A new deep neural network algorithm for multiple stopping with applications in options pricing

Abstract

In this paper, we propose a deep learning method to solve high-dimensional optimal multiple stopping problems. We represent the policies of multiple stopping problems by the composition of functions. Using the new representation, we approximate the optimal stopping policy recursively with simulation samples. We also derive lower and upper bounds and confidence intervals for the values. Finally, we apply the algorithm to the pricing of swing options, and it produces accurate results in high-dimensional problems.