# Solving high-dimensional optimal stopping problems using deep learning

@article{Becker2021SolvingHO,
title={Solving high-dimensional optimal stopping problems using deep learning},
author={Sebastian Becker and Patrick Cheridito and Arnulf Jentzen and Timo Welti},
journal={European Journal of Applied Mathematics},
year={2021},
volume={32},
pages={470 - 514}
}
• S. Becker, +1 author Timo Welti
• Published 5 August 2019
• Computer Science, Mathematics, Economics
• European Journal of Applied Mathematics
Nowadays many financial derivatives, such as American or Bermudan options, are of early exercise type. Often the pricing of early exercise options gives rise to high-dimensional optimal stopping problems, since the dimension corresponds to the number of underlying assets. High-dimensional optimal stopping problems are, however, notoriously difficult to solve due to the well-known curse of dimensionality. In this work, we propose an algorithm for solving such problems, which is based on deep… Expand

#### Topics from this paper

Financial Option Valuation by Unsupervised Learning with Artificial Neural Networks
• Computer Science, Economics
• Mathematics
• 2020
The classical problem of pricing European and American financial options, based on the corresponding PDE formulations, is studied here, using artificial neural networks in the context of unsupervised learning. Expand
Neural network regression for Bermudan option pricing
• Computer Science, Mathematics
• Monte Carlo Methods Appl.
• 2021
This work proves the convergence of the well-known Longstaff and Schwartz algorithm when the standard least-square regression is replaced by a neural network approximation, assuming an efficient algorithm to compute this approximation. Expand
Deep neural network framework based on backward stochastic differential equations for pricing and hedging American options in high dimensions
• Computer Science, Economics
• ArXiv
• 2019
This work proposes a deep neural network framework for computing prices and deltas of American options in high dimensions, and introduces the least squares residual of the associated backward stochastic differential equation as the loss function. Expand
Space-time deep neural network approximations for high-dimensional partial differential equations
• Computer Science, Mathematics
• ArXiv
• 2020
The main result of this work proves for every a and b that solutions of certain Kolmogorov PDEs can be approximated by DNNs on the space-time region $[0,T]\times [a,b]^d$ without the curse of dimensionality. Expand
Deep neural network approximations for Monte Carlo algorithms
• Computer Science, Mathematics
• ArXiv
• 2019
The main result of this paper shows that if a function can be approximated by means of some suitable discrete approximation scheme without the curse of dimensionality and if there exist DNNs which satisfy certain regularity properties and which approximate this discrete approximation Scheme, then the function itself can also be approximating with DNN's without the cursed dimensionality. Expand
N A ] 2 8 A ug 2 01 9 Deep neural network approximations for Monte Carlo algorithms
In the past few years deep artificial neural networks (DNNs) have been successfully employed in a large number of computational problems including, e.g., language processing, image recognition, fraudExpand
Deep combinatorial optimisation for optimal stopping time problems and stochastic impulse control. Application to swing options pricing and fixed transaction costs options hedging
• Economics, Mathematics
• 2020
A new method for stochastic control based on neural networks and using randomisation of discrete random variables is proposed and applied to optimal stopping time problems and seems to be competitive with the best existing algorithms both in terms of Precision and computation time. Expand
Neural Network Pricing of American Put Options
• Computer Science
• 2020
All methods present a good accuracy; however, after calibration, NNs produce better results in terms of both execution time and Root Mean Squared Error (RMSE). Expand
A differential neural network learns stochastic differential equations and the Black-Scholes equation for pricing multi-asset options
A differential neural network is applied to the pricing of financial options, where stochastic differential equations and the Black-Scholes partial differential equation represent the relation of price of option and underlying assets, and the first and second derivatives, Greeks, of option play important roles in financial engineering. Expand
Solving high-dimensional Hamilton-Jacobi-Bellman PDEs using neural networks: perspectives from the theory of controlled diffusions and measures on path space
• Computer Science, Mathematics
• Partial Differential Equations and Applications
• 2021
The potential of iterative diffusion optimisation techniques is investigated, in particular considering applications in importance sampling and rare event simulation, and focusing on problems without diffusion control, with linearly controlled drift and running costs that depend quadratically on the control. Expand

#### References

SHOWING 1-10 OF 124 REFERENCES
Deep Learning-Based BSDE Solver for Libor Market Model with Applications to Bermudan Swaption Pricing and Hedging
• Computer Science, Economics
• 2018
It is demonstrated that using backward DNN the high-dimension Bermudan swaption pricing and hedging can be solved effectively and efficiently. Expand
Implied stopping rules for American basket options from Markovian projection
• Mathematics, Economics
• 2017
This work addresses the problem of pricing American basket options in a multivariate setting, which includes among others, the Bachelier and Black–Scholes models. In high dimensions, nonlinear PDEExpand
Beating the curse of dimensionality in options pricing and optimal stopping
• Computer Science, Mathematics
• ArXiv
• 2018
This work develops a novel pure-dual algorithm that allows one to elegantly trade-off between accuracy and runtime through a parameter epsilon controlling the associated performance guarantee, with computational and sample complexity both polynomial in T and effectively independent of the dimension, in contrast to past methods typically requiring a complexity scaling exponentially in these parameters. Expand
Machine learning for pricing American options in high-dimensional Markovian and non-Markovian models
• Computer Science, Economics
• 2019
Two efficient techniques which allow one to compute the price of American basket options, called GPR Tree and GPR Exact Integration, are proposed, which solve the backward dynamic programing problem considering a Bermudan approximation of the American option. Expand
Solving high-dimensional partial differential equations using deep learning
• Mathematics, Computer Science
• Proceedings of the National Academy of Sciences
• 2018
A deep learning-based approach that can handle general high-dimensional parabolic PDEs using backward stochastic differential equations and the gradient of the unknown solution is approximated by neural networks, very much in the spirit of deep reinforcement learning with the gradient acting as the policy function. Expand
A dynamic look-ahead Monte Carlo algorithm for pricing Bermudan options
• Mathematics
• 2007
Under the assumption of no-arbitrage, the pricing of American and Bermudan options can be casted into optimal stopping problems. We propose a new adaptive simulation based algorithm for the numericalExpand
Neural network regression for Bermudan option pricing
• Computer Science, Mathematics
• Monte Carlo Methods Appl.
• 2021
This work proves the convergence of the well-known Longstaff and Schwartz algorithm when the standard least-square regression is replaced by a neural network approximation, assuming an efficient algorithm to compute this approximation. Expand
High Dimensional American Options
Pricing single asset American options is a hard problem in mathematical finance. There are no closed form solutions available (apart from in the case of the perpetual option), so many approximationsExpand
Pricing high-dimensional Bermudan options using the stochastic grid method
• Computer Science
• Int. J. Comput. Math.
• 2012
This paper proposes a stochastic grid method for estimating the optimal exercise policy and uses this policy to obtain a low-biased estimator for high-dimensional Bermudan options. Expand
Deep neural network framework based on backward stochastic differential equations for pricing and hedging American options in high dimensions
• Computer Science, Economics
• ArXiv
• 2019
This work proposes a deep neural network framework for computing prices and deltas of American options in high dimensions, and introduces the least squares residual of the associated backward stochastic differential equation as the loss function. Expand