STATES OF fNATURE

ALTERNATIVES N1: (PROSPERITY) N2: (RECESSION)

(0.6) (0.4)
A1: (BUY NEW) $ 950,000 $ - 200,000

A2: (MODIFY) $ 700,000 $ 300,000

(A) EXPECTED VALUE

EXPECTED VALUE = ∑ (Probability * Corresponding PAYOFF)

EXPECTED VALUE (FOR BUY NEW)

Expected Value = (0.6 * 950,000) + (0.4 * - 200,000)

Expected Value = $ 490,000

EXPECTED VALUE (FOR BUY NEW)

Expected Value = (0.6 * 700,000) + (0.4 * 300,000)

Expected Value = $ 540,000

STATES OF fNATURE

ALTERNATIVES N1: (PROSPERITY) N2: (RECESSION) EMV

(0.6) (0.4)
A1: (BUY NEW) $ 950,000 $ - 200,000 $ 490,000
A2: (MODIFY) $ 700,000 $ 300,000 $ 540,000

MAXIMUM EXPECTED VALUE = $ 540,000

BEST DECISION = A2 (MODIFY)
 (B) MAXIMAX CRITERION

STATES OF fNATURE

ALTERNATIVES N1: (PROSPERITY) N2: (RECESSION) MAXIMAX

(0.6) (0.4)
A1: (BUY NEW) $ 950,000 $ - 200,000 $ 950,000
A2: (MODIFY) $ 700,000 $ 300,000 $ 700,000

MAXIMAX DECISION = A1 (BUY NEW)

MAXIMAX VALUE = $ 950,000

(C) ​MAXIMIN METHOD

STATES OF fNATURE

ALTERNATIVES N1: (PROSPERITY) N2: (RECESSION) MINIMUM

(0.6) (0.4)
A1: (BUY NEW) $ 950,000 $ - 200,000 $ - 200,000
A2: (MODIFY) $ 700,000 $ 300,000 ​ $ 300,000

MAXIMIN DECISION = A2 (MODIFY)

MAXIMIN = $ 300,000
 (D) ​HURWICZ METHOD

Given Coefficient Value (a) = 0.3

EXPECTED VALUE = a * BEST PAYOFF + (1 - a) * WORST PAYOFF

EXPECTED VALUE (FOR BUY NEW)

Expected Value = (0.3 * 950,000) + (0.7 * - 200,000)

Expected Value = $ 145,000

EXPECTED VALUE (FOR MODIFY)

Expected Value = (0.3 * 700,000) + (0.7 * 300,000)

Expected Value = $ 420,000

STATES OF fNATURE

ALTERNATIVES N1: (PROSPERITY) N2: (RECESSION) Expected Value

(0.6) (0.4)
A1: (BUY NEW) $ 950,000 $ - 200,000 $ 145,000
A2: (MODIFY) $ 700,000 $ 300,000 $ 420,000

MAXIMUM EXPECTED VALUE = $ 420,000

BEST DECISION = A2 (MODIFY)
 f STATES OF NATURE

ALTERNATIVES LOW DEMAND MEDIUM DEMAND HIGH DEMAND

RAISE RATES BY 50% -9 4 12

RAISE RATES BY 20% -4 4 8
HOLD RATES CONSTANT -1 3 5
REDUCE RATES BY 20% 1 2 4

(A) MAXIMAX CRITERION

f STATES OF
NATURE

ALTERNATIVES LOW DEMAND MEDIUM HIGH MAXIMUM

DEMAND DEMAND

RAISE RATES BY 50% -9 4 12 12

RAISE RATES BY 20% -4 4 8 8
HOLD RATES CONSTANT -1 3 5 5
REDUCE RATES BY 20% 1 2 4 4

MAXIMAX DECISION = RAISE RATES BY 50%

MAXIMAX VALUE = 12
 (B) MAXIMIN CRITERION

f STATES OF
NATURE

ALTERNATIVES LOW DEMAND MEDIUM HIGH MINIMUM

DEMAND DEMAND

RAISE RATES BY 50% -9 4 12 -9

RAISE RATES BY 20% -4 4 8 -4
HOLD RATES CONSTANT -1 3 5 -1
REDUCE RATES BY 20% 1 2 4 1

MAXIMIN DECISION = REDUCE RATES BY 20%

MAXIMIN VALUE = 1

(C) ​EQUALLY LIKELY DECISION (LAPLACE)

FORMULA::
EQUALLY LIKELY VALUE FOR AN ALTERNATIVE = AVERAGE OF ALL ITS PAYOFFS
UNDER DIFFERENT STATES OF NATURE

f STATES OF
NATURE

ALTERNATIVES LOW DEMAND MEDIUM HIGH EQUALLY

DEMAND DEMAND LIKELY
VALUE
RAISE RATES BY 50% -9 4 12 2.33
RAISE RATES BY 20% -4 4 8 2.66
HOLD RATES CONSTANT -1 3 5 2.33
REDUCE RATES BY 20% 1 2 4 2.33

