MGMT 670 Homework 2
Due 11:59 pm EST on September 9. Submit via Brightspace.
1. Susan, the marketing manager of a consumer electronics company, is considering
whether to introduce a new large screen television model into the market. Susan
is aware that the success of the new model depends on the economy in the coming year
which in turn depends on the result of the upcoming presidential election. According
to her assessment, the incumbent and the challenger are equally likely to win the elec-
tions. A fixed cost of 3 million dollars has been incurred prior to making the decision
to launch the model. The payoff table (in millions) for the two alternative actions is
set up as follows (the fixed cost has been included in the calculation):
Presidential Election Outcome
Introduction of the new model Incumbent Challenger
Yes -36 45
No -3 -3
(a) Find the expected payoffs of the two alternatives. Which is better?
(b) Compute the expected value of the perfect information.
(c) Bob, a well known consultant, proposed a project for Susan to consider. Bob
argued that he could provide a prediction of the election result through a special
survey method. Susan just graduated and received an A grade in MGMT 670
two years ago. She still remembers the correct method of using Bayes’ Theorem.
After interviewing Bob, Susan obtained the following probability estimates:
P(Predicted Winner is Incumbent | Real Winner is Incumbent) = .7
P(Predicted Winner is Challenger | Real Winner is Incumbent) = .3
P(Predicted Winner is Incumbent | Real Winner is Challenger) = .2
P(Predicted Winner is Challenger | Real Winner is Challenger) = .8.
Complete the following decision tree for evaluating the expected payoff with the
prediction result (sample information) from Bob. Compute EVSI.
(d) Suppose Bob has been hired and his prediction is that the incumbent will win.
Should Susan introduce the new TV model? Support your answer.
1
Solution:
(a) EMV of introducing the new model = (-36)(.5) + (45)(.5) = 4.5 EMV of not
introducing the new model = (-3)(.5) + (-3)(.5) = -3 Therefore, introducing the
new model gives a higher expected payoff.
(b) EVwith PI = (-3)(.5) + (45)(.5) = 21. EVPI = 21 – 4.5 = 16.5
(c) EV with SI = .45*(-3) + .55*22.887 =11.24. So EVSI = 11.24 – 4.5 = 6.74.
Figure 1: Solution of decision tree
(d) Susan should not introduce the new TV model, because the expected loss will be
smaller.
2
2. Over the past 10 years, the expected return for the S&P 500 was 5.04% with a standard
deviation of 19.45%, the expected return over that same period for a core bonds fund
was 5.78% with a standard deviation of 2.13%, and the expected return for real estate
investment trusts (REITs) was 13.07% with a standard deviation of 23.17%. The
correlation between the S&P 500 and core bonds is −0.32, the correlation between the
S&P 500 and REITs is 0.74, and the correlation between core bonds and REITs is
−0.04. You are considering portfolio investments.
(a) Construct a portfolio that is 50% invested in an S&P 500 fund and 50% invested in
REITs. In percentage terms, what are the expected return and standard deviation
for such a portfolio?
(b) Construct a portfolio that is 50% invested in a core bonds fund and 50% invested in
REITs. In percentage terms, what are the expected return and standard deviation
for such a portfolio?
(c) Construct a portfolio that is 80% invested in a core bonds fund and 20% invested in
REITs. In percentage terms, what are the expected return and standard deviation
for such a portfolio?
(d) Which of the portfolios in parts (a), (b), and (c) would you recommend to an
aggressive investor? Which would you recommend to a conservative investor?
Why?
Solution:
(a) Construct a portfolio that is 50% invested in an S&P 500 fund and 50% invested in
REITs. In percentage terms, what are the expected return and standard deviation
for such a portfolio?
The expected return of the portfolio is:
E[Rp ] = wS&P · E[RS&P ] + wREIT s · E[RREIT s ]
Where:
wS&P = 0.50, wREIT s = 0.50
E[RS&P ] = 5.04% = 0.0504
E[RREIT s ] = 13.07% = 0.1307
Thus:
E[Rp ] = 0.50 × 0.0504 + 0.50 × 0.1307 = 0.09055 or 9.06%
The variance of the portfolio is:
2 2 2 2
Var(Rp ) = wS&P · σS&P + wREIT s · σREIT s + 2 · wS&P · wREIT s · Cov(S&P, REITs)
Var(Rp ) = (0.50)2 ·(0.1945)2 +(0.50)2 ·(0.2317)2 +2·0.50·0.50·(0.1945×0.2317×0.74)
Var(Rp ) = 0.25 · 0.0378 + 0.25 · 0.0536 + 0.5 · 0.0333
Var(Rp ) = 0.00945 + 0.0134 + 0.01665 = 0.0395
√
σp = 0.0395 ≈ 0.199 or 19.9%
3
(b) Construct a portfolio that is 50% invested in a core bonds fund and 50% invested in
REITs. In percentage terms, what are the expected return and standard deviation
for such a portfolio?
