OFFICIAL (CLOSED) \ NON-SENSITIVE

Lesson 08 – 09
Decision Making and Risk Profile

E210 – Operations Planning

1
 OFFICIAL (CLOSED) \ NON-SENSITIVE

E210 Operations Planning Topic Tree

2
2
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Scenario – Investment Plans

• Mrs. Lim works as a sales manager in a
retail outlet of a large chain store. She
recently received a special year-end
bonus and was trying to find an
investment plan for it.
• After some initial analysis, Mrs. Lim found
3 choices she can invest her bonus in, a
high-risk stock, a low-risk stock or a
savings account. While investing in a • Mrs. Lim understands the risk
savings account would earn a sure $3000, element involved in the investment
the payoff for investing in the two stocks plans. She tends to be more
depends on what would happen to the concerned with the potential huge
market as a whole, and if the market loss if the Market Performance is
performance is good, the payoff would be Poor as she and her husband are
quite substantial. getting old and they are going to
• The following table summarizes the retire soon. In the meantime, she
information that Mrs. Lim has collected. sees the market potential as well.

Investment Plan Market Performance

(Annual Profit/Loss in $1000) Poor Average Good
High-risk Stock -24 8 24
Low-risk Stock -12 12 16
Savings Account 3 3 3

3
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Scenario – Investment Plans

With the information on hand, Mrs. Lim discussed the investment plans with her
family members for opinions.

Alan Amy Mr. Lim

• Mrs. Lim’s son
• Decision: High-risk Stock
• He is always willing to try
his luck and is obviously
drawn to the possibility of
high payoff (24K).
• Mrs. Lim’s daughter
• Decision: Savings Account
• She always tries to play
safe and avoid losses
whenever possible.
• Mrs. Lim’s husband
• Decision: Low-risk Stock
• He neither agrees with
Alan’s optimism nor Amy’s
pessimism.

Mrs. Lim is aware of the use of utility theory in decision

Issues to be addressed in W08.
making when a decision maker’s risk attitude is to be taken • Decision making under
into consideration. Help Mrs. Lim conduct analysis in the uncertainty;
• Decision making under risk.
following contexts:
• Which investment plan should she go for without
considering her risk attitude?
• In view of her risk attitude, is there a systematic method to
decide which investment plan to go for? Why do different Issues to be addressed in W09.
people choose different alternatives based on the same • How to incorporate a decision
maker’s risk attitude into
information given? decision making?
Assume that the probabilities of the various market performance
are: Poor-20%; Average-50%; Good-30%. 4
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Lesson 08
Suggested Solution

5
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Problem Definition
• Decision Maker: Mrs. Lim
• Alternatives: The investment plans
• High-risk Stock
• Low-risk Stock
• Savings Account

• Criteria considered in the Decision Making: Annual

Profit in $1000
• States of nature: Market Performance
(Poor, Average, Good)

6
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Problem Definition
• Decision making under uncertainty
Investment Plan Market Performance
(Annual Profit/Loss in $1000) Poor Average Good
High-risk Stock -24 8 24
Low-risk Stock -12 12 16
Savings Account 3 3 3

• Decision making under risk (the probabilities of the

various market performance: Poor-20%; Average-
50%; Good-30%.

7
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Decision Making – Under Uncertainty

8
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Decision Making – Under Uncertainty

Pessimistic Optimistic
decision maker decision maker

Low-risk Stock gives the

minimum Max regret.

9
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Decision Making Under Risk – EMV and EOL

0.2×(-12K) + 0.5×12K + 0.3x(16K) = 8.4K

Maximum outcome under Opportunity Loss 0.2×(0K) + 0.5×9K +

10
‘Average’ market performance = 12K – 3K = 9K 0.3x21 = 10.8K
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Expected Value of Perfect Information (EVPI)

• At most, how much Mrs. Lim should pay if a consultant can help
her estimate the market performance with 100% accuracy?
Market Performance Probability Optimal Decision Payoff ($K)
Poor 20% Savings Account 3.00
Average 50% Low-risk Stock 12.00
Good 30% High-risk Stock 24.00

The Expected Value with Perfect Information (EVwithPI):

