The Bismut-Elworthy-Li formula for jump-diffusions and applications to Monte Carlo pricing in finance
From MaRDI portal
Publication:6477016
arXivmath/0604311MaRDI QIDQ6477016FDOQ6477016
Authors: Thomas Cass, Peter Friz 
Publication date: 13 April 2006
Abstract: We extend the Bismut-Elworthy-Li formula to non-degenerate jump diffusions and "payoff" functions depending on the process at multiple future times. In the spirit of Fournie et al [13] and Davis and Johansson [9] this can improve Monte Carlo numerics for stochastic volatility models with jumps. To this end one needs so-called Malliavin weights and we give explicit formulae valid in presence of jumps: (a) In a non-degenerate situation, the extended BEL formula represents possible Malliavin weights as Ito integrals with explicit integrands; (b) in a hypoelliptic setting we review work of Arnaudon and Thalmaier [1] and also find explicit weights, now involving the Malliavin covariance matrix, but still straight-forward to implement. (This is in contrast to recent work by Forster, Lutkebohmert and Teichmann where weights are constructed as anticipating Skorohod integrals.) We give some financial examples covered by (b) but note that most practical cases of poor Monte Carlo performance, Digital Cliquet contracts for instance, can be dealt with by the extended BEL formula and hence without any reliance on Malliavin calculus at all. We then discuss some of the approximations, often ignored in the literature, needed to justify the use of the Malliavin weights in the context of standard jump diffusion models. Finally, as all this is meant to improve numerics, we give some numerical results with focus on Cliquets under the Heston model with jumps.
Applications of statistics to actuarial sciences and financial mathematics (62P05) Applications of stochastic analysis (to PDEs, etc.) (60H30)
This page was built for publication: The Bismut-Elworthy-Li formula for jump-diffusions and applications to Monte Carlo pricing in finance
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6477016)
