Example 23.3:
Looking at example 23.1, a decision maker would choose the “do nothing” strategy since it gives the maximum value
of the smallest gains/losses.

The Maximax criterion

The Maximax criterion is based upon extreme optimism. The decision maker selects the particular strategy which
corresponds to the maximum of the maximum payoff for each strategy.

Therefore, the decision maker will proceed by identifying the maximum payoff for each alternative and then follow by
selecting the maximum (greatest) among them; this will be the alternative that maximises the outcome of every
alternative.

Example 23.4:
Looking at example 23.1, a decision maker would choose the “large plant” strategy since it gives the maximum value
of the maximum payoffs.

The Hurwicz criterion

In order overcome the disadvantages of extreme pessimism of maximin and extreme optimism of maximax criterion, K.
Hurwicz introduced the concept of coefficient of optimism or pessimism.

This allows the decision maker to take into account both the maximum and minimum payoff to the degree of optimism
or pessimism. The alternative which maximizes the sum of the weighted payoffs is then selected.

The Laplace criterion

The above three decision rules ignore most of the information in the payoff matrix, as they focus on either the best or
the worst that could happen. The Laplace criterion uses all the information by assigning equal probabilities to the possible
payoffs for each action and then selecting that alternative which corresponds to the maximum expected payoff.

The Minimax regret criterion

It is based upon the assumption that the decision maker might experience ‘regret’ after he has made the decision and
the events have occurred. The criterion follows the opportunity loss approach and proceeds by identifying the alternative
which minimizes the maximum opportunity loss within each alternative.

After computing the opportunity loss values, pick out the maximum per row and the minimax is selected from the
alternative with the minimum value from the column of maximum (row) values.

Example 23.5:
Looking at example 23.2, the opportunity loss table for minimax decision will appear as below;

228
 Chapter 23 Decision Making

States of Nature Maximum in the

Alternatives Strong Market Poor Market row
Large plant 0 180,000 180,000
Small Plant 100,000 20,000 100,000
Do nothing 200,000 0 200,000

The minimum from the column of maximum values is thus; Shs. 100,000/=. The alternative corresponding to this value
is selected. Therefore, the decision maker will choose the “small plant” strategy.

23.6: DECISION TREES

A decision tree is a pictorial display of the various decision alternatives and their corresponding outcomes. All the
possible choices are shown on the tree as branches and the possible outcomes as subsidiary branches.

The difference between a “probability tree” and a “decision tree” is that the probability tree will only show the probability
of an event occurring and will not take into account the outcome (actions) in decision making.

Structure of decision trees

It is a convention to use the symbol ‘ ‘to indicate the decision point and ‘ ‘ to denote the situation of uncertainty
or event or outcome.
Branches coming out of a decision point are immediate mutually exclusive alternative acts (action) open to the decision
maker.

Branches coming from the event point represent all possible outcomes or situation or events.

ILLUSTRATION

D2

Outcome X2

D1

Outcome Y1

D3

When the probabilities of the various events are known, they are written along the corresponding branches.
A decision tree diagram is simply a probability tree with the chosen alternatives indicated at the respective branches.

229
 Chapter 23 Decision Making

Steps in drawing up a decision tree

Step 1; - Draw the tree showing the various decision points and outcome nodes.
The tree is drawn from left to right (i.e. forward pass)

Step 2; - Fill the outcomes values and probabilities

Step 3; - Use the probabilities and outcome values to calculate the expected values at various points so that the correct
decisions are highlighted.
This stage works from right to left (i.e. backward pass)
At this stage the tree shows the expected values at the outcome nodes from which the highest expected values from
the subsequent part of the tree indicate the best decision that can be taken.

The above steps are illustrated in the examples below;

Example 23.6

An oil company may bid for drilling one of the two contracts for drilling in two different areas in Western Uganda. It is
estimated that the profit of $3 million per month would be realized in the first field and $4 million in the second field.
These profit figures were determined ignoring the cost of bidding which amounts to $3,500 for the first field and $5,000
for the second field.

Required:
5
Advise the oil company on which field they should bid for if the probability of winning a contract for the field is and the
67
8
second field is .
9
Solution profit = $3m, Bid cost= $3,500
Pr = 0.7 A

Profit= $4m, Bid cost= $5,000

Pr = 0.6 B

C
Payoffs at the different nodes;

:;<= > 3# 3,500 0.7 $2.1#

:;<= B 4# 5,000 0.6 $2.4#

:;<= C 0

The payoffs are obtained by multiplying the respective probabilities with the net estimated profit. The net estimated profit
is obtained by subtracting the bidding cost from the gross estimated profit.
The company should bid for oil drilling in the second field since it has the highest payoff (expected monetary value).

