UNIVERSITY OF TORONTO

Joseph L. Rotman School of Management

MGT337Y PROBLEM SET #3

SOLUTIONS

1. (a) B and D are not minimum variance efficient portfolios. D is not efficient because
A offers same mean for less variance. As long as A and C are not perfectly correlated,
B will also not be minimum-variance efficient. This is because some combination of
A and C will offer the same mean return yet less variance than B. This is pictured
below.
Minimum Variance Efficient Portfolio
0.25

0.2 C
Expected Return

0.15 B

ρAC = 1

0.1 A D

ρAC = 0.5

0.05
0.1 0.15 0.2 0.25 0.3 0.35
Standard Deviation of Return

(b) False, the higher is a security’s beta (and not its variance), the higher is its expected
return. A security’s variance is made up of two components: (i) market nondiversifiable
risk, and (ii) unique diversifiable risk. In a well diversified portfolio, the unique risk
gets diversified away, leaving only market risk. Only market risk (i.e., beta) requires
a risk premium and thus a higher expected return. It is possible that a security’s
variance is primarily made up of unique risk — in terms of the example, portfolio D’s

1
 30% standard deviation maybe mostly unique risk; and therefore does not require a
high expected return.
(c) The CAPM states

E[R] = Rf + β(E[Rm ] − Rf )
= 0.07 + 0.08β.

Hence, substituting in the values for each security’s E[R], we get

Portfolio A B C D
E[R] 0.10 0.15 0.20 0.10
β 0.375 1.0 1.625 0.375

(d) What is important in terms of risk is the market risk (i.e., beta) of a security. If
this security with zero expected return has a negative beta (in fact, −0.875), then its
expected return is reasonable. Since it provides a hedge against market movements,
it actually requires an expected return less than the risk-free rate. The reason the
security has a low expected return (and thus a high price) is because the security
offsets market losses — it is valuable in that it helps us out when times are very bad.

2. (a) The expected returns on each security are all equal to 17%. As an example of how
to calculate this,

E[R1 ] = 0.4(0.3) + 0.4(0.1) + 0.1(0.1) + 0.1(0.0) = 0.17 or 17%.

The standard deviations on each security are respectively, σ1 = 11%, σ2 = 46.05%,

and σ3 = 10.3%. As an example of how to calculate this,
h i1
σ1 = 0.4(0.3 − 0.17)2 + 0.4(0.1 − 0.17)2 + 0.1(0.1 − 0.17)2 + 0.1(0.0 − 0.17)2 2

= 0.11.

(b) The covariances between the returns are respectively, Cov(R1 , R2 ) = −0.0479,
Cov(R1 , R3 ) = −0.0084, and Cov(R2 , R3 ) = 0.043. As an example of how to calculate
this,

Cov(R1 , R2 ) = 0.4(0.3 − 0.17)(−0.3 − 0.17) + 0.4(0.1 − 0.17)(0.3 − 0.17)

+ 0.1(0.1 − 0.17)(0.5 − 0.17) + 0.1(0.0 − 0.17)(1.2 − 0.17) = −0.0479.

The correlations between the returns of different securities are respectively, Corr(R1 , R2 )
= −0.9456, Corr(R1 , R3 ) = −0.7414 and Corr(R2 , R3 ) = 0.9195. As an example of how
to calculate this,
Cov(R1 , R2 ) −0.0479
Corr(R1 , R2 ) ≡ = = −0.9456.
σ1 σ2 (0.11)(0.4605)

2
 (c) The expected return and standard deviation of a portfolio with equal proportions
in two securities (i and j) is given by
µp = 0.5E[Ri ] + 0.5E[Rj ],
q
σp = 0.25σi2 + 0.25σj2 + 0.5Cov(Ri , Rj ).
Substituting in the values in parts (a) and (b) above gives us the same mean return
on each of the portfolios, i.e. 17%. Substituting in the values in parts (a) and (b),
the standard deviations of the portfolios are (i) σp = 0.1792 for the equally weighted
portfolio of assets 1 and 2, (ii) σp = 0.0384 for the equally weighted portfolio of assets
1 and 3, and (iii) σp = 0.2778 for the equally weighted portfolio of assets 2 and 3.
(d) The expected return and standard deviation of a portfolio with equal proportions
in the three securities is given by
1 1 1
µp = E[R1 ] + E[R2 ] + E[R3 ],
3 3 3
1/2
1 2 2