MAXIMUM EQUALLY LIKELY VALUE = 2.66

EQUALLY LIKELY DECISION = RAISE RATES BY 20%
(D) ​MINIMAX REGRET APPROACH

REGRET TABLE:

f STATES OF
NATURE

ALTERNATIVES LOW DEMAND MEDIUM HIGH MAXIMUM

DEMAND DEMAND

RAISE RATES BY 50% 10 0 0 10

RAISE RATES BY 20% 5 0 4 5
HOLD RATES CONSTANT 2 1 7 7
REDUCE RATES BY 20% 0 2 8 ​ 8

MINIMAX REGRET VALUE = 5

MINIMAX REGRET DECISION = RAISE RATES BY 20%
(E) ​REALISM

Given Coefficient Value (a) = 0.6

EXPECTED VALUE = a * BEST PAYOFF + (1 - a) * WORST PAYOFF

EXPECTED VALUE ( FOR PORTFOLIO A)

EMV = ($ 100 * 0.6) + (1 - 0.6)* $ - 50

EMV = $ 40

EXPECTED VALUE ( FOR PORTFOLIO B)

EMV = ($ 60 * 0.6) + (1 - 0.6)* $ 40

EMV = ​$ 52

EXPECTED VALUE ( FOR PORTFOLIO C)

EMV = ($ 70 * 0.6) + (1 - 0.6)* $ 10

EMV = $ 46

BASED ON HIGH EXPECTED VALUE, BEST DECISION = PORTFOLIO B

BASED ON HIGHEST EMV OF $ 87,000, BEST DECISION IS SCORPION.

(D)

Expected Value of Perfect Information = Expected Value with perfect information -

Maximum Expected Monetary Value

EXPECTED VALUE WITH PERFECT INFORMATION = ∑ (Probability * Corresponding

BEST PAYOFF)

EXPECTED VALUE WITH PERFECT INFORMATION = (0.4* $700,000) + (0.25* $75,000)

+ (0.35 * $550,000)

EXPECTED VALUE WITH PERFECT INFORMATION = $ 491,250

Expected Value of Perfect Information = Expected Value with perfect information -

Maximum Expected Monetary Value

Expected Value of Perfect Information = $ 491,250 - $ 87,000

Expected Value of Perfect Information ​= $ 404,250

Ultra-Smart Research must be willing to pay $ 404,250 for the perfect information
service offered by the marketing firm.
 (A) MAXIMIN CRITERION

f STATES OF
NATURE

ALTERNATIVES VERY AVERAGE

FAVORABLE MARKET UNFAVORABLE MINIMUM
MARKET MARKET

BUILD NEW PLANT $ 300,000 $ 210,000 $ - 280,000 ​ ​$ - 280,000

SUB CONTRACT $ 160,000 $ 100,000 $ - 15,000 $ - 15,000
OVERTIME $ 120,000 $70,000 $ - 8,000 $ - 8,000
DO NOTHING $0 $0 $0 $0

MAXIMIN DECISION = DO NOTHING

MAXIMIN = $ 0

(C)​ EQUALLY LIKELY DECISION

FORMULA::
EQUALLY LIKELY VALUE FOR AN ALTERNATIVE = AVERAGE OF ALL ITS PAYOFFS
UNDER DIFFERENT STATES OF NATURE

f STATES OF
NATURE

ALTERNATIVES VERY AVERAGE

FAVORABLE MARKET UNFAVORABLE EQUALLY
 MARKET MARKET LIKELY VALUE

BUILD NEW PLANT $ 300,000 $ 210,000 $ - 280,000 $ 76,667

SUB CONTRACT $ 160,000 $ 100,000 $ - 15,000 ​$ 81,667
OVERTIME $ 120,000 $70,000 $ - 8,000 $ 60,667
DO NOTHING $0 $0 $0 $0

EQUALLY LIKELY VALUE = $ 81,667

EQUALLY LIKELY DECISION = SUB CONTRACT

STANDARD DEVIATION (FOR MOTEL)

VARIANCE = [(-8,000 - 12,400)^2 * 0.2] + [(15,000 - 12,400)^2 *0.4] + [(20,000 -

12,400)^2 *0.4)]
VARIANCE = 109,040,000
STANDARD DEVIATION = 10,442.222

STANDARD DEVIATION (FOR RESTAURANT)

VARIANCE = [(6,000 - 2,000)^2 * 0.2] + [(6,000 - 8,000)^2 *0.4] + [(6,000 - 6,000)^2 *0.4)]
VARIANCE = 4,800,000
STANDARD DEVIATION = 2190.89

STANDARD DEVIATION (FOR THEATRE)

VARIANCE = [(6,000 - 5,600 )^2 *0.2] + [(6,000 - 5,600)^2 *0.4] + [(5,000 - 5,600)^2 *0.4)]
VARIANCE = 240,000
STANDARD DEVIATION = 489.898
(A) ​CONSERVATIVE (MAXIMIN APPROACH)