Example 23.7

230
 Chapter 23 Decision Making

A Japanese automobile manufacture currently produces its best selling U.S model in Japan, but the relative strength of
the Japanese Yen verses the U.S dollar has been making the car very expensive for the U.S market. To ensure lower
and more stable prices, the company is considering the possibility of manufacturing cars for U.S consumers at one of
its American plants. Its payoff table, in millions of dollars, is given below;

States of Nature
Alternatives Weak dollar Probability = Moderate Dollar Strong Dollar
0.3 Probability = 0.5 Probability = 0.2
Japanese Production 10 15 25
American Production 20 18 16

Required
(i) Make a decision tree summary of the table above.
(ii) Compute the expected payoff for each production.
(iii) Compute the expected payoff with perfect information.
(iv) Compute the expected value of perfect information.
Payoffs
Solution
= $3million
Moderate dollar (15 × 0.5)
= $7.5million
1
= $5milion

Weak Dollar (20 × 0.3)

= $6million
Moderate Dollar (18 ×0.5)
= $9million

= $3.2million
2
(i) Expected payoffs

:;<= 4: E F G -HIJ /- 3 7.5 5 $4$. $ KLMML;N

:;<= %: O# /J -HIJ /- 6 9 3.2 $4 . % KLMML;N ∗∗∗ Q

(ii) Expected payoff with perfect information

EVwPI = (Best outcome of SN1) × (Pr SN1) + (Best outcome of SN2) × (Pr SN2) +(Best outcome of SN3) × (Pr SN3)

Where; N1 = Weak Dollar, N2 = Moderate Dollar and N2 = Strong Dollar

Hence;

= (20 × 0.3) + (18 × 0.5) + (25 × 0.2)

= 6 + 9 + 5 = $20 million
(iii) Expected value of perfect information
231
 Chapter 23 Decision Making

EVoPI = EVwPI - EMVmax

= $20 million - $18.2million
= $1.8million

Example 23.8

ABC electronics specializes in the production of modern electronics and consequent sale. The managers recently met
to discuss the prospects of a new equipment and they came up with the following concerning construction of a new
facility.

Strong market Fair market Poor market

Shs Shs Shs
Large size 550,000 110,000 -310,000
Medium size 300,000 129,000 -100,000
Small size 200,000 100,000 -32,000

Probability 0.6 0.3 0.1

Before taking on the construction, ABC electronics may carry out a market survey which will cost Shs. 100,000.

(a) Develop a decision tree for the table above.

(b) Calculate the expected monetary value at respective nodes.
(c) Advise ABC electronics on which alternative to take.

Solution

(a) Decision tree 1

A
2

B 5

6
(b) Expected monetary values

:;. 4: 0.6 550,000 10,000 0.3 110,000 10,000 0.1 310,000 10,000

0.6 500,000 0.3 100,000 0.1 320,000 RST. U%%, !!!

232
 Chapter 23 Decision Making

:;. %: 0.6 300,000 10,000 0.3 129,000 10,000 0.1 100,000 10,000

0.6 290,000 0.3 119,000 0.1 110,000 RST. 4V , 2!!

:;. U: 0.6 200,000 10,000 0.3 100,000 10,000 0.1 32,000 10,000

0.6 190,000 0.3 90,000 0.1 42,000 RST. 4U3, !!

:;. : 0.6 550,000 0.3 110,000 0.1 310,000 RST. UU%, !!!
:;. $: 0.6 300,000 0.3 129,000 0.1 100,000 RST. %! , 2!!

:;. 3: 0.6 200,000 0.3 100,000 0.1 32,000 RST. 4 3, !!

(c) ABC should take the alternative with maximum EMV i.e. the alternative of Node 4 = “Do no survey and construct
a large facility.

Advantages of decision trees

(i) It clearly brings out implicit assumptions and calculation for all to see, question and revise.
(ii) It allows one to understand, simplify by inspection, various assumptions and alternatives in a graphical
from, which is much easier to understand than the abstract analytical form.

Disadvantages of decision trees

(i) Decision tree diagrams become more and more complicated as the number of decision alternatives increases
and as more variables are increased.
(ii) It becomes highly complicated when interdependent alternatives and dependent variables are present in the
problem.
(iii) It assumes that the utility of money is linear.

23.7: THE EXPECTED VALUE

The expected value of an event is the product of its probability and the outcome or value of the event over a series of
trails.

