Using Model-Independent Lower Bounds to Improve Pricing of Asian Style Options in Levy Markets

Astin Bulletin, Forthcoming

44 Pages Posted: 18 Jan 2014  

Griselda Deelstra

Université Libre de Bruxelles (ULB)

Grégory Rayée

Université Libre de Bruxelles (ULB)

Steven Vanduffel

Vrije Universiteit Brussel (VUB)

Jing Yao

Vrije Universiteit Brussel (VUB)

Date Written: January 16, 2014

Abstract

Albrecher et al. (2008) have proposed model-independent lower bounds for arithmetic Asian options. In this paper we provide an alternative and more elementary derivation of their results. We use the bounds as control variates to develop a simple Monte Carlo method for pricing contracts with Asian style features. The conditioning idea that is inherent in our approach also inspires us to propose a new semi-analytic pricing approach. We compare both approaches and conclude that they both have their merits and are useful in practice. In particular, we point out that our newly proposed Monte Carlo method allows to deal with Asian style products that appear in insurance (e.g. unit linked contracts) in a very efficient way, and outperforms other known Monte Carlo methods that are based on control variates.

Keywords: Asian style options, conditional expectation, control variates, stochastic clock

JEL Classification: C00

Suggested Citation

Deelstra, Griselda and Rayée, Grégory and Vanduffel, Steven and Yao, Jing, Using Model-Independent Lower Bounds to Improve Pricing of Asian Style Options in Levy Markets (January 16, 2014). Astin Bulletin, Forthcoming. Available at SSRN: https://ssrn.com/abstract=2380319

Griselda Deelstra

Université Libre de Bruxelles (ULB) ( email )

Boulevard du Triomphe, CP210
Brussels, Brussels 1050
Belgium

Grégory Rayée

Université Libre de Bruxelles (ULB) ( email )

CP 114/04 Av FD Roosevelt 50
Brussels, 1050
Belgium

Steven Vanduffel (Contact Author)

Vrije Universiteit Brussel (VUB) ( email )

Pleinlaan 2
Brussels, Brabant 1050
Belgium

HOME PAGE: http://www.stevenvanduffel.com

Jing Yao

Vrije Universiteit Brussel (VUB) ( email )

Pleinlaan 2
Brussels, B-1050
Belgium

Paper statistics

Downloads
24
Abstract Views
248