Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Self Study Problem Set - Solutions

1.
The uncertainty in this problem comes from the salary offer that the student gets. We let Xi be
the random variable that represents the job offer that the students gets in month i, i=1, 2, 3, …. Xi
is a normally distributed random variable with mean $100,000 and standard deviation $10,000.

Let R be the student’s reservation level and let i* be the first month in which the candidate gets a
job offer that exceeds her reservation level. That is, i* is the first month where Xi > R. Then the
student’s total payoff can be written as:
Total payoff = (– 2000)*(i*-1) + Xi*

Note that
1) In months 1,…, i*-1, the candidate incurs a search cost of -$2000. In month i*, the candidate
accepts the job and gets a payoff of Xi*.
2) If i*=1, i.e., the candidate accepts a job offer in the first month itself, then the total payoff is
simply X1.

The following figure shows how the total payoff varies as a function of the reservation level. The
optimal reservation level, R*, is around $105,000.
Total payoff

120000

100000

80000

60000 Total payoff

40000

20000

0
80000 85000 90000 95000 100000 105000 110000 115000 120000
2.
(Note : the simulation results were obtained using a simulation package called Crystal Ball. You
may get slightly different numerical results if you implement the simulation in @Risk)

The demand for trees in a particular year is the uncertain or the random variable in this problem.
Let us call it D. The expected profit of the store is the objective that House store owners want to
maximize. They decide upon the order (call it O) to maximize the expected profit (call it P*).
The problem dictates that each tree is sold at a price of $30 a unit and the store incurs a loss of
$45 on each unsold tree.

First determine the number of trees sold. Notice that, if O > D, then D trees are sold. Similarly if
O < D, then O trees are sold. Thus the number of trees sold is the minimum of O and D. Number
of trees left is zero if O < D and O – min (O, D) if O > D. Now that we know number of trees
sold and leftover,

Profit (P*) = 30 * min (O, D) - 45 * (O – min (O, D)).

Now we are left with the menial job of creating a spreadsheet and simulating. Please refer to
“InventoryPlanning.xls”.

1.
a) Cell E18 in the spreadsheet forecasts the profit when we order 180 trees and demand
has the custom distribution. The expected profit turns out to be $3727.
b) Cell E43 in the spreadsheet forecasts the profit when we order 180 trees and demand is
normally distributed with mean 185 and standard deviation 57.22. We choose 10,000
runs and get the following statistics for profit:

Forecast
Statistic values
Trials 10,000
Mean $3,878
Median $5,400
Mode $5,400
Standard Deviation $2,382
Variance $5,675,314
Skewness -1.81
Kurtosis 6.12
Coeff. of Variability 0.6143
Minimum ($10,501)
Maximum $5,400
Mean Std. Error $24

i. Clearly, the mean or the expected profit is $3878.


ii. 95% confidence interval is ($3878 – 1.96 x 24, $3878 + 1.96 x 24) = ($3832,
$3925).
[Run the same experiment for 1000 runs. Do you get the same answers?]
iii. There are various ways of finding this probability. Observe that this is simply
the probability that a normal random variable with mean 185 and standard
deviation 57.22 exceeds 180. However, something that provides better
insights is a closer look at the frequency chart.

Observe that there is one bar at the end which dominates. This bar
corresponds to the cases when D was indeed greater than the O. When D ≥ O
=180, 180 trees are sold and hence the profit is 180 x $ 30 = $ 5400. So the
probability that demand exceeds quantity is same as the probability that profit
is equal to $5400. We find this from the frequency chart to be 0.53.

2.
To find out the optimum profit first thing we need to do is to come up with reasonable values
of order. When one considers the custom distribution, it is clear that one would never order
below $100 (one has to forego an opportunity to make a sure shot). Ordering above $300
means we will always end up with a surplus. So we take various values from $100 to $300
with increments of $10 to get the following table and graph.
Expected Profit - Trees Ordered

Expected Profit (in $) $4,000


$3,500
$3,000
$2,500
$2,000 Assumption 1
$1,500
$1,000
$500
$0
90 140 190 240 290
Trees Ordered

Expected
Trees
leftover Profit
100 3000
110 3147.825
120 3295.65
130 3443.475
140 3591.3
150 3739.125
160 3735.15
170 3731.175
180 3727.2
190 3723.225
200 3719.25
210 3451.275
220 3183.3
230 2915.325
240 2647.35
250 2379.375
260 1966.05
270 1552.725
280 1139.4
290 726.075
300 312.75

It is easy to discern either from the table in that the optimal order quantity is 150.

