On smoothing of the Crank-Nicolson scheme and higher order schemes for pricing barrier options

  • B. A. Wade*
  • , A. Q.M. Khaliq
  • , M. Yousuf
  • , J. Vigo-Aguiar
  • , R. Deininger
  • *Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

49 Scopus citations

Abstract

Most option pricing problems have nonsmooth payoffs or discontinuous derivatives at the exercise price. Discrete barrier options have not only nonsmooth payoffs but also time dependent discontinuities. In pricing barrier options, certain aspects are triggered if the asset price becomes too high or too low. Standard smoothing schemes used to solve problems with nonsmooth payoff do not work well for discrete barrier options because of discontinuities introduced in the time domain when each barrier is applied. Moreover, these unwanted oscillations become worse when estimating the hedging parameters, e.g., Delta and Gamma. We have an improved smoothing strategy for the Crank-Nicolson method which is unique in achieving optimal order convergence for barrier option problems. Numerical experiments are discussed for one asset and two asset problems. Time evolution graphs are obtained for one asset problems to show how option prices change with respect to time. This smoothing strategy is then extended to higher order methods using diagonal (m, m)-Padé main schemes under a smoothing strategy of using as damping schemes the (0, 2 m - 1) subdiagonal Padé schemes.

Original languageEnglish
Pages (from-to)144-158
Number of pages15
JournalJournal of Computational and Applied Mathematics
Volume204
Issue number1
DOIs
StatePublished - 1 Jul 2007
Externally publishedYes

Keywords

  • Barrier options
  • Black-Scholes PDE
  • Crank-Nicolson scheme
  • Padé schemes
  • Smoothing

ASJC Scopus subject areas

  • Computational Mathematics
  • Applied Mathematics

Fingerprint

Dive into the research topics of 'On smoothing of the Crank-Nicolson scheme and higher order schemes for pricing barrier options'. Together they form a unique fingerprint.

Cite this