STATES OFfNATURE

ALTERNATIVE S1 S2 S3 MINIMUM
(0.4) (0.35) (0.25)

A1 50 80 30 30
A2 35 50 40 35

A3 60 30 50 30

MAXIMIN VALUE = 35
MAXIMIN DECISION = ALTERNATIVE A2

(C) ​BAYE’S DECISION RULE (EMV)

STATES OFfNATURE

ALTERNATIVE S1 S2 S3 EMV
(0.4) (0.35) (0.25)
A1 50 80 30 ​$ 55.50
A2 35 50 40 $ 41.50
A3 60 30 50 $ 47

EXPECTED PAYOFF = ∑ (Probability * Corresponding PAYOFF)

EMV (FOR ALTERNATIVE 1)

EMV = (0.4 * 50 + 0.35 * 80 + 0.25 * 30)

EMV = $ 55.50

EMV (FOR ALTERNATIVE 2)

EMV = (0.4 * 35 + 0.35 * 50 + 0.25 * 40)

EMV = $ 41.50

EMV (FOR ALTERNATIVE 3)

EMV = (0.4 * 60 + 0.35 * 30 + 0.25 * 50)

EMV = $ 47

BASED ON HIGHEST EMV, BEST DECISION IS ALTERNATIVE A1.

REGRET TABLE:
STATES OFfNATURE

DECISION STRONG DEMAND WEAK DEMAND MAXIMUM

SMALL COMPLEX, d1 12 0 12
MEDIUM COMPLEX, d2 6 2 6
LARGE COMPLEX, d3 0 16 16

MINIMAX VALUE = 6
MINIMAX REGRET APPROACH DECISION = CHOOSE MEDIUM COMPLEX

(D) EMV DECISION

STATES OFfNATURE

DECISION STRONG DEMAND WEAK DEMAND EMV

(0.8) (0.2)
SMALL COMPLEX, d1 8 7 7.8
MEDIUM COMPLEX, d2 14 5 12.6
LARGE COMPLEX, d3 20 -9 14.2

EXPECTED COST = ∑ (Probability * Corresponding Cost)

FOR SMALL COMPLEX

EXPECTED COST = (0.8 * 8) + (0.2 * 7) = 7.8

FOR MEDIUM COMPLEX

EXPECTED COST = (0.8 * 14) + (0.2 * 7) = 12.6

FOR LARGE COMPLEX

EXPECTED COST = (0.8 * 20) + (0.2 * -9) = 14.2

SINCE LARGE COMPLEX HAS THE MAXIMUM EMV, THEREFORE CHOOSE LARGE
COMPLEX.

CASELOAD

DECISION MODERATE HIGH VERY HIGH EXPECTED COST

(0.1) (0.3) (0.6)
REASSIGN STAFF 50 60 85 $ 74,000
NEW STAFF 60 60 60 $ 60,000
REDESIGN 40 50 90 $ 73,000

MINIMUM EXPECTED COST = $ 60,000

DECISION = NEW STAFF

(C) ​OPPORTUNITY LOSS TABLE

CASELOAD

DECISION MODERATE HIGH VERY HIGH EOL

(0.1) (0.3) (0.6)
REASSIGN STAFF 10 10 25 $ 19,000
NEW STAFF 20 10 0 $ 5,000
REDESIGN 0 0 30 $ 18,000

EXPECTED OPPORTUNITY LOSS = ∑ (Probability * Corresponding Regret Cost)

FOR REASSIGN STAFF

EXPECTED REGRET COST = (0.1 * 10) + (0.3 * 10) + (0.6 * 25)

EXPECTED REGRET COST = $ 19 (IN THOUSANDS)

FOR NEW STAFF

EXPECTED REGRET COST = (0.1 * 20) + (0.3 * 10) + (0.6 * 0)

EXPECTED REGRET COST = $ 5 (IN THOUSANDS)

FOR REDESIGN

EXPECTED REGRET COST = (0.1 * 0) + (0.3 * 0) + (0.6 * 30)

EXPECTED REGRET COST = $ 18 (IN THOUSANDS)
Expected value of perfect information = Minimum Expected Regret Loss
Expected value of perfect information = $ 5000

Expected Value of Perfect Information = Expected Value with perfect information -

Maximum Expected Monetary Value

Expected Value with perfect information = $ 3,450 - $ 2,650

Expected Value with perfect information = $ 800

EVPI is the maximum amount the investor will be willing to pay for additional
information about a decision problem, i.e., if we have perfect information about the
state of nature before the decision is made, how much is this information worth. In
this case, it is $ 800.

(D) Let payoff for stock under good economy X.

EMV (for stock) = 0.2X + (0.3 * 2000) + (0.5 * 5,000)

EMV (for bond) = $ 2,650

AT POINT OF INDIFFERENCE

EMV (for stock) = EMV (for bond)

0.2X + 600 - 2500 = 2,650
0.2X = 4550
X = $ 22,750

Payoff for Stock under good economy = $ 22,750