Mertons Jump Diusion Model

Peter Carr (based on lecture notes by Robert Kohn) Bloomberg LP and Courant Institute, NYU

Continuous Time Finance

Lecture 5 Wednesday, February 16th, 2005

Introduction

Mertons 1976 JFE article Option pricing when underlying stock returns are discontinuous was the rst to explore jump diusion models. Jump diusion models address the issue of fat tails. Recent reference: A. Lipton, Assets with jumps, RISK, Sept. 2002, 149-153. When the underlying can jump to any level, the market is not complete, since there are many more states than assets. How to come up with a unique price for options in this setting? Mertons novel proposal: Assume that the extra randomness due to jumps can be diversied away.

Mathematical Impact of Jumps

Black-Scholes PDE becomes a partial integrodierential equation. Fourier transform is a convenient tool for solving PIDE (especially in a constant coecient setting). Contrast a one-dimensional diusion for returns y = ln S : dy = dt + dw, where and can be functions of y and t, with the SDE in a jump-diusion setting: dy = dt + dw + JdN Here, the jump magnitudes J are i.i.d. r.v.s, i.e. the jump-size J is selected by drawing from a pre-specied probability distribution. Itos Lemma: if v (x, t) is smooth enough, v (y (t), t) is again a jump-diusion, with 2vxx)dt + vxdw + [v (y (t) + J, t) v (y (t), t)]dN. d[v (y (t), t)] = (vt + vx + 1 2

Dynamics of Conditional Expectations

Now consider the expected nal-time payo u(x, t) = Ey(t)=x [w(y (T ))] Here w(x) is an arbitrary payo (later it will be the payo of an option). Solves a backward Kolmogorov equation ut + Lu = 0 for t < T , with u(x, T ) = w(x) at t = T . The operator L is
2 Lu = ux + 1 2 uxx + E [u(x + J, t) u(x, t)] .

(1)

The expectation in the last term is over the probability distribution of jumps (in the log price).

 Let u solve (1), and apply It os formula:

T T

u(y (T ), T ) u(x, t) =
0

(ux)(y (s), s) dw +
0 T

(us + ux + 1 2uxx)(y (s), s) ds 2

+
0

[u(y (s) + J, s) u(y (s), s)]dN.

Take the expectation, noting that the jump magnitudes, J , are independent of the Poisson jump occurence process, N : E ([u(y (s) + J, s) u(y (s), s)]dN ) = E ([u(y (s) + J, s) u(y (s), s)]) ds. Thus when u solves (1), we get: E [u(y (T ), T )] u(x, t) = 0. This gives the result, since u(y (T ), T ) = w(y (T )).

Adding Interest Rates (Finally)

Assume a (nonzero) constant interest rate r A similar argument shows u(x, t) = Ey(t)=x er(T t)w(y (T )) solves ut + Lu ru = 0 for t < T , with u(x, T ) = w(x) at t = T , using the same operator L. The probability distribution solves the forward Kolmogorov equation, ps L p = 0 for s > 0, with p(z, 0) = p0 (z ) p0 is the initial probability distribution, L is the adjoint of L.

 What is the adjoint of the new jump term? For any functions (z ), (z ) we have

E [[ (z + J ) (z )] (z ) dz = since
E [ (z

(z )E [[ (z J ) (z )] dz = J )] dz .

+ J )] (z ) dz =

(z )E [ (z

We have: L p = 1 ( 2p)zz (p)z + E [p(z J ) p(z )] , 2 And so:

2 ( p)zz + (p)z E [p(z J, s) p(z, s)] = 0. ps 1 2

Hedging and the Risk-Neutral Process

Assume that one can only trade the underlying stock and a riskfree asset. Further assume that Mertons jump idusion process governs log prices. Without further assumptions, no-arbitrage cannot be used to give a unique price. Still, the payo w(S ) should be the discounted nal-time payo under the riskneutral dynamics. Stock price dynamics under statistical measure P are:
2 J dS = ( + 1 ) Sdt + Sdw + ( e 1)SdN. 2

Hedging and the Risk-Neutral Process

Recall that the stock dynamics under statistical measure P are:
2 J dS = ( + 1 2 )Sdt + Sdw + (e 1)SdN.

Merton proposed that the risk-neutral process be determined by two considerations: (a) it has the same volatility and jump statistics i.e. it diers from the subjective process only by having a dierent drift; and (b) under the risk-neutral process ertS is a martingale, i.e. dS rSdt has mean value 0. Risk-neutral process is dS = (r E [eJ 1])Sdt + Sdw + (eJ 1)SdN. (2)

Hedging and the Risk-Neutral Process

Applying It os formula once more, we see that under the risk-neutral dynamics y = log S satises
2 J E [ e 1])dt + dw + JdN dy = (r 1 2

Can we use = r 1 2 E [eJ 1] to price options? 2 Need to be able to hedge. Try hedging a long position in the option by a short position of units of stock: d[u(S (t), t)] dS = utdt + uS ([ + 1 2]Sdt + Sdw) + 1 u 2S 2 dt 2 2 SS +[u(eJ S (t), t) u(S (t), t)]dN ([ + 1 2]Sdt + Sdw) (eJ 1)SdN. 2 Market is incomplete, no choice of makes this portfolio risk-free.

10

 Choose = uS (S (t), t) Randomness due to dw cancels, leaving only the uncertainty due to jumps:
2 2 J J portfolio gain = (ut + 1 2 S uSS )dt + {[u(e S (t), t) u(S (t), t)] uS (e S S )}dN.

Merton: assume jumps uncorrelated with the marketplace. Impact of such randomness can be eliminated by diversication. According the the Capital Asset Pricing Model, for such an investment (whose is zero) only the mean return is relevant to pricing. So the mean return on our hedge portfolio should be the risk-free rate: (ut + 1 2S 2 uSS )dt + E [u(eJ S (t), t) u(S (t), t) (eJ S S )uS ]dt = r(u SuS )dt. 2 (3) Obtain the backward Kolmogorov equation describing the discounted nal-time payo under the risk-neutral dynamics (2): 2S 2 uSS ru + E [u(eJ S, t) u(S, t)] = 0. ut + (r E [eJ 1])SuS + 1 2

11

Convex Payos
Calls and puts have convex payos in S As a result, the jump term in (3) is positive: E [u(eJ S (t), t) u(S (t), t) (eJ S S )uS ] 0. Between jumps, the hedge portfolio rises slower than risk free rate. Jumps work in favour of the option holder.

12

Why Model Jumps?

Better able to t smile. There exists a consistent theoretical framework. Can experiment with adapting the stock hedge or hedgin with options. The model can be calibrated to plain vanilla options and used to price and (partially) hedge exotic options.

13