Open Access
October, 2019 A Method-of-lines Approach for Solving American Option Problems
Min-Sun Horng, Tzyy-Leng Horng, Chih-Yuan Tien
Taiwanese J. Math. 23(5): 1253-1270 (October, 2019). DOI: 10.11650/tjm/181010

Abstract

The early exercise property of American option changes the original Black-Scholes equation to an inequality that cannot be solved via traditional finite difference method. Therefore, finding the early exercise boundary prior to spatial discretization is a must in each time step. This overhead slows down the computation and the accuracy of solution relies on if the early exercise boundary can be accurately located. A simple numerical method based on finite difference and method of lines is proposed here to overcome this difficulty in American option valuation. Our method averts the otherwise necessary procedure of locating the optimal exercise boundary before applying finite difference discretization. The method is efficient and flexible to all kinds of pay-off. Computations of American put, American call with dividend, American strangle, two-factor American basket put option, and two-factor convertible bond with embedded call and put options are demonstrated to show the efficiency of the current method.

Citation

Download Citation

Min-Sun Horng. Tzyy-Leng Horng. Chih-Yuan Tien. "A Method-of-lines Approach for Solving American Option Problems." Taiwanese J. Math. 23 (5) 1253 - 1270, October, 2019. https://doi.org/10.11650/tjm/181010

Information

Received: 2 April 2018; Revised: 2 October 2018; Accepted: 23 October 2018; Published: October, 2019
First available in Project Euclid: 8 November 2018

zbMATH: 07126947
MathSciNet: MR4012378
Digital Object Identifier: 10.11650/tjm/181010

Subjects:
Primary: 62P05 , 65M06 , 65M20 , 91G60 , 97M30

Keywords: American option , American strangle option , callable and putable convertible bond , finite difference method , method of lines , two-factor American basket put option

Rights: Copyright © 2019 The Mathematical Society of the Republic of China

Vol.23 • No. 5 • October, 2019
Back to Top