

EXPERT ANSWER
a)
Expected value of each case is given by:
EVsmall = Probability of worst case*CF of worst case + Probability of basecase*CF of base case + Probability of best case*CF of best case
EVsmall = 0.1*400 + 0.6*500+0.3*660= 40+300+198 =538
Similarly
EVmedium = 0.1*(-250) + 0.6*650 + 0.3*800 = 605
EV large = 0.1*(-400)+0.6*580+990*0.3 = 605
Thus based on Expected value we would go for Medium or Large size as both are having EV=605
Demand Scenario | A*D | ||||
Center Size | Worst-Case | Base-Case | Best-Case | Expected Value | |
A | Small | 400 | 500 | 660 | 538 |
B | Medium | -250 | 650 | 800 | 605 |
C | Large | -400 | 580 | 990 | 605 |
D | Probability | 10% | 60% | 30% |
b)
Risk Profile calculations | |||||||||||||
1 | 2 | 3 | (1+2+3)/3 | Avg-x | (Avg-x)^2 | sum of (Avg-x)^2*prob | Squareroot (Variance) | ||||||
Center Size | Worst-Case | Base-Case | Best-Case | Average | Worst-Case | Base-Case | Best-Case | Worst-Case | Base-Case | Best-Case | Variances | SD | |
A | Small | 400 | 500 | 660 | 520 | 120 | 20 | -140 | 14400 | 400 | 19600 | 7560 | 86.94826048 |
B | Medium | -250 | 650 | 800 | 400 | 650 | -250 | -400 | 422500 | 62500 | 160000 | 127750 | 357.42132 |
C | Large | -400 | 580 | 990 | 390 | 790 | -190 | -600 | 624100 | 36100 | 360000 | 192070 | 438.2579149 |
D | Probability | 10% | 60% | 30% |
Sample calculation for medium size
Average = (-250+650+800)/3 = 400
Avg-x = 400 – (-250) = 650 (worst)
= 400 – 650 = -250 (base)
= 400- 800 = -400 (best)
(Avg- x)^2 = 6502 = 422500 , (-250)^2 = 62500 , (-400)^2 = 160000
Variance = Prob * (Avg- x)^2
= 422500*0.1 + 0.6 * 62500 + 0.3*160000 = 122750
SD = Sqrt(Var) = 122750^0.5 = 357.42
Caclulation above shows that Standard deviation ie risk for medium size is 357.42 while that for LArge is 438.26. Thus Large size has more risk for same EV. Hence Medium size should be prefered
C)
EV without perfect information = 605,000
Center Size | Worst-Case | Base-Case | Best-Case | |
Small | 400 | 500 | 660 | |
Medium | -250 | 650 | 800 | |
Large | -400 | 580 | 990 | |
Probability | 10% | 60% | 30% | |
Payoff using perfect prediction | 400 | 650 | 990 | Max of above values |
EVwPI | 727 | 400*0.1+650*0.6+990*0.3 |
Expected value with perfect information = 727
Value of perfect Information = Expected value with perfect information – Expected value withou perfect information
Value of perfect Information = 727-605 = 122 ie $122,000
To pursue additional information would benefit the city since it would possibly increase its net cash flow up $122,000
d)
Demand Scenario | sum of A*D | ||||
Center Size | Worst-Case | Base-Case | Best-Case | Expected Value | |
A | Small | 400 | 500 | 660 | 528 |
B | Medium | -250 | 650 | 800 | 515 |
C | Large | -400 | 580 | 990 | 507 |
D | Probability | 20% | 50% | 30% |
This would change the recommendation to small size with EV 528,000
e)
Demand Scenario | sum of A*D | ||||
Center Size | Worst-Case | Base-Case | Best-Case | Expected Value | |
A | Small | 400 | 500 | 660 | 564 |
B | Medium | -250 | 650 | 800 | 710 |
C | Large | -400 | 580 | 990 | 744 |
D | Probability | 0% | 60% | 40% |
This would change recommendation to Large size with EV 744,000
Base case EV= 605,000
Difference = 744,000-605,000 = 139,000 which is less than expenditure
of $150,000 on promotional campaign. Thus i would not recommend promotional campaign based on expected value only
Using the campaign would reduce the net cash flow of all of these centers
but since the mayor is concerned about the risk of losing money, possibly
affecting their re-election, this route would be more favorable to them to eliminate risks.