Practice Final Exam III

FM423
Asset Markets

Part B
Question 2. (25 marks)

(a) (8 marks) Consider a perpetuity with annual payments starting in year 5. The first
payment in year 5 is £15; the payment then grows at a constant rate of 1.5% per year
forever. Consider a forward contract that delivers this perpetuity in year 3 (that is,
the forward price is paid in year 3).

(i) (4 marks) What is the forward price at date 0 if the term structure is flat at
r=1.8%?
Solution: Forward price is set such that the value of the forward contract is zero.
1 15
So, it should be F = 1+r r−g
=4911.59.
(ii) (4 marks) What is the value of a long position in this forward contract one year
later, if in year 1 the term structure shifts to 1.6% and remains flat?
1 1 15
Solution: The value of the long forward one year later is 1.016 2 ( 1.016 0.016−0.015 −

F )=9544.33.

(b) (8 marks) Consider the following two-factor model for the returns of three well-diversified
assets (i.e., with no idiosyncratic risk):

rA = 0.4 + I1 + 2I2
rB = 0.5 + 2I1 + I2
rC = 0.1 − I1 − I2

(i) (4 marks) Assume that the APT holds, compute the expected return of factor
replicating portfolio 1 (with respect to I1 ), the expected return of factor replicating
portfolio 2 (with respect to I2 ), and the risk-free rate. Assume that we have a
new asset, asset D (also with no idiosyncratic risk), with exposures 1 to factor 1
and -1 to factor 2. What is its expected return?
Solution: The risk-free asset has zero exposure to both factors, hence the weights
wA ,wB ,wC of assets A, B, C that replicate it should satisfy the following system

c LSE FM423 Page 1 of 10
 of equations:
wA + 2wB − wC = 0
2wA + wB − wC = 0
wA + wB + wC = 1
Solving this yields, wA = wB = 0.2, wC = 0.6, so that the risk-free rate is
rf = 0.2 ∗ 0.4 + 0.2 ∗ 0.5 + 0.6 ∗ 0.1 = 0.24. Similarly, for FRP1 the LHS is
the same but the RHS is (1,0,1), so that the weights are wA = −0.2, wB = 0.8,
wC = 0.4 and the expected return is r1 = 0.36. Finally for FRP2 the LHS is
the same but the RHS is (0,1,1), so that the weights are wA = 0.8, wB = −0.2,
wC = 0.4 and the expected return is r2 = 0.26.
The expected return of D is
rD = rf + βD,1 ∗ (r1 − rf ) + βD,2 ∗ (r2 − rf )
= 0.24 + 1 ∗ (36 − 0.24) − 1 ∗ (0.26 − 0.24)
= 0.34

(ii) (2 marks) Without knowing the return generating processes of the three assets
described above, how would you empirically estimate the risk premia of the two
factors.
Solution: Similar to the test for the CAPM: first-pass regression to estimate factor
loadings on each factor; then run cross sectional regressions of portfolio returns
on the factor loadings to estimate the factor risk premium.
(iii) (2 marks) Suppose that the two factors above are drawn from a large pool of
potential factors—for example, by selecting the ones with statistically significant
risk premia. What is a potential issue of this approach? What would you do to
mitigate the issue?
Solution: The validity of the factors are only within the time periods and type of
assets of the sample; to solve for this implement an out of sample test. It’s also
difficult to interpret the coefficients because there might be no economic principle
supporting the selected factor because it’s selected based on a statistical criteria.
To solve this issue, one may pre-select some factors based on economic reasoning
and then run the regressions.

(c) (9 marks)
(i) (3 marks) When a company announces surprisingly good earnings, its stock price
jumps up immediately, but then keeps rising in the weeks thereafter. A recent
study has found that this effect (i.e., the subsequent upward drift) is particularly
strong for firms that announce earnings on the same day as many other firms.
In other words, the greater the number of firms announcing earnings on a par-
ticular day, the stronger the subsequent drift for those firms after their earnings
announcements. What is a plausible interpretation of this evidence?

c LSE FM423 Page 2 of 10
 Solution: Investors have limited attention and/or processing capacity; so they do
not adjust their expectations sufficiently.
(ii) (3 marks) A recent study shows that most of the long-run return reversal is con-
centrated around subsequent earnings announcement dates. For example, a firm
with high past 5-year returns as of April 2013 tends to have low stock returns
around the subsequent earnings announcement (e.g., July 2013). What is a plau-
sible interpretation of this evidence?
Solution: Investors extrapolate past earnings growth to the future, and are dis-
appointed when subsequent earnings are announced.
(iii) (3 marks) Two securities with identical future cash flows may at times be traded
at different prices (e.g., the twin stock puzzle). List two possible explanations for
this phenomenon.
Solution: Implementation costs; noise trade risk.

c LSE FM423 Page 3 of 10
Question 3. (25 marks)
(a) (10 marks) Consider a binomial model with the up move u = 1.2, down move d = 0.9,
risk free rate R = 1+10% = 1.1, and initial stock price S = 100.

