Randomized quasi-Monte Carlo methods in pricing securities

Abstract

Today quasi-Monte Carlo methods are used successfully in computational finance and economics as an alternative to the Monte Carlo method. One drawback of these methods, however, is the lack of a practical way of error estimation. To address this issue several researchers introduced the so-called randomized quasi-Monte Carlo methods in the last decade. In this paper we will present a survey of randomized quasi-Monte Carlo methods, and compare their efficiencies with the efficiency of the Monte Carlo method in pricing certain securities. We will also investigate the effects of Box-Muller and inverse transformation techniques when they are applied to low-discrepancy sequences. © 2004 Elsevier B.V. All rights reserved.