σp = (σ1 + σ22 + σ32 ) + (Cov(R1 , R2 ) + Cov(R1 , R3 ) + Cov(R2 , R3 )) .
9 9
Substituting in the values in parts (a) and (b) gives us µp = 0.17 and σp = 0.1525.
3. See hw3a.xls for answers.
4. Dow’s asset beta is irrelevant. Since both companies are considering an investment in
natural gas production, both companies are facing similar risk, namely the risk (i.e.,
beta) of natural gas production. Superior Oil’s asset beta is the same as natural gas,
i.e., 0.85 and it is the appropriate risk of the investment in natural gas production.
Both companies, therefore, have the same project cost of capital. From the CAPM,
E[R] = Rf + β(E[Rm ] − Rf )
= 0.075 + 0.85(0.08) = 0.143 or 14.3%.
Both companies will make the same decision to accept or not — they will use the 14.3%
rate to discount future expected cash flows and check whether the NPV is positive.
There is not enough information to definitively evaluate the investment. However, we
know that the NPV of the projects is positive at a 15% discount rate and negative at
a 21% discount rate. Hence, as long as the investment produces expected cash flows
that will be positive in the future (i.e., no change in the signs of expected future cash
flows), the NPV of natural gas production will be positive.
5. (a) Since the electronics company is going to invest in the same line of business as that
of Microchips Inc., the relevant beta (β) will be the asset beta (βA ) of Microchips Inc.
D E
βA = βD + βE
D+E D+E
D
! !
E 1
= D βD + D βE .
E
+1 E
+1

3
 D
Substituting in the values of E
, βE and βD for Microchips Inc.:

0.3 1 1.46
   
βA = 0.2 + 1.4 = = 1.123.
1.3 1.3 1.3
Therefore, from the CAPM,

E[RA ] = 0.08 + 1.12(0.19 − 0.08) = 0.2035 or 20.35%.

The appropriate discount rate is, therefore, 20.35% for the electronics company when
investing in computer chips.
(b) The first step is to calculate the electronics company’s new βA . Once we have its
new βA , we can compute its new βE .
The company’s new βA is a weighted average of 10% of the computer chips project’s
beta and 90% of the company’s initial βA . First, we need to find the initial βA :
D
! !
1
βAold = E
βD + D
D βE
+1 E E
+1
0.5 1
   
= 0.0 + 2.0 = 1.33.
1.5 1.5
Hence, the beta of the new assets (old plus new project) is:

βAnew = 0.1 × 1.123 + 0.9 × 1.33 = 1.31.

What is the beta of its equity now? Given that the debt to equity ratio is unchanged,
use the condition:
D
! !
1
βAnew = βD + D D
E
βE
+1 E E
+1
0.5 1
   
⇒ 1.31 = 0.0 + βE
1.5 1.5
⇒ βE = 1.97.

6. (a) The monthly interest rate is rm =12%/12=1% and the depreciation rate is d = 1%.
The present value of rent payments is:
!2 !3
1−d 1−d 1−d
PV(no renovation) = 1000 + 1000 + 1000 + 1000 + ...
1 + rm 1 + rm 1 + rm
 !2 !3 
1−d 1−d 1−d
= 1000 1 + + + + . . .
1 + rm 1 + rm 1 + rm
1 1 + rm
= 1000 1−d = 1000 = $50, 500
1− 1+rm
rm + d

4
(b) If George renovates immediately, the first time he gets rent is a month from now.
Thus, the present value of the stream of rent payments is:
1000 1000 1000
PV(renovation) = + 2
+ + ...
1 + rm (1 + rm ) (1 + rm )3
" #
1000 1 1 1
= 1+ + + + ...
1 + rm 1 + rm (1 + rm )2 (1 + rm )3
1000 1 1000
= × 1 = = $100, 000
1 + rm 1 − 1+rm r