(i) (3 marks) Calculate the price of a European call option with exercise price K =
100 and n = 3 periods left until expiry.
Solution: The risk neutral probability of the “up” state is
R−d 2
q= = .
u−d 3
3 +
3


q k (1 − q)3−k uk d3−k S − K
P
The price of European Option is: C = k
=
k=0
3 +
3! 2k
P  k 3−k
(3−k)!k! 33 100 1.2 0.9 − 1 = 26.09.
k=0
(ii) (3 marks) Calculate the price of an American put option with the same exercise
price and time to expiry as in (a)(i).
Solution: For the American put, one has to work backwards through the binomial
tree, determining at each node whether it is optimal to exercise or not.
u3 S = 172.8
E=−, N E=0
%
u2 S = 144
E=−, N E=0
% &
uS = 120 u2 dS = 129.6
E=−, N E=0.26 E=−, N E=0
% & %
S = 100 udS = 108
E=0, N E=3.19 E=−, N E=0.85
& % &
dS = 90 ud2 S = 97.2
E=10, N E=6.27 E=2.8, N E=0
& %
d2 S = 81
E=19, N E=9.91
&
d3 S = 72.9
E=27.1, N E=0
At each node of the binomial tree, three numbers are listed:
the current stock price St , which is of the form uk d3−k S
the value of the option if exercised, denoted by E, which is equal to 100-St (a “-”
symbol means that the exercise value is negative), the value of the option if it is
not exercised at that node, denoted by NE, which is equal to the discounted risk-
neutral expectation of the next period value of the option. The optimal exercise
policy is indicated in bold font. The date 0 value of the option is 3.19.
(iii) (4 marks) Calculate the price of an option which, at the end of the third period,
gives its holder the right to purchase the underlying stock at the minimum price
realized over the life of the option.
Solution: Let the exercise value of this option at the terminal date be CT . Then
CT = [ST −K]+ , where ST is the price of the underlying asset at the terminal date,

c LSE FM423 Page 4 of 10
 and K is the minimum price of the underlying asset over the life of the option.
Since this is a path-dependent option, we need to look at the non-recombining
binomial tree as shown in the following figure.
u3 S = 172.8
K=100
%
u2 S = 144 −→ u2 dS = 129.6
K=100
%
uS = 120 −→ udS = 108 −→ u2 dS = 129.6
K=100
% & ud2 S = 97.2
K=97.2
S = 100
& % u2 dS = 129.6
K=90
dS = 90 −→ udS = 108 −→ ud2 S = 97.2
K=90
&
d2 S = 81 −→ ud2 S = 97.2
K=81
&
d3 S = 72.9
K=72.9
At each node the current price of the underlying asset is shown (of the form
uk d3−k S). At the terminal nodes the value of K is shown as well. The date 0
price of the option, C, is given by the risk-neutral expectation of CT , discounted
over three periods.
72.8 ∗ 2 + 29.6
Cuu = = 53.09
1.1 ∗ 3
29.6 ∗ 2 + 0
Cud = = 17.94
1.1 ∗ 3
39.6 ∗ 2 + 7.2
Cdu = = 26.18
1.1 ∗ 3
16.2 ∗ 2 + 0
Cdd = = 9.82
1.1 ∗ 3
Cuu ∗ 2 + Cud
Cu = = 37.61
1.1 ∗ 3
Cdu ∗ 2 + Cdd
Cd = = 18.84
1.1 ∗ 3
Cu ∗ 2 + Cd
C = = 28.49
1.1 ∗ 3

(b) (5 marks) Consider two European-type

√ options that have payoffs P1 (ST ) = max(1 −
ST , 0) and P2 (ST ) = max(1 − ST , 0) , where stock price ST can take any value
greater than zero. Both options have the same maturity date and are written on the
same stock. Which of these options costs more at date 0? Explain your answer.
√
Solution: These options are in the money only when ST ≤ 1. In this range, ST ≥ ST .
Consequently, the first option has strictly larger payoff at the final date, and hence is
more expensive.

c LSE FM423 Page 5 of 10
 (c) (10 marks) Now consider a stock with Geometric Brownian Motion dynamics as in the
Black & Scholes model:

dSt
= µdt + σdWt ,
St
µ and σ are positive known constants. There is also a money-market account with
dynamics

dBt
= rdt,
Bt
where r is a positive known constant.

