On Lévy processes, Malliavin calculus and market models with jumps (Q1849791)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | On Lévy processes, Malliavin calculus and market models with jumps |
scientific article |
Statements
On Lévy processes, Malliavin calculus and market models with jumps (English)
0 references
1 December 2002
0 references
This paper deals with the Malliavin calculus for Lévy processes and applications for option hedging in a jump-diffusion model. The authors introduce the notion of iterated integral of \(f\in L^2(R_{+}^{n})\) with respect to processes \(H_{t}^{(i)}\), \(i=i_1,\dots,i_{n}\), where \(H_{t}^{(i)}= \sum_{j=1}^{i}a_{ij}(X_{t}^{(j)}-m_{j}t)\); \(X_{t}^{1}=X_{t}\), \(X_{t}^{(i)}=\sum_{0<s\leq t}(\Delta X_{s})^{i}, i\geq 2\); \(E(X_{t}^{(i)})=m_{i}t\); \(X_{t}\) is the Lévy process such that \(\exists \varepsilon>0\), \(\delta>0\), \(\int_{(-\varepsilon,\varepsilon)^{c}}e^{\delta|x|} \nu(dx)<\infty\); \(\nu\) is the Lévy measure of \(X\); the constants \(a_{ij}\) are chosen in such a way that \(a_{i1}=1\) and the martingales \(H^{(i)}\) are pairwise strongly orthogonal. Properties of the iterated integrals as well as properties of the Malliavin derivatives such as chaotic and the predictable representations of the square integrable random variables, and Clark-Ocone formula, are studied. Some formulas that help to compute the Malliavin derivatives for a simple Lévy process \(X_{t}=\sigma W_{t}+ \alpha_{1}N_{1}(t)+\dots+\alpha_{k}N_{k}(t)\), \(t\geq 0\), are obtained. Here \(W_{t}\) is a standard Brownian motion, \(N_{j}(t)\), \(j=1,\dots,k\), are independent Poisson processes; \(\sigma>0\), \(\alpha_{1},\dots,\alpha_{k}\), are non-zero constants. A theorem on approximation of the Lévy process of the form \(X_{t}=\sigma W_{t}+\sum_{j=1}^{N_{t}}Z_{j}\) by simple Lévy processes is proved. Formulas for the approximate hedge of a European call are derived in the jump-diffusion model \(dA(t)=rA(t) dt\), \(A(0)=1\), \(r>0\), \(dS(t)=S(t-) dX(t)\), \(S(0)=s_0\), where \(s_0\) is a constant, \(X(t)\) is the Lévy process.
0 references
Lévy processes
0 references
Malliavin calculus
0 references
jump diffusion model
0 references
option hedging
0 references