A1.11 B1.16

Stochastic Financial Models
Part II, 2003

(i) In the context of a single-period financial market with dd traded assets, what is an arbitrage? What is an equivalent martingale measure?

A simple single-period financial market contains two assets, S0S^{0} (a bond), and S1S^{1} (a share). The period can be good, bad, or indifferent, with probabilities 1/31 / 3 each. At the beginning of the period, time 0 , both assets are worth 1 , i.e.

S00=1=S01S_{0}^{0}=1=S_{0}^{1}

and at the end of the period, time 1 , the share is worth

S11={a if the period was bad, b if the period was indifferent c if the period was good S_{1}^{1}= \begin{cases}a & \text { if the period was bad, } \\ b & \text { if the period was indifferent } \\ c & \text { if the period was good }\end{cases}

where a<b<ca<b<c. The bond is always worth 1 at the end of the period. Show that there is no arbitrage in this market if and only if a<1<ca<1<c.

(ii) An agent with C2C^{2} strictly increasing strictly concave utility UU has wealth w0w_{0} at time 0 , and wishes to invest his wealth in shares and bonds so as to maximise his expected utility of wealth at time 1 . Explain how the solution to his optimisation problem generates an equivalent martingale measure.

Assume now that a=3/4,b=1a=3 / 4, b=1, and c=3/2c=3 / 2. Characterise all equivalent martingale measures for this problem. Characterise all equivalent martingale measures which arise as solutions of an agent's optimisation problem.

Calculate the largest and smallest possible prices for a European call option with strike 1 and expiry 1, as the pricing measure ranges over all equivalent martingale measures. Calculate the corresponding bounds when the pricing measure is restricted to the set arising from expected-utility-maximising agents' optimisation problems.