Algorithmic Differentiation: Adjoint Greeks Made Easy

12 Pages Posted: 4 Apr 2011

See all articles by Luca Capriotti

Luca Capriotti

Columbia University

Michael B. Giles

University of Oxford - Oxford-Man Institute of Quantitative Finance

Date Written: April 2, 2011

Abstract

We show how algorithmic differentiation can be used as a design paradigm to implement the adjoint calculation of sensitivities in Monte Carlo in full generality and with minimal analytical effort. With several examples we illustrate the workings of this technique and demonstrate how it can be straightforwardly implemented to reduce the time required for the computation of the risk of any portfolio by orders of magnitude.

Keywords: Algorithmic Differentiation, Monte Carlo Simulations, Derivatives Pricing

Suggested Citation

Capriotti, Luca and Giles, Michael B., Algorithmic Differentiation: Adjoint Greeks Made Easy (April 2, 2011). Available at SSRN: https://ssrn.com/abstract=1801522 or http://dx.doi.org/10.2139/ssrn.1801522

Luca Capriotti (Contact Author)

Columbia University ( email )

3022 Broadway
New York, NY 10027
United States

Michael B. Giles

University of Oxford - Oxford-Man Institute of Quantitative Finance ( email )

Eagle House
Walton Well Road
Oxford, Oxfordshire OX2 6ED
United Kingdom

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
3,205
Abstract Views
9,948
Rank
7,045
PlumX Metrics