Paper: arXiv 2402.01441

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

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.

Complexity vs Empirical Score

  • Math Complexity: 6.5/10
  • Empirical Rigor: 4.0/10
  • Quadrant: Lab Rats — high math complexity, limited empirical testing

Why this score: The paper employs deep reinforcement learning algorithms like DDPG and A2C, which involve advanced mathematical concepts and neural network architectures, but lacks visible code, detailed datasets, or extensive statistical metrics in the provided excerpt, indicating a focus on theoretical modeling over implementation-heavy validation.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Integrate Sentiment & DRL for Profitable/Robust Trading"] --> B["Data Acquisition & Preprocessing<br>Historical Stock Data + Financial News Text"]
  B --> C["Methodology - Key Components<br>Sentiment Analysis Module<br>DRL Trading Agents"]
  C --> D["Computational Process:<br>Dynamic Ensemble Framework<br>- Extract Real-Time Sentiment<br>- Select Best Agent Based on Sentiment"]
  D --> E["Backtesting & Evaluation<br>vs. Fixed-Ensemble, Single Agents, & Market Metrics"]
  E --> F["Key Outcomes & Findings<br>1. Higher Profitability & Lower Risk<br>2. Dynamic Selection beats Fixed Scheduling<br>3. Framework is Simplicity-Efficient for Live Trading"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes integrating sentiment analysis with deep reinforcement learning ensemble algorithms for stock trading, creating a dynamic, sentiment‑based agent switching strategy.

Source sentence
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.

Assumes 3

  • The trading environment assumes liquidity, non‑negative balance, and transaction costs. execution
    Source sentence
    In this environment, an agent will aim to maximize their expected cumulative change in portfolio value following several market constraints and assumptions (liquidity, non-negative balance, and transaction costs).
  • Agents are switched when the sentiment score exceeds a predetermined absolute threshold. model
    Source sentence
    If the sentiment score for a given period exceeds a predetermined absolute threshold, then we switch to using the agent which currently holds the highest validation score for the given period.
  • Sentiment is captured using daily headlines from leading financial news sources and the AFINN‑en‑165 lexicon. data
    Source sentence
    To capture market sentiment, we write a script that gathers daily headlines from leading financial news sources (Wall Street Journal, Bloomberg, etc.) and extracts their sentiments using the AFINN‑en‑165 sentiment lexicon.

Data

  • Historical stock price data market-prices public
    2010-01-01 to 2016-12-31 (training) and
    Source sentence
    All deep‑learned agents were trained on stock data from 1/1/2010 to 12/31/2016 and then evaluated from 01/01/2017 to 01/01/2019.
  • Wall Street Journal Archive articles · Wall Street Journal text-news public
    Source sentence
    For sentiment analysis, we used business articles from the Wall Street Journal Archive, with 15 articles gathered daily to generate an approximation of daily financial sentiment.

Recognised sources: Bloomberg

Costs

Unclear - sources disagree The platform classifier and the paper-text extraction give different answers, so no value is asserted.

Platform classifier: Not modeled

Paper text extraction: Modeled - Transaction costs are included as a market constraint in the environment definition.

Source sentence (the authors' words)
In this environment, an agent will aim to maximize their expected cumulative change in portfolio value following several market constraints and assumptions (liquidity, non-negative balance, and transaction costs).

Finds 3

  • Sentiment‑Ensemble achieves a cumulative return of over 40% over the two‑year test period, an annualized return of 18.36%, and the highest Sharpe (1.32) and Sortino (1.87) ratios among all baselines. performance
    Source sentence
    Our Sentiment‑Ensemble strategy significantly outperforms both the Dow Jones and the ensemble strategy, resulting in a cumulative return of over 40% over the two‑year period and an annualized return of 18.36%
  • Dynamic sentiment‑based agent switching provides the primary performance gain over the conventional fixed‑interval ensemble switching. performance
    Source sentence
    demonstrate the contribution of each component, and find that the majority of our improvement in trading ability can be attributed to dynamic agent switching as expected.
  • A single DDPG agent underperforms all ensemble‑based methods and even the Dow Jones benchmark. performance
    Source sentence
    We also note that an individual deep reinforcement strategy (DDPG) falls behind all methods, including the DJIA.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 29 related papers). Coverage is partial, so this is neither replication nor refutation.

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.

Cited References (27)

  1. FinRL : A deep reinforcement learning library for automated stock trading in quantitative finance (2020)
  2. Deep Reinforcement Learning for Automated Stock Trading: An Ensemble Strategy (2020)
  3. Continuous control with deep reinforcement learning (2019) — link
  4. Dynamic stock-decision ensemble strategy based on deep reinforcement learning (2023)
  5. Deterministic policy gradient algorithms (2014)
  6. Proximal Policy Optimization Algorithms (2017) — link
  7. Trust Region Policy Optimization (2017) — link
  8. Playing Atari with Deep Reinforcement Learning (2013) — link
and 19 more
  1. Asynchronous Methods for Deep Reinforcement Learning (2016) — link
  2. A new ANEW : evaluation of a word list for sentiment analysis in microblogs (2011)
  3. Top of Mind - U.S. Outperformance: At a Turning Point? (2023)
  4. Practical Deep Reinforcement Learning Approach for Stock Trading (2022) — link
  5. Evolutionary Multiobjective Optimization Approach for Evolving Ensemble of Intelligent Paradigms for Stock Market Modeling (2005)
  6. Stock Prediction Using Event-based Sentiment Analysis (2013)
  7. A new ANEW: Evaluation of a word list for sentiment analysis in microblogs (2011)
  8. A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem (2017) — link
  9. Mercury: A Deep Reinforcement Learning-Based Investment Portfolio Strategy for Risk-Return Balance (2023) — link
  10. A Novel Ensemble Learning Approach for Stock Market Prediction Based on Sentiment Analysis and the Sliding Window Method (2023)
  11. Big Data: Deep Learning for fnancial sentiment analysis (2018)
  12. Stock trend prediction user sentiment analysis (2023)
  13. Ensemble Investment Strategies Based on Reinforcement Learning (2022)
  14. Stock Prediction Using Twitter Sentiment Analysis (2012)
  15. Stock trend prediction using news sentiment analysis (2016) — link
  16. Deep Reinforcement Learning Approach for Trading Automation in The Stock Market (2022)
  17. Addressing Function Approximation Error in Actor-Critic Methods (2018) — link
  18. VADER: A Parsimonious Rule-Based Model for Sentiment Analysis of Social Media Text (2014) — link
  19. Loughran-McDonald Master Dictionary w/ Sentiment Word Lists

0 of 27 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.