The probability distribution of a random variable is a list of the values of random variables with their corresponding
probabilities of occurrence.
If + is a random variable which can take any one of the values +4 , +% , … , +N with respective probabilities
X4 , X% , … , XN , then the expected value of X, denoted by E(X) is defined as;

Y G6 Y6 GZ YZ ⋯ G\ Y\

Where; each G] ^ 0 and G6 GZ ⋯ G\ 1

Example 23.9

The revenue curve , R, of Basaza and Sons is known to be _ 20,000` while the cost curve , C, is known to be
a 10,000 b` where ` represents the number of units produced while b represents the variable costs with the
following probabilities;

233
 Chapter 23 Decision Making

B probabilities
10,000 0.1
20,000 0.3
28,000 0.2
Required;
(a) If Basaza expects a production of 1,000 units, compute the expected profit.
(b) Given that the probability of selling 1,000 units is 0.5 and the probability of producing zero units is also 0.5,
compute the expected profit.
Solution

(a)

a 10,000 b`

_ 20,000`

Profit = Revenue – Cost

G -c/ 20,000` 10,000 b`

G -c/ 20,0000` 10,000 b`

G -c/ 20,000 ` 10,000 b`

G -c/ 20,0000 1,000 10,000 b `

G -c/ 20,0000 1,000 10,000 b 1,000

But E(b) is given as in the table below;

b Prob E(b)
10,000 0.1 1,000
20,000 0.3 6,000
28,000 0.2 5,600
12,600

Hence;

G -c/ 20,0000 1,000 10,000 12,600 1,000

G -c/ 20,000,000 10,000 12,600 1,000 Shs. 7,390,000

G -c/ def. 2, UV!, !!!

(b)
Units sold prob Expected sales
1,000 0.5 500
0 0.5 0
∑' + $!!

234
 Chapter 23 Decision Making

G -c/ 20,0000` 10,000 b`

G -c/ 20,000 ` 10,000 b`

G -c/ 20,0000 ` 10,000 b `

E(b) = 12,600, E(x)= 500

Hence;
G -c/ 20,0000 500 10,000 12600 500
Shs. 3,690,000

Properties of expected values

(i) The expected value of a constant is a constant itself i.e; ' h h.

(ii) The expected value of a product of a constant and a random variable is the product of the constant and the
expected value of the random variable. i.e;

' hi h' i
(iii) The expected value of the sum or difference of two random variables is equal to the sum or difference of the
expected values of the individual random variables;

' i j ' i ' j

' i j ' i ' j
(iv) The expected value of a product of two independent random variables is equal to the product of their individual
expected values. i.e;

' ij ' i .' j

(v) The expected value of a random variable is also equal to its mean;

' i k ∑ XL iL

(vi) The variance of a random variable is given by;

Y lZ YZ m Y nZ

∑ `Z ` m∑ ` ` nZ

∑ +% o + k%

(vii) The standard deviation of a random variable is given by;

p √r / J

Example 23.10

ABC Ltd is in the process of introducing its newly developed shaving machine. Data on sales results has been obtained
plus the corresponding probabilities.

Sales 50 100 150 200 250 300

Probabilities 0.10 0.30 0.30 0.15 0.10 0.05

The cost of introduction is Shs. 1,000,000 and unit selling price Shs. 10,000.

235
 Chapter 23 Decision Making

Required
(i) Compute the expected sales of ABC Ltd
(ii) Compute the expected profit from the above data.
(iii) Compute the sales variance and sales standard deviation

Solution

Sales + Probabilities Expected Sales

50 0.10 5
100 0.30 30
150 0.30 45
200 0.15 30
250 0.10 25
300 0.05 15
∑' + 4$!

(i) Expected sales = 150 units.

(ii) Expected profit = Expected sales – cost of introduction

= (150 × 10,000) – 1,000,000
= 1,500,000 – 1,000,000
= Shs. 500,000

(iii) Sales variance and Standard deviation

Sales + G ` `G ` ` ZG `
50 0.10 5 250
100 0.30 30 3,000
150 0.30 45 6,750
200 0.15 30 6,000
250 0.10 25 6,250
300 0.05 15 4,500
∑' + 150 ∑ +% o + 26,750

Sales variance = 26,750 – (150)2 = 4,250 units

Sales standard deviation = s4,250 65.2 units

Example 23.11

Patel, an Asian investor is in the process of considering to invest in either of the two projects A and B. The table below
shows data that was collected and forwarded to him about each project.

PROJECT A PROJECT B
Value Probability Value Probability
Shs “000” Shs “000”
Optimistic outcome 6,000 0.2 7,000 0.1
Most likely outcome 4,000 0.5 5,000 0.6
Pessimistic outcome 3,000 0.3 2,000 0.2

Required
Compute the expected value for each project and advise Patel on which project to invest in order ensure value for his
money.

236
 Chapter 23 Decision Making

Solution

We shall begin by computing the expected value for each project as below;

PROJECT A PROJECT B
Value Prob Expected Value Prob Expected
Shs “000” value Shs “000” value
Optimistic outcome 6,000 0.2 1200 7,000 0.1 700
Most likely outcome 4,000 0.5 2000 5,000 0.6 300
Pessimistic outcome 3,000 0.3 900 2,000 0.3 600
Project EV 4,100 4,300
Basing on the computed expected value, Project B would be preferred as it has the highest value.

