Solution 4

Problem 1
Assume that there are two risky assets listed in the market. Expected returns and risks for the
risky assets are

𝐸[𝑅] 𝜎
7.5% 22%
10%% 27%
a) Determine the analytical formulation of the expected return of a portfolio as a function of
the standard deviation of the portfolio in case the two risky assets are perfectly positively
correlated.
b) Determine the analytical formulation of the expected return of a portfolio as a function of
the standard deviation of the portfolio in case the two risky assets are perfectly negatively
correlated.

Solution
a) The feasible set of portfolios can be represented as a curve in the mean variance space. We
can write

µ𝑃 = 𝑤µ1 + (1 − 𝑤)µ2

𝜎𝑃2 = 𝑤 2 𝜎12 + (1 − 𝑤)2 𝜎22 + 2𝑤(1 − 𝑤)𝜌12 𝜎1 𝜎2

We know that for 𝜌12 = 1

𝜎𝑃2 = 𝑤 2 𝜎12 + (1 − 𝑤)2 𝜎22 + 2𝑤(1 − 𝑤)𝜎1 𝜎2

𝜎𝑃 = 𝑤𝜎1 + (1 − 𝑤)𝜎2

and the standard deviation of the portfolio is a linear combination of the standard deviations
of the two securities. In order to get the equation of the frontier we can solve for w
𝜎𝑃 = 𝑤(𝜎1 − 𝜎2 ) + 𝜎2
𝜎𝑃 − 𝜎2
𝑤=
𝜎1 − 𝜎2

and plugging this value into the definition of the expected return
𝜎𝑃 − 𝜎2 𝜎𝑃 − 𝜎2
µ𝑃 = µ1 + (1 − )µ
𝜎1 − 𝜎2 𝜎1 − 𝜎2 2
𝜎𝑃 − 𝜎2 𝜎1 − 𝜎𝑃
µ𝑃 = µ1 + µ
𝜎1 − 𝜎2 𝜎1 − 𝜎2 2
 (𝜎𝑃 − 𝜎2 )µ1 + (𝜎1 − 𝜎𝑃 )µ2
µ𝑃 =
𝜎1 − 𝜎2
µ1 − µ2 𝜎1 µ2 − 𝜎2 µ1
µ𝑃 = 𝜎𝑃 +
𝜎1 − 𝜎2 𝜎1 − 𝜎2

0.075 − 0.10 0.22 ∙ 0.10 − 0.27 ∙ 0.075

µ𝑃 = 𝜎𝑃 +
0.22 − 0.27 0.22 − 0.27
µ𝑃 = 0.5𝜎𝑃 − 0.035

b) For 𝜌12 = −1
𝜎𝑃2 = 𝑤 2 𝜎12 + (1 − 𝑤)2 𝜎22 − 2𝑤(1 − 𝑤)𝜎1 𝜎2

𝜎𝑃 = 𝑤𝜎1 − (1 − 𝑤)𝜎2 = −𝑤𝜎1 + (1 − 𝑤)𝜎2

and we will consider the positive solution so

𝑤𝜎1 − (1 − 𝑤)𝜎2 ≥ 0

𝑤𝜎1 − 𝜎2 + 𝑤𝜎2 ≥ 0
𝜎2
𝑤≥
𝜎1 + 𝜎2

so we have that
𝜎2
𝑤𝜎1 − (1 − 𝑤)𝜎2 𝑖𝑓 𝑤 ≥
𝜎1 + 𝜎2
𝜎𝑃 = { 𝜎2
−𝑤𝜎1 + (1 − 𝑤)𝜎2 𝑖𝑓 𝑤 <
𝜎1 + 𝜎2

In order to get the frontier equation we can again solve for w

𝜎𝑃 + 𝜎2 𝜎2
𝑖𝑓 𝑤 ≥
𝜎 +𝜎 𝜎1 + 𝜎2
𝑤 = { 𝜎1 − 𝜎2 𝜎2
2 𝑃
𝑖𝑓 𝑤 <
𝜎1 + 𝜎2 𝜎1 + 𝜎2

and plugging this value into the definition of the expected return, we have
𝜎𝑃 + 𝜎2 𝜎2
(µ1 − µ2 ) + µ2 𝑖𝑓 𝑤 ≥
𝜎 +𝜎 𝜎1 + 𝜎2
µ𝑃 = {𝜎1 − 𝜎2 𝜎2
2 𝑃
(µ1 − µ2 ) + µ2 𝑖𝑓 𝑤 <
𝜎1 + 𝜎2 𝜎1 + 𝜎2
 µ1 − µ2 µ1 𝜎2 + µ2 𝜎1 𝜎2
𝜎𝑃 + 𝑖𝑓 𝑤 ≥
𝜎1 + 𝜎2 𝜎1 + 𝜎2 𝜎1 + 𝜎2
µ𝑃 = µ1 − µ2 µ1 𝜎2 + µ2 𝜎1 𝜎2
− 𝜎𝑃 + 𝑖𝑓 𝑤 <
{ 𝜎1 + 𝜎2 𝜎1 + 𝜎2 𝜎1 + 𝜎2