(i) (5 marks) What is the ∆ (i.e., the sensitivity of the price with respect to the
underlying asset price) and Γ (i.e., the sensitivity of the ∆ with respect to the
underlying asset price) of a straddle portfolio (i.e., buying a European call and a
European put together with same strike price K and maturity T ), as a function
of the delta and gamma of the corresponding call option, ∆C and ΓC ? Explain
under which condition the straddle is delta-neutral.
(ii) (2 marks) Explain why investors prefer low gamma or even gamma-neutral port-
folios.
(iii) (3 marks) What is the sign of ν (i.e., the sensitivity of the price with respect to
σ) of this straddle and what is ν right before the straddle expires?

Solution: Given call-put parity, ∆C = ∆P + 1 and ΓC = ΓP . Then the straddle’s

delta is 2∆C − 1 and gamma is 2ΓC . To have delta-neutral straddle, we must have
∆ = 0 = 2∆C − 1, i.e. ∆C = 0.5, which in general happens when both put and call
options are ATM. Since in practice it is not feasible to be continuously delta-hedged,
a reinforcing strategy is to select positions that, in addition to being close to delta-
neutral, have deltas that are relatively insensitive to movements in the stock price, i.e.
low gammas. A gamma-neutral position has a gamma of zero and should use some
other asset except underlying stock.
Since both call and put option have same positive ν, the ν for straddle is also positive.
Considering ν is the sensitivity of the price with respect to the stock volatility, when
the maturity is coming, ν should go to zero. The price is no longer sensitive to stock
volatility.

c LSE FM423 Page 6 of 10
Question 4. (25 points)

(a) (5 points) A UK-based trader is given the following quotes (where GBP is the home
currency):

Spot exchange rate St [GBP/U SD] = 0.7000

Spot exchange rate St [GBP/SEK] = 0.0800
6 month forward exchange rate Ft,t+6 [GBP/U SD] = 0.6880
6 month forward exchange rate Ft,t+6 [GBP/SEK] = 0.0840
6 month GBP interest rate per annum 1%
6 month USD interest rate per annum 4.50%
6 month SEK interest rate per annum 1.80%
(GBP=British Pound, USD=US Dollar, SEK=Swedish Krona)

Is there an arbitrage opportunity based on these quotes? If yes, describe the strategy
(i.e., describe both the long and short sides of the arbitrage trade). Further assume
that the trader has £1 million to work with (i.e., the absolute value of the long or short
position at time 0 cannot exceed £1 million), what is the maximum profit realized in
month 6?
Solution: (1 points)Using CIP for GBP/USD:

1 + 0.01
2
Ft,t+6 [GBP/U SD] = 0.7 ∗ = 0.6880
1 + 0.045
2

The forward quoted in the question for GBP/USD is correct and arbitrage is not
possible.
(1 points) Using CIP for GBP/SEK:

1 + 0.01
2
Ft,t+6 [GBP/SEK] = 0.08 ∗ = 0.0797
1 + 0.018
2

Since the forward calculated from the CIP is different from the one quoted in the
question arbitrage is possible.
(1 points) At t = 0, borrow £1m for 6 months at the GBP interest rate. Repay with
interest at t = 6 months, which generates a cash flow of −£1m∗(1.005) = −£1, 005, 000
at t = 6 months.
(1 points) Also at t = 0, convert £1m into SEK (12.5m SEK) and invest for 6 months
at SEK interest rate. At t = 0, enter into forward to sell 12.5m ∗ 0.9% = 12, 612, 500
SEK in 6 months for GBP.
(1 points) This generates £1m ∗ (1/0.08) ∗ (1.009) ∗ 0.084 = £1, 059, 450 at t = 6
months. In all the profit is £1, 059, 450 − £1, 005, 000 = £54, 450 and the cost is
£1m − £1m = 0, which is an arbitrage.

c LSE FM423 Page 7 of 10
 (b) (3 points) Do you expect the covered interest parity (CIP) to hold in the market?
Justify your answer and state any necessary conditions.
Solution: The CIP should hold in the market. All the quotes/rates are assumed to be
risk free and are known today. Thus if the CIP does not hold then arbitrage is possible
as in the example above. The CIP as a no arbitrage pricing relation is also conditional
on capital being free to flow across markets.

