FINA 3103: Intermediate Investments

Problem Set 4 Solution

Prof. Don Noh

April 19, 2024

Problem 1
(a) Assume CAPM holds
E[Ri ] − R f = β i (E[Rm ] − R f ) (1)

In the case of Coca-Cola, we have

4.30% − 0.80% = 0.41(E[Rm ] − R f ), (2)

which implies the market risk premium is

4.30% − 0.80%
E[Rm ] − R f = = 8.54%. (3)
0.41

Apply CAPM (1)for Amazon and Walt Disney, we have

E[R A ] = 8.23%, E[RW ] = 12.32%. (4)

From the investment fraction, we have their corresponding weights

′
w = [0.5 0.2 0.3] (5)

Finally the expected return of the portfolio is

E[R p ] = 0.5 × 8.23% + 0.2 × 4.30% + 0.3 × 12.32% = 8.67%. (6)

(b) The Beta of the portfolio in (a) is

β p = 0.5 × 0.87 + 0.2 × 0.41 + 0.3 × 1.35 = 0.922. (7)

1
 To have the same expected return as the portfolio in (a), one only need a portfolio with
the same beta. Suppose we allocate $x to the market portfolio of all stocks with beta of 1,
then solve for x

x 10, 000 − x
0.922 = ×1+ ×0 ⇒ x = 9, 220 (8)
10, 000 10, 000

(c) Portfolio in (b) should contain more stocks than that in (a), and hence can benefit more
from diversification, reducing idiosyncratic risks.
Comments:
• Risk averse investors do not necessarily hold riskless asset. Think about the investor
who puts his entire wealth in the tangency portfolio. He is still risk averse, and
those who are more risk averse than him would starting investing in the riskless
asset.
• Higher weight on riskless asset does not necessarily imply lower risk if the com-
position of risky assets also changes. Consider a market portfolio X with all stocks
that has beta equal to 1, i.e., R X − Rb = Rm − Rb , and another market portfolio Y
with all stocks that has beta equal to 2, i.e., RY − Rb = 2(Rm − Rb ), assuming that
idiosyncratic risks are fully diversified away in both market portfolios. Now com-
pare a portfolio A with zero riskless asset and 100% portfolio X, and a portfolio B
with 50% riskless asset and 50% portfolio Y. Then we can write R A = R X = Rm and
R B = 0.5Rb + 0.5RY = 0.5Rb + 0.5 × [2(Rm − Rb ) + Rb ] = Rm . Thus these two portfolios
A and B have the same expected return E[Rm ], and the same variance Var(Rm ).

Problem 2
(a) Regression results are reported in Table I. See details in the Excel.

Table I
CAPM estimates
Standard errors are reported in parentheses. ∗∗∗ , ∗∗ , and ∗ indicate p < 0.01,
p < 0.05, and p < 0.1, respectively.

Germany Japan UK USA

−0.0003 −0.0028 −0.0006 0.0015*
α
(0.0019) (0.0022) (0.0012) (0.0008)
1.3047*** 0.8592*** 0.9446*** 0.9272***
β
(0.0450) (0.0535) (0.0287) (0.0193)

(b) The riskiest (i.e. highest β) country is Germany.

2
 (c) The safest (i.e. lowest β) country is Japan.
(d) No, we cannot reject the CAPM at the 5% significance level. The null hypothesis is α = 0.
To carry out the test, we compare the absolute values of t-stats of α to 1.96, or p-values to
0.05, or see if 0 falls in the 95% confidence intervals. All of the countries have absolute
values of t-stats smaller than 1.96, p-values greater than 5%, and 0 lying in the 95%
confidence intervals. Hence, all alpha’s are statistically insignificant, and we cannot reject
CAPM for any of these countries.

Problem 3
There are three constraints for the problem:
1. All weights on the given portfolios are non-negative.

wi ≥ 0 ∀i = 1, 2, . . . , 5 (9)

2. The weight of risk-free asset is no less than −1.

5
w f ∶= 1 − ∑ wi ≥ −1 (10)
i=1

3. The portfolio has a beta of β.

5
β p ∶= ∑ wi β i = β (11)
i=1

The objective is to maximize the total expected return

5 5
(1 − ∑ wi ) Rb + ∑ wi E[Ri ]. (12)
i=1 i=1

Express expected return in terms of alpha and beta, then we have E[Ri ] = Rb + αi + β i (E[Rm ] −
Rb ). Substitute this into above objective function and we can show the objective is

5
max Rb + β p (E[Rm ] − Rb ) + ∑(αi wi ). (13)
w
i=1

Note that only the last summation term in the objective function involves the control variables
wi , and that other terms are constants. Hence, the objective is equivalent to

5
max ∑(αi wi ) (14)
w
i=1

3
given constraints in Equation 9, Equation 10 and Equation 11.
(a) Put the problem into solver in excel or other programs and set β = 1, and we obtain the
following solution:

w1 w2 w3 w4 w5 wf
1.471 0 0 0 0 -0.471

See the Excel file for detailed calculations. There are many ways to set up the objective
and constraints in Excel solver.
(b) The only change now is to set set β = 2. We obtain the following solution:

w1 w2 w3 w4 w5 wf
0 1.661 0 0 0.339 -1