0.075 − 0.10 0.075 ∙ 0.27 + 0.10 ∙ 0.22 0.27

𝜎𝑃 + 𝑖𝑓 𝑤 ≥
µ𝑃 = { 0.22 + 0.27 0.22 + 0.27 0.22 + 0.27
0.075 − 0.10 0.075 ∙ 0.27 + 0.10 ∙ 0.22 0.27
− 𝜎𝑃 + 𝑖𝑓 𝑤 <
0.22 + 0.27 0.22 + 0.27 0.22 + 0.27
−0.0510 𝜎𝑃 + 0.0862 𝑖𝑓 𝑤 ≥ 0.551
µ𝑃 = {
+0.0510 𝜎𝑃 + 0.0862 𝑖𝑓 𝑤 < 0.551
Problem 2
In a financial market there are two risky assets A and B and a risk-free security. The returns
over the next year of the risky assets are uncorrelated and the expected returns and variances
are µA = 0.03, µB = 0.12, σA = 0.09 and σB = 0.18. A portfolio manager is holding the
following portfolio:
risk-free Asset A Asset B
Manager 1500 $ 4313 $ 3187 $
a) Find out the composition, the expected return and risk of the tangency portfolio.
b) The manager allocation optimizes a mean-variance utility function, E(rp )− 21 γσp2 with γ = 7,
find the return r0 of the risk-free asset.
c) Find the composition of the minimum variance portfolio with an expected return equal to
4%. Is this portfolio efficient?

Solution
a) By the two-fund separation theorem, the manager is holding a portfolio made of the tangency
portfolio and the risk-free asset. The composition of the tangency portfolio can be found
looking at the the risky holdings of the manager.

4313
wA = = 0.575
4313 + 3187
wB = 1 − wA = 0.425

We can calculate the expected return and the variance of the tangency portfolio:
µT = w A µA + w B µB
= 0.575 · 0.03 + 0.425 · 0.12 = 6.825%
and
σT2 = wA2 2
σA + wB2 2
σB
= 0.575 · 0.092 + 0.4252 · 0.182 = 0.00853
2
√
σT = 0.00853 = 9.24%

b) We know that investor is optimizing her portfolio, so we have:

 
1 2
max E(rp ) − γσp
w 2
 
1 2 2
max r0 + w(µT − r0 ) − γw σT
w 2
Thus we know that the FOC must be valid wih a fractional allocation to the tangency
portfolio given by w? = 75/90 hence
µT − r0 − γwσT2 = 0
r0 = µT − γwσT2
75
= 0.06825 − 7 · · 0.0085
90
= 1.85%
c) Every possible combination of the two assets is a minimum variance portfolio since we have
only two risky assets. Thus, we can find the portfolio with expected return µp = 4% by
solving:

µp = wA µA + (1 − wA )µB
0.04 = wA · 0.03 + (1 − wA ) · 0.12
0.08
wA = = 0.8889
0.09
To see if this portfolio is efficient, we can look at the return of the absolute minimum variance
portfolio. We can find the composition of this portfolio by minimizing the portfolio variance:
 
2 2 2 2
min wA σA + (1 − wA ) σB
wA

and by taking the FOC, we have:

2wA σA2 − 2(1 − wA )σB2 = 0

wA (σA2 + σB2 ) − σB2 = 0
σB2 0.182
wA = = = 0.8
σB2 + σA2 0.182 + 0.092
wB = (1 − wA ) = 0.2

The expected return of this portfolio is:

µmvp = 0.8µA + 0.2µB

= 0.8 · 0.03 + 0.2 · 0.12 = 0.048

and since µp < µmvp we conclude that the minimum variance portfolio with expected return
equal to 4% is not efficient.

Problem 3
Assume that only two risky assets are traded in the market and that they have the following
expected returns, risks and correlation

A B
µ 0.04 0.10
σ 0.10 0.15
ρ 0.3

a) Calculate the composition of the absolute minimum variance portfolio and plot the mean
variance efficient frontier. Which condition (in terms of ρAB , σA and σB ) must be satisfied
in order for this portfolio not to be fully invested in security A?