(c) (5 points) The return of the US stock market index, in US dollar (USD) terms, has a
variance of 0.09. The return of the UK market index, in British pound (GBP) terms,
has a variance of 0.04. The correlation in returns between the US and UK markets,
both expressed in GBP terms, is 0.1. Finally, the variance of the changes in the ex-
change rate of GBP/USD (i.e., the variance of the currency return) is 0.03.

A UK-based asset manager invests 75% of his capital in the UK stock market index
and 25% in the US stock market index. What is the approximate standard deviation
of the portfolio return expressed in GBP terms? State any assumptions you make in
your calculation.
Solution: (4 points) The variance of the return on the portfolio where all terms are
expressed in GBP is calculated as follows:

σr2p in £
= wU2 K σr2U K + wU2 S σr2U S in £
+ 2wU K σrU K wU S σrU S in £
ρ(rU K , rU S )in £

σr2p in £
2
= (0.75) (0.2) + 2
(0.25)2 σr2U S in £
+ 2(0.75)(0.25)(0.2)σrU S in £
(0.1)

(6 points)

σr2U S in £
≈ σr2U S in $
+ σs2GBP/U SD
σr2U S in £
≈ 0.09 + 0.03 = 0.12

In our approximation above we assume the interaction terms (1 variance and 6 covari-
ance terms) are negligible.
Therefore:
1
σr2p in £
= (0.75)2 (0.2)2 + (0.25)2 (0.12) + 2(0.75)(0.25)(0.2)(0.12) 2 (0.1)
σr2p in £
= 0.1805

(d) (6 points) A UK-based investor has £1 million of capital. The investor wants to
construct a portfolio that invests only in stocks A and B and that is uncorrelated with
the market. The table below shows the regression results of stocks A and B’s excess
returns on contemporaneous excess market returns (i.e., regression results based on
the CAPM).

c LSE FM423 Page 8 of 10
 Stock A Stock B
Total Variance 0.1125 0.05
Intercept and Beta 0.01 + 1.5(rm − rf ) 0.03 + 0.5(rm − rf )
Standard deviation of residual 0.15 0.2

The average excess return of the market is 10% and the risk free rate is 2%. Assume
that the idiosyncratic risk of stocks A and B is uncorrelated. What weights should the
investor have in stocks A and B? What is the Sharpe ratio of the portfolio?
Solution: (2 points)For the portfolio to be uncorrelated with the market we set the
beta of the portfolio to 0.

βp = wA βA + (1 − wA )βB
−βB
wA =
βA − βB
−0.5
wA = = −0.5
1.5 − 0.5
wB = 1.5

The investor should be long 150% in Stock B and short 50% in Stock A.
(1 points)The Sharpe ratio of the portfolio is calculated from:

E(rp ) − rf
Sp =
σrp
E(rp ) − rf = wA [E(rA ) − rf ] + wB [E(rB ) − rf ]
E(rp ) − rf = −0.5[0.01 + 1.5(0.1)] + 1.5[0.03 + 0.5(0.1)] = 0.04

(3 points)Since the beta of the portfolio is 0 and the idiosyncratic risks of Stocks A
and B are independent:

σr2p = βp2 σr2m + σe2p

= σe2p
2
X
= wi2 σe2i
i=1
= (−0.5)2 (0.15)2 + (1.5)2 (0.2)2 = 0.095625
σrp = 0.30923

Therefore the Sharpe ratio of the portfolio is:

0.04
Sp = = 0.1294
0.30923

c LSE FM423 Page 9 of 10
 (e) (3 points) In event studies, we often have to calculate abnormal returns. How are
abnormal returns defined? What assumptions do we have to make?
Solution: We have to take a stand on the asset pricing model. For example, under
CAPM
ARit = (rit − rf t ) − (α̂i + β̂i (rmt − rf t ))

(f) (3 points) Suppose, in an event study of how analyst downgrades might impact firm
value, we observe the following pattern of average cumulative abnormal returns (day
0 is the announcement day of the downgrade). What can we infer from this pattern?
Is this pattern consistent with market efficiency?

Solution: Investors overreact to the downgrade on the announcement day, there’s a

subsequent reversal.

c LSE FM423 Page 10 of 10