A Numerical Study of Radial Basis Function Based Methods for Options Pricing under the One Dimension Jump-diffusion Model
For computational finance practitioners, this provides an alternative numerical method for option pricing under jump-diffusion models, though it is an incremental application of existing RBF techniques.
This study applies meshless Radial Basis Function (RBF) methods to price options under jump-diffusion models, achieving second-order spatial accuracy for both European and American options, and first-order temporal accuracy for American options.
The aim of this chapter is to show how option prices in jump-diffusion models can be computed using meshless methods based on Radial Basis Function (RBF) interpolation. The RBF technique is demonstrated by solving the partial integro-differential equation (PIDE) in one-dimension for the American put and the European vanilla call/put options on dividend-paying stocks in the Merton and Kou jump-diffusion models. The radial basis function we select is the Cubic Spline. We also propose a simple numerical algorithm for finding a finite computational range of an improper integral term in the PIDE so that the accuracy of approximation of the integral can be improved. Moreover, the solution functions of the PIDE are approximated explicitly by RBFs which have exact forms so we can easily compute the global integral by any kind of numerical quadrature. Finally, we will not only show numerically that our scheme is second order accurate in both spatial and time variables in a European case but also second order accurate in spatial variables and first order accurate in time variables in an American case.