Solution
a) In order to find the composition of the absolute minimum variance portfolio, we have to
solve the following problem:
  
2
min wmv σA2 + (1 − wmv )2 σB2 + 2wmv (1 − wmv )ρAB σA σB
wmv

∂σP2
= 2wmv σA2 − 2(1 − wmv )σB2 + 2ρAB σA σB − 4wmv ρAB σA σB = 0
∂wmv
σ 2 − ρAB σA σB 0.0225 − 0.0045
wmv = 2 B 2 = = 0.76596
σA + σB − 2ρAB σA σB 0.01 + 0.0225 − 0.009

Thus, we have that the minimum variance portfolio has weights wA = 0.76596, wB =
0.23404, expected return µmv = 0.05404 and standard deviation σmv = 0.09334.

The minimum variance portfolio is not fully invested in risky asset A when

σB2 − σAB
wmv = <1
σA2 + σB2 − 2σAB
σB2 σAB < σA2 + σB2 − 2σAB
σA > ρAB σB
σA
> ρAB
σB

This condition is satisfied in the current setting as:

0.10
= 0.666 > 0.3.
0.15

Problem 4
Two risky assets a and b are listed in the market. A rational investor with a mean variance
utility function E[U ] = µp − 21 γσp2 with γ = 2 has the following expectations:

a b
µ 0.05 0.10
σ 0.10 0.20

a) If we know that this investor is holding a portfolio with the following composition wa =
0.28261 and wb = 0.71739, can we infer the expectation of this investor about the correlation
level ρab between two assets?

b) A risk-free asset with a return r0 = 2.5% becomes available in the economy and we observe
a second rational mean variance investor, more risk averse (γ2 = 5) than the first one, is
holding the following portfolio wa = 0.429299, wb = 0.353540, and w0 = 0.217161. Assuming
that the two investors share the same expectations about the future behavior of the risky
assets, can we predict how the first investor will modify her portfolio in order to incorporate
the newly-available risk-free asset?
Solution
a) The investor can find the optimal portfolio composition by maximizing her expected utility:
 
1 2
max E(rp ) − γσp
wa 2
 
1  2 2 2 2 2

max (wa µa + (1 − wa )µb ) − γ wa σa + (1 + wa − 2wa )σb + (2wa − 2wa )σa σb ρab
wa 2

The FOC is:

µa − µb − γ wa σa2 + wa σb2 − σb2 + σa σb ρab − 2wa σa σb ρab = 0

 

µa − µb
wa σa2 + wa σb2 − σb2 + σa σb ρab − 2wa σa σb ρab =
γ
µa − µb + γ(σb2 − σa σb ρab )
wa =
γ(σa2 + σb2 − 2σa σb ρab )

We can also solve the FOC for ρab and knowing the composition of the optimal portfolio wa ,
we have that:
µa − µb − γ(wa σa2 + wa σb2 − σb2 )
ρab =
γ(σa σb − 2wa σa σb )
0.05 − 0.1 − 2 · (0.28261 · 0.01 + 0.28261 · 0.04 − 0.04)
ρab = = 0.1
2 · (0.1 · 0.2 − 2 · 0.28261 · 0.1 · 0.2)

b) If the two investors share the same expectations, they also share the same tangency portfolio,
which has the following composition:
0.429299
wa,t = = 0.548387
0.429299 + 0.353540
0.353540
wb,t = = 0.451613
0.429299 + 0.353540

Thus, we can compute the expected return and standard deviation of the tangency portfolio:

µt = wa,t µa + wb,t µb = 0.548387 · 0.05 + 0.451613 · 0.10 = 7.258%

σt2 = wa,t
2
σa2 + wb,t
2
σb2 + 2wa,t wb,t σa σb ρab
= 0.5483872 · 0.01 + 0.4516132 · 0.04 + 2 · 0.548387 · 0.451613 · 0.1 · 0.2 · 0.1
= 0.012156

We know that the first investor will calculate the new optimal portfolio composition maxi-
mizing:
 
1 2
max E(rp ) − γσp
w 2
 
1 2 2
max (wµt + (1 − w)r0 ) − γw σt
w 2
The FOC for this problem is:

µt − r0 − γwσt2 = 0
µt − r0 0.072581 − 0.025
wt = 2
= = 1.9571
γσt 2 · 0.012156

Knowing the investment in the tangency portfolio, we can find the investment in the three
assets:

wa = wt wa,t
= 1.957 · 0.548387 = 1.0732
wb = wt wb,t
= 1.957 · 0.451613 = 0.88385
w0 = 1 − wt
= 1 − 1.9571 = −0.9571.