Machine learning methods for American-style path-dependent contracts

42 Pages Posted: 5 Dec 2023

Date Written: November 28, 2023

Abstract

In the present work, we introduce and compare state-of-the-art algorithms, that are now classified under the name of machine learning, to price Asian and look-back products with early-termination features. These include randomized feed-forward neural networks, randomized recurrent neural networks, and a novel method based on signatures of the underlying price process. Additionally, we explore potential applications on callable certificates. Furthermore, we present an innovative approach for calculating sensitivities, specifically Delta and Gamma, leveraging Chebyshev interpolation techniques.

Keywords: Amerasian options, Look-back options, Callable certificates, Early termination, Random networks, Signature methods, Least-square Monte Carlo, Chebyshev Greeks

JEL Classification: C63, G13

Suggested Citation

Gambara, Matteo and Livieri, Giulia and Pallavicini, Andrea, Machine learning methods for American-style path-dependent contracts (November 28, 2023). Available at SSRN: https://ssrn.com/abstract=4646847 or http://dx.doi.org/10.2139/ssrn.4646847

Matteo Gambara

INAIT SA ( email )

Av. du Tribunal-Fédéral 34
Lausanne, 1005
Switzerland

Giulia Livieri

Scuola Normale Superiore ( email )

Piazza dei Cavalieri, 7
Pisa, 56126
Italy

Andrea Pallavicini (Contact Author)

Intesa Sanpaolo ( email )

Largo Mattioli 3
Milan, MI 20121
Italy

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
116
Abstract Views
4,911
Rank
448,296
PlumX Metrics