The expected return of the portfolio is:
E[Rp ] = wCore Bonds · E[RCore Bonds ] + wREIT s · E[RREIT s ]
Where:
wCore Bonds = 0.50, wREIT s = 0.50
E[RCore Bonds ] = 5.78% = 0.0578
E[RREIT s ] = 13.07% = 0.1307
Thus:
E[Rp ] = 0.50 × 0.0578 + 0.50 × 0.1307 = 0.09425 or 9.43%
The variance of the portfolio is:
2 2 2 2
Var(Rp ) = wCore Bonds ·σCore Bonds +wREIT s ·σREIT s +2·wCore Bonds ·wREIT s ·Cov(Core Bonds, REITs)
Var(Rp ) = (0.50)2 ·(0.0213)2 +(0.50)2 ·(0.2317)2 +2·0.50·0.50·(−0.04×0.0213×0.2317)
Var(Rp ) = 0.25 · 0.000453 + 0.25 · 0.0536 − 0.00025
Var(Rp ) = 0.000113 + 0.0134 − 0.00025 = 0.013263
√
σp = 0.013263 ≈ 0.115 or 11.5%
(c) Construct a portfolio that is 80% invested in a core bonds fund and 20% invested in
REITs. In percentage terms, what are the expected return and standard deviation
for such a portfolio?
The expected return of the portfolio is:
E[Rp ] = wCore Bonds · E[RCore Bonds ] + wREIT s · E[RREIT s ]
Where:
wCore Bonds = 0.80, wREIT s = 0.20
E[RCore Bonds ] = 5.78% = 0.0578
E[RREIT s ] = 13.07% = 0.1307
Thus:
E[Rp ] = 0.80 × 0.0578 + 0.20 × 0.1307 = 0.04624 + 0.02614 = 0.07238 or 7.24%
The variance of the portfolio is:
2 2 2 2
Var(Rp ) = wCore Bonds ·σCore Bonds +wREIT s ·σREIT s +2·wCore Bonds ·wREIT s ·Cov(Core Bonds, REITs)
Var(Rp ) = (0.80)2 · (0.0213)2 + (0.20)2 · (0.2317)2 + 2 · 0.80 · 0.20 · (−0.000197)
Var(Rp ) = 0.64 × 0.00045369 + 0.04 × 0.05370929 − 0.00006304
Var(Rp ) = 0.00029036 + 0.00214837 − 0.00006304 = 0.00237569
√
σp = 0.00237569 ≈ 0.04873 or 4.87%
4
(d) Which of the portfolios in parts (a), (b), and (c) would you recommend to an
aggressive investor? Which would you recommend to a conservative investor?
Why?
• For an aggressive investor, I would recommend the portfolio constructed in
part (b), which has 50% invested in a core bonds fund and 50% invested in
REITs. This portfolio has the highest expected return of 9.43%, though it
also comes with the moderate risk, indicated by the standard deviation of
11.5%.
• For a conservative investor, I would recommend the portfolio constructed in
part (c), which has 80% invested in a core bonds fund and 20% in REITs.
This portfolio has a lower expected return of 7.24% but with significantly
lower risk, with a standard deviation of 4.87%.
3. According to a survey, one out of five investors have exchange-traded funds in their
portfolios. Consider a sample of 20 investors.
(a) Compute the probability that exactly 4 investors have exchange-traded funds in
their portfolios.
(b) Compute the probability that at least 2 of the investors have exchange-traded
funds in their portfolios.
(c) If you found that exactly 12 of the investors have exchange-traded funds in their
portfolios, would you doubt the accuracy of the survey results?
(d) Compute the expected number of investors who have exchange-traded funds in
their portfolios.
Solution:
(a) Compute the probability that exactly 4 investors have exchange-traded funds in
their portfolios.
The probability mass function of the binomial distribution is:
n k
P (X = k) = p (1 − p)n−k
k
For k = 4: 4 16
20 1 4
P (X = 4) =
4 5 5
Calculating this:
20 20!
= = 4845
4 4!(20 − 4)!
4
1
= 0.0016
5
5
16
4
= 0.0281
5
P (X = 4) = 0.218
(b) Compute the probability that at least 2 of the investors have exchange-traded
funds in their portfolios.
To find P (X ≥ 2), we use:
P (X ≥ 2) = 1 − P (X < 2) = 1 − (P (X = 0) + P (X = 1))
First, compute P (X = 0) and P (X = 1):
0 20 20
20 1 4 4
P (X = 0) = = ≈ 0.0115
0 5 5 5
1 19
20 1 4
P (X = 1) =
1 5 5
20
= 20
1
1
1
= 0.20
5
19
4
≈ 0.0144
5
P (X = 1) = 20 × 0.20 × 0.0144 ≈ 0.058
Thus:
P (X < 2) = P (X = 0) + P (X = 1) ≈ 0.0115 + 0.058 = 0.0695
P (X ≥ 2) = 1 − 0.0695 = 0.9305
(c) If you found that exactly 12 of the investors have exchange-traded funds in their
portfolios, would you doubt the accuracy of the survey results?
