Authors:
(1) Xiao-Yang Liu, Hongyang Yang, Columbia University (xl2427,hy2500@columbia.edu);
(2) Jiechao Gao, University of Virginia (jg5ycn@virginia.edu);
(3) Christina Dan Wang (Corresponding Author), New York University Shanghai (christina.wang@nyu.edu). Table of Links Abstract and 1 Introduction 2 Related Works and 2.1 Deep Reinforcement Learning Algorithms 2.2 Deep Reinforcement Learning Libraries and 2.3 Deep Reinforcement Learning in Finance 3 The Proposed FinRL Framework and 3.1 Overview of FinRL Framework 3.2 Application Layer 3.3 Agent Layer 3.4 Environment Layer 3.5 Training-Testing-Trading Pipeline 4 Hands-on Tutorials and Benchmark Performance and 4.1 Backtesting Module 4.2 Baseline Strategies and Trading Metrics 4.3 Hands-on Tutorials 4.4 Use Case I: Stock Trading 4.5 Use Case II: Portfolio Allocation and 4.6 Use Case III: Cryptocurrencies Trading 5 Ecosystem of FinRL and Conclusions, and References 4.4 Use Case I: Stock Trading We use FinRL to reproduce both [50] and [51] for stock trading. The ensemble strategy [51] combines three DRL algorithms (PPO [42], A2C [32] and DDPG [26]) to improve the robustness. The implementation is easy with FinRL. We choose three algorithms (PPO, A2C, DDPG) in the agent layer, and an environment with start and end dates in the environment layer. The implementations of DRL algorithms and data preprocessing are transparent to users, alleviating the programming and debugging workloads. Thus, FinRL greatly facilitates the strategy design, allowing users to focus on improving the trading performance. Fig. 5 and Table 3 show the backtesting performance on Dow 30 constituent stocks, accessed at 2020/07/01. The training period is from 2009/01/01 to 2020/06/30 on a daily basis, and the testing period is from 2020/07/01 to 2021/06/30. The performance in terms of multiple metrics is consistent with the results reported in [51] and [50], and here we show results in a recent trading period. We can see from the DJIA index that the trading period is a bullish market with an annual return of 32.84%. The ensemble strategy achieves a Sharpe ratio of 2.81 and an annual return of 52.61%. It beats A2C with a Sharpe ratio of 2.24, PPO with a Sharpe ratio of 2.23, DDPG with a Sharpe ratio of 2.05, DJIA with a Sharpe ratio of 2.02, and min-variance portfolio allocation with a Sharpe ratio of 1.98, respectively. Therefore, the backtesting performance demonstrates that FinRL successfully reproduces the ensemble strategy [51]. This paper is available on arxiv under CC BY 4.0 DEED license. Authors: (1) Xiao-Yang Liu, Hongyang Yang, Columbia University (xl2427,hy2500@columbia.edu); (2) Jiechao Gao, University of Virginia (jg5ycn@virginia.edu); (3) Christina Dan Wang (Corresponding Author), New York University Shanghai (christina.wang@nyu.edu). Authors: Authors: (1) Xiao-Yang Liu, Hongyang Yang, Columbia University (xl2427,hy2500@columbia.edu); (2) Jiechao Gao, University of Virginia (jg5ycn@virginia.edu); (3) Christina Dan Wang ( Corresponding Author ), New York University Shanghai (christina.wang@nyu.edu). Corresponding Author Table of Links Abstract and 1 Introduction Abstract and 1 Introduction 2 Related Works and 2.1 Deep Reinforcement Learning Algorithms 2 Related Works and 2.1 Deep Reinforcement Learning Algorithms 2.2 Deep Reinforcement Learning Libraries and 2.3 Deep Reinforcement Learning in Finance 2.2 Deep Reinforcement Learning Libraries and 2.3 Deep Reinforcement Learning in Finance 3 The Proposed FinRL Framework and 3.1 Overview of FinRL Framework 3 The Proposed FinRL Framework and 3.1 Overview of FinRL Framework 3.2 