Learning Exercise Policies for American Options
; Proceedings of the Twelth International Conference on Artificial Intelligence and Statistics, PMLR 5:352-359, 2009.
Options are important instruments in modern finance. In this paper, we investigate reinforcement learning (RL) methods—in particular, least-squares policy iteration (LSPI)—for the problem of learning exercise policies for American options. We develop finite-time bounds on the performance of the policy obtained with LSPI and compare LSPI and the fitted Q-iteration algorithm (FQI) with the Longstaff-Schwartz method (LSM), the standard least-squares Monte Carlo algorithm from the finance community. Our empirical results show that the exercise policies discovered by LSPI and FQI gain larger payoffs than those discovered by LSM, on both real and synthetic data. Furthermore, we find that for all methods the policies learned from real data generally gain similar payoffs to the policies learned from simulated data. Our work shows that solution methods developed in machine learning can advance the state-of-the-art in an important and challenging application area, while demonstrating that computational finance remains a promising area for future applications of machine learning methods.