Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Learning the Market: Sentiment-Based Ensemble Trading Agents

Andrew Ye, James Xu, Vidyut Veedgav, Yi Wang, Yifan Yu, Daniel Yan, Ryan Chen, Vipin Chaudhary, Shuai Xu

TL;DR

This work proposes and study the integration of sentiment analysis and deep reinforcement learning ensemble algorithms for stock trading by evaluating strategies capable of dynamically altering their active agent given the concurrent market environment, and designs a simple-yet-effective method for extracting financial sentiment.

Abstract

We propose and study the integration of sentiment analysis and deep reinforcement learning ensemble algorithms for stock trading by evaluating strategies capable of dynamically altering their active agent given the concurrent market environment. In particular, we design a simple-yet-effective method for extracting financial sentiment and combine this with improvements on existing trading agents, resulting in a strategy that effectively considers both qualitative market factors and quantitative stock data. We show that our approach results in a strategy that is profitable, robust, and risk-minimal - outperforming the traditional ensemble strategy as well as single agent algorithms and market metrics. Our findings suggest that the conventional practice of switching and reevaluating agents in ensemble every fixed-number of months is sub-optimal, and that a dynamic sentiment-based framework greatly unlocks additional performance. Furthermore, as we have designed our algorithm with simplicity and efficiency in mind, we hypothesize that the transition of our method from historical evaluation towards real-time trading with live data to be relatively simple.

Learning the Market: Sentiment-Based Ensemble Trading Agents

TL;DR

This work proposes and study the integration of sentiment analysis and deep reinforcement learning ensemble algorithms for stock trading by evaluating strategies capable of dynamically altering their active agent given the concurrent market environment, and designs a simple-yet-effective method for extracting financial sentiment.

Abstract

We propose and study the integration of sentiment analysis and deep reinforcement learning ensemble algorithms for stock trading by evaluating strategies capable of dynamically altering their active agent given the concurrent market environment. In particular, we design a simple-yet-effective method for extracting financial sentiment and combine this with improvements on existing trading agents, resulting in a strategy that effectively considers both qualitative market factors and quantitative stock data. We show that our approach results in a strategy that is profitable, robust, and risk-minimal - outperforming the traditional ensemble strategy as well as single agent algorithms and market metrics. Our findings suggest that the conventional practice of switching and reevaluating agents in ensemble every fixed-number of months is sub-optimal, and that a dynamic sentiment-based framework greatly unlocks additional performance. Furthermore, as we have designed our algorithm with simplicity and efficiency in mind, we hypothesize that the transition of our method from historical evaluation towards real-time trading with live data to be relatively simple.
Paper Structure (20 sections, 4 figures, 5 tables)

This paper contains 20 sections, 4 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Overview
  3. The Market as a Deep Learning Environment
  4. Language-Based Trading Agents
  5. Our Motivation
  6. Methodology
  7. Deep Reinforcement Learning
  8. Deep Deterministic Policy Gradient
  9. Proximal Policy Optimization
  10. Advantage Actor-Critic
  11. Capturing Market Sentiment
  12. Dynamic Learning Adaptation
  13. Choosing an Agent
  14. Switching Agents
  15. Ablation studies
  16. ...and 5 more sections

Figures (4)

  • Figure 1: Stock trading as a reinforcement learning problem.
  • Figure 2: Overview of proposed algorithm.
  • Figure 3: Performance of our algorithm (blue) against the ensemble strategy (green), a DDPG agent (red) and the Dow Jones Industrial Average (purple).
  • Figure 4: Return over time of Sentiment-Ensemble (blue) and ensemble (green) against A2C (dashed-red), DDPG (dashed-purple), PPO (dashed-yellow), and TD3 (dashed-blue).