Application Layer 3.2 Application Layer 3.3 Agent Layer 3.3 Agent Layer 3.4 Environment Layer 3.4 Environment Layer 3.5 Training-Testing-Trading Pipeline 3.5 Training-Testing-Trading Pipeline 4 Hands-on Tutorials and Benchmark Performance and 4.1 Backtesting Module 4 Hands-on Tutorials and Benchmark Performance and 4.1 Backtesting Module 4.2 Baseline Strategies and Trading Metrics 4.2 Baseline Strategies and Trading Metrics 4.3 Hands-on Tutorials 4.3 Hands-on Tutorials 4.4 Use Case I: Stock Trading 4.4 Use Case I: Stock Trading 4.5 Use Case II: Portfolio Allocation and 4.6 Use Case III: Cryptocurrencies Trading 4.5 Use Case II: Portfolio Allocation and 4.6 Use Case III: Cryptocurrencies Trading 5 Ecosystem of FinRL and Conclusions, and References 5 Ecosystem of FinRL and Conclusions, and References 4.4 Use Case I: Stock Trading We use FinRL to reproduce both [50] and [51] for stock trading. The ensemble strategy [51] combines three DRL algorithms (PPO [42], A2C [32] and DDPG [26]) to improve the robustness. The implementation is easy with FinRL. We choose three algorithms (PPO, A2C, DDPG) in the agent layer, and an environment with start and end dates in the environment layer. The implementations of DRL algorithms and data preprocessing are transparent to users, alleviating the programming and debugging workloads. Thus, FinRL greatly facilitates the strategy design, allowing users to focus on improving the trading performance. Fig. 5 and Table 3 show the backtesting performance on Dow 30 constituent stocks, accessed at 2020/07/01. The training period is from 2009/01/01 to 2020/06/30 on a daily basis, and the testing period is from 2020/07/01 to 2021/06/30. The performance in terms of multiple metrics is consistent with the results reported in [51] and [50], and here we show results in a recent trading period. We can see from the DJIA index that the trading period is a bullish market with an annual return of 32.84%. The ensemble strategy achieves a Sharpe ratio of 2.81 and an annual return of 52.61%. It beats A2C with a Sharpe ratio of 2.24, PPO with a Sharpe ratio of 2.23, DDPG with a Sharpe ratio of 2.05, DJIA with a Sharpe ratio of 2.02, and min-variance portfolio allocation with a Sharpe ratio of 1.98, respectively. Therefore, the backtesting performance demonstrates that FinRL successfully reproduces the ensemble strategy [51]. This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv

Part of HackerNoon's growing list of open-source research papers, promoting free access to academic material.

The Future of AI in Finance

End-to-End Solutions for Cryptocurrency Trading

REINFORCEMENT . tech

Too Long; Didn't Read

CTA: Download the FREE AI Productivity Shift report

FinRL's Implementation of DRL Algorithms for Stock Trading

About Author

Comments

TOPICS

THIS ARTICLE WAS FEATURED IN

Related Stories

Untitled Story

How an Open Model and a Pile of Data are Changing Time Series Analysis

Fraud Detection to Robo Investing: AI in Finance is Rising

From AI-Powered Trading To Regulation and Compliance: What Does 2024 Look Like for Investment Tech?

The Intersection of AI and Finance Will Create New Careers

Four Use-Cases for AI Automation in 2024

How an Open Model and a Pile of Data are Changing Time Series Analysis

Fraud Detection to Robo Investing: AI in Finance is Rising

From AI-Powered Trading To Regulation and Compliance: What Does 2024 Look Like for Investment Tech?

The Intersection of AI and Finance Will Create New Careers

Four Use-Cases for AI Automation in 2024

Light-Mode

Classic

Newspaper

Dark-Mode

Neon Noir

Minty

HN StartUps