Working with the second assumption that the demand is normal, it again seems reasonable to
consider order quantities between 100 and 300. We get the following table and graph.

Trees Expected
leftover Profit
100 $2,868
110 $3,107
120 $3,324
130 $3,515
140 $3,673
150 $3,792
160 $3,869
170 $3,898
180 $3,878
190 $3,808
200 $3,686
210 $3,514
220 $3,293
230 $3,029
240 $2,728
250 $2,392
260 $2,027
270 $1,639
280 $1,232
290 $812
300 $382

Expected Profit - Trees Ordered

$4,000
Expected Profit (in $)

$3,500
$3,000
$2,500
$2,000 Assumption 2
$1,500
$1,000
$500
$0
90 140 190 240 290
Trees Ordered

From the table one finds out that the expected profit attains the maximum value at 170.

This example is instructive as to how nature of assumptions makes a difference to choices we


make. We get two different order quantities, 170 assuming normal distribution and 150
assuming custom distribution. What distribution closely captures real life phenomenon?
It is also worth noting that the difference in the optimal quantity is not vary large and even if
we work with slightly wrong assumptions one reaches at intelligent enough conclusions.
Even if the real world were to have normally distributed demand (which makes more sense)
and we assumed the custom distribution and decide on 150 we are not very worse off.
3.
The random variables in this case are the returns on each stock. Lets RR, RI, and RT denote the
returns on the Infosys, Reliance and Tata‘s stock respectively. We know that
RR is normally distributed with mean 30% and standard deviation 55%. Similarly, RI is normal
with mean 25%, standard deviation 40% and RT is normal with mean 20% and standard
deviation 20%. The return R* from a portfolio choice is

R*(θR ,θI,θT) = θR* RR + θI * RI + θT* RT

Now as outlined in the problems statement we enumerate all possible portfolio choices (θR ,θI,θT)
and simulate to get the variance and the mean return for a particular portfolio. Then we plot the
variance on the X-axis and mean return on the Y-axis to get the mean-variance graph given
below.

Mean-Variance Graph

1.32

1.30

1.28
Expected Return

1.26

1.24

1.22

1.20

1.18
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Variance

4.
(a) P(At least 4 years experience) = P(4 years experience) + P(5 or more years
experience) = 15/100 + 35/100 = 50/100.

(b) P(At least 4 years experience | At least 3 years experience) = P(At least 4
years experience)/P(At least 3 years experience) = (50/100)/(80/100) = 50/80.
5.
Let M be the mileage of a customer.
(a) 130 + 0.2 M = 195 implies M = 325. Bill drove 300 + 325 = 625 miles.
(b) ECR rate is better if and only if M > 625. P(M > 625) = 0.13 + 0.09 + 0.08 =
0.30.
(c) Since DRA is less expensive, that means M <= 625. From the table below, it
follows that E(Cost) = $151.

Miles Cost P(Cost) P(Cost | M <= Cost x P


625)
200 130 0.07 0.10 13
300 130 0.19 0.27 35
400 150 0.23 0.33 49
500 170 0.14 0.20 34
600 190 0.07 0.10 19
Total 0.70 1.00 151

6.
Let F and E respectively denote the events that the first and second balls drawn are black. Now,
given that the first ball selected is black, there are size remaining black balls and five white balls,
and so P(E|F) = 6/11. As P(F) is clearly 7/12, our desired probability is P(E and F) = P(F)*P(E|F)
= 7/12*6/11 = 42/132.

7.
Let X denote the time it takes Helen to complete a homework assignment. P(Go
to bed in time) = P(X < 4) = P(Z < (4 - 3.5)/1.2) = P(Z < 0.42) = 0.6628.

8.
Let D denote the distance traveled.
(a) Since the mean of D is 150 feet and the net is 30 feet long, he should place the
nearest edge of the net at 150 – 15 = 135 feet from the cannon. Therefore, the
net will cover a landing between 135 and 165 feet from the cannon, which is
the interval where the given normal distribution has the most area, and hence
the most probability of landing on the net.
(b) P(135 < D < 165) = 0.9332 - 0.0668 = 0.8664.

9.
Once Monday's bid is made, Newtowne's optimal strategy is to accept the bid if
it is a $3,000,000 bid and reject it if the bid is for $2,000,000. If Monday's bid
is rejected, then accept Tuesday's bid, regardless of the amount offered. The
EMV of this strategy is $2,600,000.

You might also like