Interpretable Optimal Stopping

Management Science, Forthcoming

INSEAD Working Paper No. 2020/01/TOM

85 Pages Posted: 3 Oct 2018 Last revised: 9 Jan 2020

See all articles by Dragos Ciocan

Dragos Ciocan


Velibor Mišić

University of California, Los Angeles (UCLA) - Anderson School of Management

Date Written: January 8, 2020


Optimal stopping is the problem of deciding when to stop a stochastic system to obtain the greatest reward, arising in numerous application areas such as finance, healthcare and marketing. State-of-theart methods for high-dimensional optimal stopping involve approximating the value function or the continuation value, and then using that approximation within a greedy policy. Although such policies can perform very well, they are generally not guaranteed to be interpretable; that is, a decision maker may not be able to easily see the link between the current system state and the policy's action. In this paper, we propose a new approach to optimal stopping, wherein the policy is represented as a binary tree, in the spirit of naturally interpretable tree models commonly used in machine learning. We show that the class of tree policies is rich enough to approximate the optimal policy. We formulate the problem of learning such policies from observed trajectories of the stochastic system as a sample average approximation (SAA) problem. We prove that the SAA problem converges under mild conditions as the sample size increases, but that computationally even immediate simplifications of the SAA problem are theoretically intractable. We thus propose a tractable heuristic for approximately solving the SAA problem, by greedily constructing the tree from the top down. We demonstrate the value of our approach by applying it to the canonical problem of option pricing, using both synthetic instances and instances using real S&P-500 data. Our method obtains policies that (1) outperform state-of-the-art noninterpretable methods, based on simulation-regression and martingale duality, and (2) possess a remarkably simple and intuitive structure.

Keywords: Optimal Stopping, Approximate Dynamic Programming, Interpretability, Decision Trees, Option Pricing

JEL Classification: C60

Suggested Citation

Ciocan, Dragos and Misic, Velibor, Interpretable Optimal Stopping (January 8, 2020). Management Science, Forthcoming, INSEAD Working Paper No. 2020/01/TOM, Available at SSRN: or

Dragos Ciocan (Contact Author)

INSEAD ( email )

Boulevard de Constance
77305 Fontainebleau Cedex

Velibor Misic

University of California, Los Angeles (UCLA) - Anderson School of Management ( email )

110 Westwood Plaza
Los Angeles, CA 90095-1481
United States

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

Paper statistics

Abstract Views
PlumX Metrics