Arguments in favour of Expected values

(i) Simple to understand and calculate.

(ii) Represents whole distribution by a simple figure.
(iii) It arithmetically takes into account the expectation of all outcomes.

Arguments against the use of Expected values

(a) By representing the whole distribution by a simple figure it ignores other characteristics possessed by the
distribution such as skewness and range.
(b) It is based on the assumption that the decision maker is risk averse which may not be the case.

EXERCISES 19

19.1
Metaballs manufacturing company Ltd is considering introducing two new products in the market. The company has the
following options;

Option 1: Introduce both products.

Option 2: Introduce either of the products.
Option 3: Introduce none of the products, depending on their performance in the market.

An analysis of the product’s likely performance indicates the probability of a good performance as 30%, fair performance
as 50% and poor performance as 20% and the sales revenue as shown below in the table;

Decision States of nature

Good performance Fair performance Poor performance
Shs “millions” Shs “millions” Shs “millions”
Neither 0 0 0
Product 1 only 10.0 5.2 2.4
Product 2 only 8.4 4.8 2.4
Both 7.6 8.8 3.2
Required;
(i) Compute the expected monetary value for each decision.
(ii) What decision would you recommend?
(iii) Formulate the opportunity loss table.
(iv) Compute the expected opportunity loss for each decision.
(v) Determine the expected value of perfect information.

237
 Chapter 23 Decision Making

19.2 The sales people at Jomayi properties sell up to 9 houses per month. The probability distribution of a sales person
selling x houses in a month is as follows

Sales (x) 0 1 2 3 4 5 6 7 8 9
Prob f(x) 0.05 0.10 0.15 0.20 0.15 0.10 0.10 0.05 0.05 0.05

Required;
Any sales person selling more houses than the amount equal to the mean plus two standard deviation receives a bonus.
Determine the number of houses per month that a sales person should sell to receive a bonus. (Variance = 5.49)

19.3 Homeland Computer company is considering a plant expansion that will enable the company to begin production
of a new computer product. The company’s Executive Director must determine whether to make the expansion a medium
or large scale project. Uncertainty exists in the demand for the new product, which for planning purposes may be low,
medium or high demand. The probability estimates for demand are 0.20, 0.50, and 0.30 respectively. The firm’s
accountants have developed the following annual profit (in thousands of shillings) forecasts for the medium and large
scale expansion projects.
Medium scale expansion Large scale expansion
Profit Probability Profit Probability
Low 50 0.20 0 0.20
Demand Medium 150 0.50 100 0.50
High 200 0.30 300 0.30

Required
(i) Which decision is preferred for the objective of maximizing the expected profit?
(ii) Which decision is preferred for the objective of minimising the risk or uncertainty?
(iii) From your answers in (i) and (ii) above, should the company go for medium scale expansion or the large
scale expansion? Explain.

19.4 Jama investments (U) Ltd recently acquired an option to purchase the shares of Uganda Clays Ltd. The company
can either sell the option to Summit Ltd who has offered a profit of Shs. 22 million or it can exercise the option and sell
the shares in the stock market in three months time. The return on the shares will depend on the state of the stock
market at the time. The stock market could be bullish, constant or bearish. Jama Ltd estimates to make a profit of shs.
320 million if the market is bullish, Shs. 80 million if the market is constant and a loss of Shs. 120 million if the market is
bearish. Jama Ltd also estimates that the probability of the stock market being bullish is 0.1. However the company is
unable to reach a conclusive decision about the probability oF the market remaining constant or being bearish.
Required;
(i) Draw a decision tree for this problem.
(ii) If the probability of the stock market being constant is p, calculate Jama’s expected profit in terms of p.
(iii) For what values of p would it be advisable for Jama Ltd to sell the option?

19.5 Farm products Ltd produces and sells fresh milk. The milk has to be sold within a day otherwise it will get spoilt.
Each distribution crate carries 20 packets of milk. Each crate costs Shs. 300 to produce and sells for Shs. 450. If demand
exceeds supply, a special production of milk is made at a cost of Shs. 400 per crate leaving the selling price constant
while unsold milk at the end of the day is given free of charge to the workers. FPL sells milk in multiples of hundred of
crates. The table below shows the distribution of sale of crates of milk over the last 80 days.

Number of milk Crates sold 100 200 300 400 500

Frequency of sales (days) 8 20 32 12 8

(i) Compute the pay-off table for the above.

(ii) Determine the optimal number of crates of milk that FPL should produce per day using (a) the Maximax
criterion (b) Maximin criterion (c) the expected monetary vary criterion.
(iii) What is the optimal number of crates of milk FPL should produce in order to minimize risk?

238