Overcoming the Feature Selection Issue in the Pricing of American Options

24 Pages Posted: 28 Jan 2022

See all articles by Peter Lind

Peter Lind

Aalborg University Business School

Date Written: January 21, 2022

Abstract

The feedforward neural network Monte Carlo method (FNNMC) exhibits more robustness and
accuracy than the state-of-the-art least squares Monte Carlo method (LSM) in pricing several
American-style options. Specifically, the FNNMC price estimates are accurate for basket options,
where the FNNMC price errors are more than four times smaller than the LSM with the best
choice of basis functions. By training the neural network the FNNMC avoids the issue of choosing
a proper set of basis functions. Hence we circumvent manually engineering the features for each
type of option. Furthermore, we explore in-depth the hyperparameter selection for the FNNMC.
In the exploration, we use a novel approach called price grid search, where the search is done at
the price level instead of at the usual regression level.

Keywords: American options, option theory, least squares Monte Carlo method, deep learning, feedforward neural network Monte Carlo method

JEL Classification: G13, C02, C45, C61

Suggested Citation

Lind, Peter Pommergård, Overcoming the Feature Selection Issue in the Pricing of American Options (January 21, 2022). Available at SSRN: https://ssrn.com/abstract=4014593 or http://dx.doi.org/10.2139/ssrn.4014593

Peter Pommergård Lind (Contact Author)

Aalborg University Business School

FibigerStræde 2, 41
Aalborg Ø, North Jutland 9220
Denmark
(+45) 9940 2727 (Phone)

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

Paper statistics

Downloads
149
Abstract Views
387
rank
284,015
PlumX Metrics