0.2×3K + 0.5×12K + 0.3x24K = $13.8K

• Without engaging the consultant, the ‘Low-risk Stock’ gives the

maximum Expected Monetary Value (Max EMV) of $8.4K.
• The Expected Value of Perfect Information (EVPI)
= EVwithPI – Max EMV
This is the highest amount Mrs. Lim should pay if a
= $13.8K - $8.4K = $5.4K consultant can help her estimate the market
performance with 100% accuracy.
The EVPI of $ 5.4K is also equal to EOL of11the best
alternative: the Low-risk Stock.
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Conclusion
• Mrs. Lim must understand how decisions are made and which
decision-making tools she should use under different conditions.
• What makes the difference between a good and a bad decision:
 Whether the decision is made based on logic;

 Whether all available data and possible alternatives are considered in the decision
making.

• For the problem statement,

 Decision making under uncertainty: Mrs. Lim will select different investment plans
under different strategy (Maximax, Maximin, Equal Likelihood, Hurwicz, MinMax
Regret).

 Decision making under risk: Based on the EMV criterion, Mrs. Lim should consider
selecting the Low-Risk Stock for a maximum EMV of $8,400. The decision is the same
under the EOL criterion

 Mrs. Lim should at most pay the consultant $5,400 for the perfect information.

 Mrs. Lim’s risk attitude has not been analysed and not taken into consideration in her
decision making.
To be covered in W09
12
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Lesson 09
Suggested Solution

13

13
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Suggested Soution
• Decision Maker: Mrs. Lim
• Alternatives: Investment Plans
 High-risk Stock
 Low-risk Stock
 Savings Account
• Criteria considered in the Decision Making:
 Profit/payoff

• States of nature: Market Performance

 Poor
 Average
 Good

14
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Suggested Solution
• Payoffs

• Mrs. Lim’s family members have different opinions

about the investment plans. Who do you think is
really taking the risk?
• Which decision would you advise Mrs. Lim?
• Research and help Mrs. Lim to make decision based
on her willingness to take risk. Incorporate her
attitude towards risk into the decision.
15
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Payoff Table with EMV

Applying what we learned in W08, based on the highest EMV, Mrs. Lim
should choose “Low-risk Stock”.
However, depending on an individual’s willingness to take the risk of losing
money, the decision could be different.
• Alan is always willing to try his luck and is drawn to the possibility of
high payoff ($24K). He will go for the “High-risk Stock”.
• Amy, on the other hand, is more unwilling to risk losing money. She
will decide that “Savings Account” is the best alternative.
• Mr. Lim, neither agrees with his son’s optimism nor his daughter’s
pessimism. He will go for the “Low-risk Stock”.

16
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Utility Function Assessment – Alan

• To determine the utility for $12K (payoff), we first assign a
utility value of 0 to the worst outcome (-$24K) and a utility
value of 1 to the best outcome ($24K).
i.e. U(-$24K) = 0; U($24K) = 1

• Then we must identify the probability p at which Alan is

indifferent between the following 2 alternatives:
 Alternative1: Receive $12K with certainty
 Alternative 2: Receive $24K with probability
p and lose $24K with probability (1-p)

17
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Utility Function Assessment – Alan

Finding the Utility Value
• Assume that, to be indifferent between the alternatives, the p
for $24K has to be at least 0.45.
 If the actual probability is less than 0.45, then Alternative 1
is preferred.
 If the actual probability is greater than 0.45, Alternative 2 is
preferred.
• Utility of $12K (payoff) is calculated as follows:
Expected utility of Alternative 1 = Expected utility of Alternative 2
U($12K) = U($24K)(p) + U($-24K)*(1-p)
Alan's Utility Function
= (1)(p) + (0)(1-p) Amount of Payoff ($’000) Utility Value U(X)
-24 0

=p -12
3
8
= 0.45 12 0.45
16
• Repeat the steps to determine 24 1

Alan’s utility values for the other payoffs.

