MathFinance 346/Stat391
Homework 1
Due Wednesday January 21, 2004
The first two problems concern foreign exchange. In problem 1, the same assumptions as were made in the Lecture
are assumed to be in force; in problem 2, the coefficients are assumed to be time-dependent.
Problem 1: Compute the arbitrage price at time t = 0 of the call option discussed in the notes for Lecture 9 by
calculating the discounted risk-neutral expectation under the risk-neutral measure QB for pound-sterling investors.
Verify that the price agrees with that computed using the risk-neutral measure QA for dollar investors.
Problem 2: Assume that the shart rates rtA and rtB for the US and UK Money Markets are time-dependent, but
nonrandom and continuous. Assume also that the exchange rate Yt between $ and £ (that is, Yt = number of Pounds
Sterling that one dollar will buy at time t) obeys the Stochastic Differential Equation
dYt = µt Yt dt + σt Yt dWt
where σt is a positive, nonrandom, continuous function of t. Here Wt is a standard Wiener process under QB , the
risk-neutral probability measure for £ investors.
What relation holds among the functions rtA , rtB , µt , and σt ?
The remaining problems concern an important SDE , the “mean-reversion” equation that governs the Ornstein-
Uhlenbeck process. For these you will need the Itô formula, the Girsanov theorem, and the exponential martingales
discussedc in the Lecture. Assume that under the probability measure P , the process Wt is a standard Wiener process.
Problem 3: Let f (t) be a continuous, nonrandom function of t. What is the distribution of the random variable
Z t
f (s) dWs ?
0
Rt
Hint: Evaluate the moment generating function E exp{λ 0
f (s) dWs }. Using what you saw in the Lecture, you sould
be able to do this with virtually no calculations.
Problem 4: The Ornstein-Uhlenbeck process Xt with spring constant α > 0 is the solution to the Stochastic Differ-
ential Equation
(1) dXt = −αXt dt + dWt .
The initial point X0 is arbitrary. (A) Find an explicit solution to this stochastic differential equation of the form
Z t
(2) Xt = X0 + g(t) f (s) dWs
0
where g(t) and f (t) are continuous, nonrandom functions of t. Hint: Itô. (B) What is the distribution of the random
variable Xt ? (C) What is the covariance of the random variables X(t), X(s)?
Problem 5: Suppose that under P the process Wt is a standard Wiener process, and that Q is the probability measure
on the class FT of events observable up to time T that is defined by
Z T Z T
dQ
(3) = exp{− Ws dWs − Ws2 ds/2}.
dP FT 0 0
(A) What is the nature of the process Wt under the measure Q?
(B) Show that the Radon-Nikodym derivative may be written as
Z T
dQ 1
(4) = exp{− (WT2 − T ) − Ws2 ds/2}.
dP FT 2 0
In this form, the Radon-Nikodym derivative is useful both for simulations and for theoretical investigations, the latter
using the Feynman-Kac formula, about which you will undoubtedly hear more later.
