Momentum-Space Approach to Asymptotic Expansion for Stochastic Filtering and other Problems

Abstract

This paper develops an asymptotic expansion technique in momentum space. It is shown that Fourier transformation combined with a polynomial-function approximation of the nonlinear terms gives a closed recursive system of ordinary differential equations (ODEs) as an asymptotic expansion of the conditional distribution appearing in stochastic filtering problems. Thanks to the simplicity of the ODE system, higher order calculation can be performed easily. Furthermore, solving ODEs sequentially with small sub-periods with updated initial conditions makes it possible to implement a substepping method for asymptotic expansion in a numerically efficient way. This is found to improve the performance significantly where otherwise the approximation fails badly. The method may be useful for other applications, such as, option pricing in finance as well as measure-valued stochastic dynamics in general.