ε-Monotone Fourier Methods for Optimal Stochastic Control in Finance

48 Pages Posted: 26 Feb 2019

See all articles by Peter Forsyth

Peter Forsyth

University of Waterloo - David R. Cheriton School of Computer Science

George Labahn

University of Waterloo - David R. Cheriton School of Computer Science

Date Written: February 25, 2019

Abstract

Stochastic control problems in finance often involve complex controls at discrete times. As a result, numerically solving such problems using, for example, methods based on partial differential or integrodifferential equations inevitably gives rise to low-order (usually at most second-order) accuracy. In many cases, Fourier methods can be used to efficiently advance solutions between control monitoring dates, and numerical optimization methods can then be applied across decision times. However, Fourier methods are not monotone, and as a result they give rise to possible violations of arbitrage inequalities. This is problematic in the context of control problems, where the control is determined by comparing value functions. In this paper, we give a preprocessing step for Fourier methods that involves projecting the Green’s function onto the set of linear basis functions. The resulting algorithm is guaranteed to be monotone (to within a tolerance), ℓ ∞-stable and satisfies an ε-discrete comparison principle. In addition, the algorithm has the same complexity per step as a standard Fourier method and second-order accuracy for smooth problems.

Keywords: monotonicity, Fourier methods, discrete comparison, optimal stochastic control, finance

Suggested Citation

Forsyth, Peter and Labahn, George, ε-Monotone Fourier Methods for Optimal Stochastic Control in Finance (February 25, 2019). Journal of Computational Finance, Vol. 22, No. 4, 2019, Available at SSRN: https://ssrn.com/abstract=3341491

Peter Forsyth (Contact Author)

University of Waterloo - David R. Cheriton School of Computer Science ( email )

200 University Avenue West
Waterloo, ON
Canada

George Labahn

University of Waterloo - David R. Cheriton School of Computer Science ( email )

200 University Avenue West
Waterloo, ON
Canada

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
1
Abstract Views
503
PlumX Metrics