Esscher transform and the duality principle for multidimensional semimartingales (Q983888)

The authors consider options that depend on several assets; in general, the valuation of such options requires the knowledge of the joint distribution and a high-dimensional integration. The aim of the paper is to simplify this valuation problem by relating it to its dual option pricing problem. The main result describes the triplet of predictable characteristics of one-dimensional semimartingales -- defined as the inner product of a vector with the driving multidimensional semimartingale -- under the dual measure. The dual measures are constructed via an Esscher transformation. As an application, swap options and quanto options are related to standard European call or put options, for general semimartingale models. For semimartingales with independent increments, a duality relationship is derived between an option depending on three assets and a standard call or put option. This yields a significant reduction in the computational complexity of these valuation problems. Explicit calculations for jump models are provided.