A stable local radial basis function method for option pricing problem under the Bates model

Abstract

[EN] We propose a local mesh-free method for the Bates¿Scott option pricing model, a 2D partial integro-differential equation (PIDE) arising in computational finance. A Wendland radial basis function (RBF) approach is used for the discretization of the spatial variables along with a linear interpolation technique for the integral operator. The resulting set of ordinary differential equations (ODEs) is tackled via a time integration method. A potential advantage of using RBFs is the small number of discrete equations that need to be solved. Computational experiments are presented to illustrate the performance of the contributed approach.