18
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Comparison of Utility Functions

• Utility values for Alan, Amy, Mrs. Lim and Mr. Lim were given in
tables below.
Alan's Utility Function Amy's Utility Function
Amount of Payoff ($’000) Utility Value U(X) Amount of Payoff ($’000) Utility Value U(X)
-24 0 -24 0
-12 0.05 -12 0.64
3 0.2 3 0.93
8 0.32 8 0.96
12 0.45 12 0.98
16 0.65 16 0.99
24 1 24 1

Mrs. Lim's Utility Function Mr. Lim's Utility Function

Amount of Payoff ($’000) Utility Value U(X) Amount of Payoff ($’000) Utility Value U(X)
-24 0 -24 0
-12 0.53 -12 0.24
3 0.86 3 0.52
8 0.9 8 0.64
12 0.92 12 0.72
16 0.94 16 0.83
24 1 24 1
19
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Utility Curves
• Once we determined the utility for payoffs between losing $24K
and a profit of $24K for Mrs. Lim and her family members, we
can establish the general shape of the utility functions.

Utility Curves for Mrs. Lim and her family From the four
1 curves, we can
see that Alan is
0.8 risk seeking,
Mrs. Lim and
0.6 Amy
Amy are risk
Utility

Alan averse, and Mr.

0.4
Mrs. Lim Lim is risk
Mr. Lim neutral.
0.2

0
-30 -20 -10 0 10 20 30
Payoff ($'000)

20
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Maximizing Expected Utility – Alan

Original Payoff Table
To calculate the
Expected Utility (EU),
apply the same
approach as finding the
EMV.

E.g. High-risk Stock

=0.2*0+0.5 *0.32+0.3*1
=0.46

Utility Function determined via Certainty Equivalence The best decision would
be to select “High-risk
Investment Plan Market Performance Stock” as it has the
(Utility Value) Poor Average Good Expected Utility
20% 50% 30%
highest expected utility.
High-risk Stock 0 0.32 1 0.46
Low-risk Stock 0.05 0.45 0.65 0.43
Savings Account 0.2 0.2 0.2 0.20

21
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Maximizing Expected Utility – Amy

Original Payoff Table

To calculate the Expected

Utility (EU), apply the same
approach as finding the EMV.

Eg. Low-risk Stock

=0.2*0.64+0.5*0.98+0.3*0.99
Utility Function determined via Certainty Equivalence =0.92

The best decision would be to

Investment Plan Market Performance
select “Savings Account” as it
(Utility Value) Poor Average Good Expected Utility
has the highest expected
20% 50% 30%
utility.
High-risk Stock 0 0.96 1 0.78
Low-risk Stock 0.64 0.98 0.99 0.92
Savings Account 0.93 0.93 0.93 0.93

22
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Using Exponential Utility Function – Amy

• Assume that Amy has a Risk Tolerance (R) of $5K.
Determine the best decision based on the Exponential Utility
Function.
Investment Plan Market Performance
(Utility Value) Poor Average Good Expected Utility
20% 50% 30%
High-risk Stock -120.51 0.80 0.99 -23.41
Low-risk Stock -10.02 0.91 0.96 -1.26
Savings Account 0.45 0.45 0.45 0.45

The best decision here would be “Savings Account” as it has the highest
expected utility.

23
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Maximizing Expected Utility – Mrs. Lim

Original Payoff Table

To calculate the Expected Utility

(EU), apply the same approach
as finding the EMV.

E.g. Low-risk Stock

=0.2*0.53+0.5 *0.92+0.3*0.94
=0.85
Utility Function determined via Certainty Equivalence
The best decision would be to
select “Savings Account” as it
Investment Plan Market Performance
(Utility Value) Poor Average Good Expected Utility has the highest expected utility.
20% 50% 30%
High-risk Stock 0 0.9 1 0.75
Low-risk Stock 0.53 0.92 0.94 0.85
Savings Account 0.86 0.86 0.86 0.86

24
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Maximizing Expected Utility – Mr. Lim

Original Payoff Table

To calculate the Expected

Utility (EU), apply the same
approach as finding the EMV.

