Supervised Machine Learning with Control Variates for American Option Pricing
2018
scientific article
english
- American options
- Monte Carlo
- Gaussian processes
- Kriging
- LSM
- supervised learning
- Heston Model
- control variates
EN In this paper, we make use of a Bayesian (supervised learning) ap-proach in pricing American options via Monte Carlo simulations. We first presentGaussian process regression (Kriging) approach for American options pricing andcompare its performance in estimating the continuation value with the Longstaff andSchwartz algorithm. Secondly, we explore the control variates technique in combina-tion with Kriging to further improve the estimation of the continuation value. Thismethod allows to reduce dramatically the standard errors and to improve the stabilityof the Kriging approach. For illustrative purposes, we use American put options ona stock whose dynamics is given by Heston model, and use European options on thesame stock as control variates.
207 - 217
CC BY-NC-ND (attribution - noncommercial - no derivatives)
public
15
15