A MONOTONE SCHEME FOR HIGH-DIMENSIONAL FULLY NONLINEAR PDES

• 成果类型: Article
• 署名作者: Guo, Wenjie; Zhang, Jianfeng; Zhuo, Jia
• 署名单位: Fudan University; University of Southern California
• 刊物名称: ANNALS OF APPLIED PROBABILITY
• ISSN/ISSBN: 1050-5164
• DOI: 10.1214/14-AAP1030
• 发表日期: 2015
• 页码: 1540-1580
• 关键词: STOCHASTIC DIFFERENTIAL-EQUATIONS; linear parabolic equations; American options; backward sdes; approximation; Discretization; simulation

摘要

this paper we propose a feasible numerical scheme for high-dimensional, fully nonlinear parabolic PDEs, which includes the quasi-linear PDE associated with a coupled FBSDE as a special case. Our paper is strongly motivated by the remarkable work Fahim, Touzi and Warin [Ann. Appl. Probab. 21 (2011) 1322-1364] and stays in the paradigm of monotone schemes initiated by Barles and Souganidis [Asymptot. Anal. 4 (1991) 271-283]. Our scheme weakens a critical constraint imposed by Fahim, Touzi and Warin (2011), especially when the generator of the PDE depends only on the diagonal terms of the Hessian matrix. Several numerical examples, up to dimension 12, are reported.