Eg. Low-risk Stock

=0.2*0.24+0.5 *0.72+0.3*0.83
Utility Function determined via Certainty Equivalence =0.66

The best decision would be to

Investment Plan Market Performance
select “Low-risk Stock” as it
(Utility Value) Poor Average Good Expected Utility
has the highest expected
20% 50% 30%
High-risk Stock 0 0.64 1 0.62 utility.
Low-risk Stock 0.24 0.72 0.83 0.66
Savings Account 0.52 0.52 0.52 0.52

25
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Calculation of Risk Premium

• Assume that Alan is indifferent between the following 2
alternatives:
Alternative 1: Get $2500 with certainty;
Alternative 2: 10% chance of losing $5000, 30% chance of earning
$2500, and 60% chance of earning $3000.
• The expected monetary value for Alternative 2 is:
EMV = 0.1*(-$5000) + 0.3*($2500) + 0.6*($3000) = $2050
• The risk premium = $2050 – $2500= -$450
• Alan is willing to bear higher risk.
Though the Certainty Equivalent is $450 higher than the EMV of the
risky Alternative 2, he is willing to risk losing this $450: forgo the sure
amount of $2500 and choose the uncertain Alternative 2 in
anticipation of gaining $3000.
He is not willing to accept a Certainty Equivalent of $2500 but prefers
the risky Alternative 2 that has a lower EMV of $2050.
 Hence, Alan is Risk Seeking (Risk Premium is negative).

26
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Calculation of Risk Premium

• Assume that Amy is indifferent between the following 2 alternatives:
Alternative 1: Get $1500 with certainty;
Alternative 2: 10% chance of losing $5000, 30% chance of earning $2500, and
60% chance of earning $3000.

• The expected monetary value for Alternative 2 is:

EMV = 0.1*(-$5000) + 0.3*($2500) + 0.6*($3000) = $2050
• The risk premium = $2050 – $1500= $550
• Amy prefers lower risk:
 She is willing to accept a Certainty Equivalent of $1500 to avoid the risk in the uncertain
Alternative 2 that has a higher EMV of $2050.

 She is willing to forgo $550 out of the expected value of $2050 to avoid the risk of losing
$5000. Hence, Amy is risk averse (Risk Premium is positive).

• Similar to Amy, Mrs. Lim is also risk averse. She will have a positive
Risk Premium as well.

27
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Calculation of Risk Premium

• Assume that Mr. Lim is indifferent between the following 2
alternatives:
Alternative 1: Get $2050 with certainty;
Alternative 2: 10% chance of losing $5000, 30% chance of earning
$2500, and 60% chance of earning $3000.
• The expected monetary value for Alternative 2 is:
EMV = 0.1*(-$5000) + 0.3*($2500) + 0.6*($3000)
= $2050
• The risk premium = $2050 – $2050= $0
• Mr. Lim is risk neutral (Risk Premium is 0).

28
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Recommendations
• As Mrs. Lim is risk averse, she should consider Amy’s
recommendation of “Savings Account” rather than
Alan’s recommendation of “High Risk Stock” or Mr.
Lim’s recommendation of “Low Risk Stock” as her
investment plan.
• Other considerations, such as Mrs. Lim’s long-term
investment strategies, market performance, etc.,
should also be included.

29
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Learning Outcomes
• Apply appropriate decision making criteria (Maximax, Minimax, Minimax
Regret, Equal likelihood, Criterion of Realism/Hurwizc) in decision
making under uncertainty when probabilities of occurrences of future
events are unknown.
• Apply Expected Monetary Value (EMV) and Expected Opportunity Loss
(EOL) methods in decision making under risk when probabilities of
occurrences of future events can be estimated.
• Calculate Expected Value of Perfect Information (EVPI) to limit what
decision maker should be spending on perfect information.
• Interpret the common utility curves to differentiate types of decision
maker (risk-averse, risk-neutral and risk-seeker).
• Apply Certainty Equivalent method to construct the utility function of any
decision maker.
• Calculate Risk Premium to determine a decision maker’s risk attitude.
• Apply Exponential Utility Function to calculate a risk-averse decision
maker’s utility.
• Calculate expected utility to determine the appropriate decision for a
decision maker.
30
 OFFICIAL (CLOSED) \ NON-SENSITIVE

Overview of E210 Operation Planning Module

Decision
Operations LP – Integer Making using
Planning Programming Decision Tree
Overview
LP -
Transportation Network
Model Diagram and
Decision
Making with Min. Spanning
Dependent Tree
Decisions

LP formulation LP – Binary
Process Strategy and EXCEL Integer
& Capacity Solver with Programming Decision
Requirements Sensitivity Making using
analysis Utility Dissimilarity
Function Index and MST

Linear
Programming Decision
(LP) Graphical Making under
Method uncertainty
and risk

We are
here !
31