Deep Reinforcement Learning on a Multi-Asset Environment for Trading

18 Pages Posted: 28 Jun 2021

See all articles by Ali Hirsa

Ali Hirsa

Columbia University

Branka Hadji Misheva

Zurich University of Applied Sciences

Joerg Osterrieder

University of Twente; Bern Business School

Jan-Alexander Posth

ZHAW School of Management and Law

Date Written: June 15, 2021

Abstract

Financial trading has been widely analyzed for decades with market participants and academics always looking for advanced methods to improve trading performance. Deep reinforcement learning (DRL), a recently reinvigorated method with significant success in multiple domains, still has to show its benefit in the financial markets. We use a deep Q-network (DQN) to design long-short trading strategies for futures contracts. The state space consists of volatility-normalized daily returns, with buying or selling being the reinforcement learning action and the total reward defined as the cumulative profits from our actions. Our trading strategy is trained and tested both on real and simulated price series and we compare the results with an index benchmark. We analyze how training based on a combination of artificial data and actual price series can be successfully deployed in real markets.

The trained reinforcement learning agent is applied to trading the E-mini S&P 500 continuous futures contract. Our results in this study are preliminary and need further improvement.

Keywords: Reinforcement Learning, Deep Learning, Finance, Trading Strategies

JEL Classification: G00

Suggested Citation

Hirsa, Ali and Hadji Misheva, Branka and Osterrieder, Joerg and Posth, Jan-Alexander, Deep Reinforcement Learning on a Multi-Asset Environment for Trading (June 15, 2021). Available at SSRN: https://ssrn.com/abstract=3867800 or http://dx.doi.org/10.2139/ssrn.3867800

Ali Hirsa

Columbia University ( email )

500 West 120th Street
New York, NY 10027

HOME PAGE: http://www.ieor.columbia.edu/faculty/ali-hirsa

Branka Hadji Misheva

Zurich University of Applied Sciences ( email )

IDP
Technikumstrasse 9
Winterthur, CH 8401
Switzerland

Joerg Osterrieder (Contact Author)

University of Twente ( email )

Drienerlolaan 5
Departement of High-Tech Business and Entrepreneur
Enschede, 7522 NB
Netherlands

Bern Business School ( email )

Brückengasse
Institute of Applied Data Sciences and Finance
Bern, BE 3005
Switzerland

Jan-Alexander Posth

ZHAW School of Management and Law ( email )

Technoparkstrasse 2, P.O. Box
Winterthur, 8401
Switzerland

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
159
Abstract Views
627
Rank
336,214
PlumX Metrics