To assess if 12 investors is an unusual result, we calculate P (X = 12):
12 8
20 1 4
P (X = 12) = = 8.66 × 10−5
12 5 5
The probability is extremely small, so observing 12 investors is highly unusual
and would likely cast doubt on the accuracy of the survey.
6
(d) Compute the expected number of investors who have exchange-traded funds in
their portfolios.
The expected value E[X] for a binomial distribution is given by:
E[X] = n · p
1
E[X] = 20 · =4
5
So, the expected number of investors with exchange-traded funds is 4.
4. To qualify for membership in Mensa, the international high-IQ society, an individual
must score in the top 2% of the population on a Full Scale IQ (FSIQ) test. Given that
IQ scores are normally distributed with a mean of 100 and a standard deviation of 16,
what IQ score is required to meet this criterion? Moreover, what is the chance that a
randomly selected person has an IQ score above 140?
Solution: (1) To determine the IQ score required to be in the top 2% of the population
with a mean of 100 and a standard deviation of 16, follow these steps:
(a) Find the z-score for the 98th percentile: The z-score corresponding to the
98th percentile is approximately z = 2.055 (it’s also OK if you use z = 2.05 or
z = 2.06).
(b) Convert the z-score to an IQ score: Use the formula
X =µ+z·σ
where:
• µ is the mean (100),
• σ is the standard deviation (16),
• z is the z-score (2.055).
Plugging in the values:
X = 100 + 2.055 · 16
X = 100 + 32.88
X = 132.88
Therefore, a person would need to score approximately 133 (rounding up to the
nearest whole number) to be in the top 2% of the population and qualify for
Mensa.
(2) To find the probability of having an IQ score above 140:
(a) Calculate the Z-score for an IQ of 140:
X −µ
Z=
σ
where:
7
• X is the IQ score (140),
• µ is the mean (100),
• σ is the standard deviation (16).
Plugging in the values:
140 − 100 40
Z= = = 2.5
16 16
(b)Find the probability corresponding to this Z-score: From standard Z-tables or using
statistical software, the cumulative probability for a Z-score of 2.5 is approximately
0.9938.
Thus, the probability of having an IQ score above 140 is:
P (X > 140) = 1 − P (X ≤ 140) = 1 − 0.9938 = 0.0062
Therefore, the chance that a randomly selected person has an IQ score above 140 is
approximately 0.62%.
5. According to a survey, nearly 40% of adult cell phone owners have downloaded an
application (“apps”) to their cell phone. Anyone who has not downloaded an app
clearly does not use apps. In a random sample of 50 adult cell phone owners, how
likely is it to find that more than 50% have downloaded an app on their phone?
Solution: To determine how likely it is to find that more than 50% of a random sample
of 50 adult cell phone owners have downloaded an app, we use the binomial distribution.
Define:
• Proportion of adults with apps: p = 0.4
• Sample size: n = 50
We want to find P (X > 25) where X is the number of adults with apps in the sample.
Since X follows a binomial distribution X ∼ Binomial(n = 50, p = 0.4), we can use
the normal approximation because n > 20, np = 20 > 5 and n(1 − p) = 30 > 5.
Mean and standard deviation of the binomial distribution:
• Mean: µ = np = 50 × 0.4 = 20
√ √
• Standard deviation: σ = np(1 − p) = 50 × 0.4 × 0.6 ≈ 12 ≈ 3.464
p
Determine the Probability Using the Normal Approximation
We want to find the probability that more than 50% of the sample (i.e., more than 25
individuals) have downloaded an app:
8
P (X > 25) = 1 − P (X ≤ 25)
Using the continuity correction factor (0.5), this is approximated by:
25.5 − 20
P (X > 25.5) = P Z >
3.464
Calculate the z-score:
25.5 − 20
Z= ≈ 1.59
3.464
Using standard normal distribution tables (or a calculator), find the probability that
Z > 1.59:
P (Z > 1.59) ≈ 0.0562
The probability that more than 50% of the sample of 50 adult cell phone owners have
downloaded an app is approximately 0.0562, or 5.62%.
Note: it is also OK if the students did not use the continuity correction factor (0.5),
in that case, the solution would be as follows.
Using the normal approximation, we convert X > 25 to a standard normal variable Z:
X −µ 25 − 20
Z= = ≈ 1.445
σ 3.464
We find the cumulative probability up to Z = 1.445. From the standard normal
distribution table or a calculator:
P (Z ≤ 1.445) ≈ 0.9265
Thus:
P (X > 25) = 1 − P (Z ≤ 1.445) = 1 − 0.9265 = 0.0735
Therefore, the probability that more than 50% of a random sample of 50 adult cell
phone owners have downloaded an app is approximately 0.0735 or 7.35%.
