A computational weighted finite difference method for American and barrier options in subdiffusive Black–Scholes model

•The system describing the fair price of American put option in subdiffusive Black–Scholes model is derived.•The weighted finite difference method for the class of problems is introduced.•The formula for the optimal choice of discretization parameter is given.•The Longstaff–Schwartz method is ineffi...

Full description

Saved in:
Bibliographic Details
Published in:Communications in nonlinear science & numerical simulation 2021-05, Vol.96, p.105676, Article 105676
Main Authors: Krzyżanowski, Grzegorz, Magdziarz, Marcin
Format: Article
Language:English
Subjects:
Algorithms
American option numerical evaluation
Diffusion
Diffusion barriers
Finite difference method
Finite element analysis
Fluid dynamics
Mathematical analysis
Physics
Schwartz method
Subdiffusion
Time fractional Black–Scholes model
Weighted finite difference method
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•The system describing the fair price of American put option in subdiffusive Black–Scholes model is derived.•The weighted finite difference method for the class of problems is introduced.•The formula for the optimal choice of discretization parameter is given.•The Longstaff–Schwartz method is inefficient for the subdiffusive models. Subdiffusion is a well established phenomenon in physics. In this paper we apply the subdiffusive dynamics to analyze financial markets. We focus on the financial aspect of time fractional diffusion model with moving boundary i.e. American and barrier option pricing in the subdiffusive Black–Scholes (B–S) model. Two computational methods for valuing American options in the considered model are proposed - the weighted finite difference (FD) and the Longstaff–Schwartz method. In the article it is also shown how to valuate numerically wide range of barrier options using the FD approach.
ISSN:1007-5704
1878-7274
DOI:10.1016/j.cnsns.2020.105676