The NPV of immediate renovation is $100,000-$50,000=$50,000. It is positive, so the

project would be worth taking if it were the only investment George has access to.
Here, however, he has another possibility: not renovating the apartment. This second
choice yields the NPV of $50,500 (since the cost of not renovating is 0), so it is more
attractive. Thus, the right choice is not renovating the apartment.
(c) You can easily check that the NPV of waiting one month is $50,504.95. Can George
do even better than that? The table below summarizes the cash flows he gets if he
rents for 3 months before he renovates:

Renovate after 3 months

Month 0 1 2 3 4 ...
Cashflows 1000 1000(1 − d) 1000(1 − d)2 −50000 1000 . . .

and the NPV is $51,470.49. Thus, delaying renovation for 3 months is even better.
In the renovation month the NPV of all remaining cash flows is $50,000, as we computed
in part (b). That is, if George renovates after one month, the overall NPV is 1000 +
50000/(1 + rm ). If he renovates in two months, it is 1000 + 1000(1 − d)/(1 + rm ) +
50000/(1 + rm )2 , etc. The general formula for NPV of renovating after t months is:
!2 !t−1
1−d 1−d 1−d 50000
1000 + 1000 + 1000 + . . . + 1000 +
1 + rm 1 + rm 1 + rm (1 + rm )t
 !2 !t−1 
1−d 1−d 1−d 50000
= 1000 1 + + + ... + +
1 + rm 1 + rm 1 + rm (1 + rm )t
 t
1−d
1− 1+rm 50000
= 1000 1−d + .
1 − 1+r m
(1 + rm )t

(d) Waiting involves a tradeoff: the expenditure is postponed, but the rent payments
decrease in the meantime. To choose optimally, George should use the formula from
part (c) to compute the NPV for different delay time, t. Then he should select the
month that yields the highest payoff.

5
Suppose we are at the end of month t, whether to renovate immediately or to wait
for another month depends on the benefit and cost of delaying the renovation for one
more month. The benefit of delaying includes the time value of the renovation cost,
50000 − 50000/(1 + rm ), plus the rent for month t + 1 at the old rate, 1000(1 − d)t .
And the cost of delaying is the rent for month t + 2 at the new rate discounted back
to the end of month t, that is, 1000/(1 + rm ). We can see that, for any t, the cost of
delaying the renovation for another month is the same, while the benefit of delaying
is decreasing in t. The profit maximizing strategy is to find the smallest t such that
the benefit of delaying the renovation is smaller than the cost of delaying for another
month.
50000 1000
50000 − + 1000(1 − d)t < ⇒ t > 69.96.
1 + rm 1 + rm
Since the renovation only takes place at the end of the month, we will rent for 70
months before the renovation. If we choose to renovate at the end of the 70th month,
the NPV will be $62,963.18.
The graph below depicts the NPV as a function of the delay. You could get a similar
graph from Excel or your favorite program/ programming language. The optimal
choice corresponds with the maximum value of NPV – George should rent for 70
months before he renovates the apartment.
65

62.5

60
NPV

57.5

55

52.5

50
0 50 100 150 200 250
Delay (Months)

(e) The value of the option is the difference between the NPV of the delayed project
(computed in part (d)) and the NPV the project would have without the option to
delay (computed in part (a)): $62,963.18−$50,500=$12,463.18.

6
Notice that the value of the option is considerably larger than the difference in NPVs
of renovating immediately or never renovating, $500 (as computed in (b)). Lesson: you
should be very careful about when to exercise real options, as the potential gains (or
losses!) could be a sizable portion of the value of the project.
(f) George not only lost the option to wait (worth $12,463.18) but he also chose the
wrong project. Recall from part (b) that without the option to delay, the NPV of reno-
vating is $500 lower than the NPV of never renovating. Thus, George made $12,963.18
less than he would have if he had taken MGT337.
Of course, George was also swindled by the black market contractors he hired, so he
actually wasted even more money. But we will save that story for another assignment.