An integrated momentum strategy based on entropy and behavioral overreaction: Evidence from Vietnam

  • 10 Views
  • 2 Downloads

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License

Type of the article: Research Article

Abstract
The increasing behavioral volatility and informational complexity of emerging stock markets such as Vietnam create a critical need for more advanced analytical approaches to identify reliable momentum signals. This study aims to develop and validate an integrated momentum-based trading strategy specifically designed for the Vietnamese stock market. Using price and trading volume data for all stocks listed on the VNINDEX from January 2015 to February 2025, the methodology combines permutation-based entropy measures to capture short-term structural patterns with a formation–holding period framework to analyze medium- and long-term dynamics through Continuing Overreaction. The empirical results reveal a pronounced structural divergence in momentum behavior across investment horizons. Short-term momentum is persistent and strongly associated with low-complexity price and volume patterns, indicating coordinated behavioral trading and temporary predictability. In contrast, medium- and long-term Continuing Overreaction effects exhibit consistently negative values across various formation and holding horizons, suggesting that excess trading intensity leads to systematic mean reversion rather than sustained momentum. Backtesting over the period from January 2023 to February 2025 demonstrates that the proposed integrated strategy substantially outperforms a passive VNINDEX buy-and-hold benchmark, achieving a Sharpe ratio of 3.96 compared to 0.64 for the market. The superior performance remains robust across alternative portfolio construction settings and reflects improved downside risk control rather than increased return volatility. These findings indicate that integrating entropy-based complexity measures with volume-driven behavioral indicators provides a more effective framework for enhancing risk-adjusted returns in emerging stock markets.

view full abstract hide full abstract
    • Figure 1. Permutation Entropy of price and volume measures for VNINDEX stocks (m = 2-5, 2015–2025)
    • Figure 2. Permutation Transition Entropy of price and volume measures for VNINDEX stocks (m = 2-5, 2015–2025)
    • Figure 3. Most frequent permutation transitions of daily closing prices for VNINDEX stocks (m = 2-5, 2015–2025)
    • Figure 4. Most frequent permutation transitions of log returns for VNINDEX stocks (m = 2-5, 2015–2025)
    • Figure 5. Most frequent permutation transitions of trading volume for VNINDEX stocks (m = 2-5, 2015–2025)
    • Figure 6. Most frequent permutation transitions of log-volume changes for VNINDEX stocks (m = 2-5, 2015–2025)
    • Figure 7. Continuing Overreaction (CO) values of Winner-minus-Loser portfolios across J/K horizons for VNINDEX stocks (2015–2025)
    • Figure 8. Performance comparison between the VNINDEX benchmark and the integrated strategy (Jan 2023 – Feb 2025)
    • Table 1. Permutation Entropy and Permutation Transition Entropy for VNINDEX at different embedding dimensions (m = 2-5)
    • Table 2. Continuing Overreaction results for J/K-month strategies (Dec 2015–Feb 2025)
    • Table 3. VNINDEX benchmark performance (Jan 2023 – Feb 2025)
    • Table A1. Annual summary of VNINDEX data (2015–2025)
    • Table A2. Backtesting results for integrated Permutation Entropy – Permutation Transition Entropy – Continuing Overreaction model (m = 2)
    • Table A3. Backtesting results for the integrated model (m = 3)
    • Table A4. Backtesting results for the integrated model (m = 4)
    • Table A5. Backtesting Results for the Integrated Model (m = 5)
    • Table A6. Descriptive statistics for VNINDEX benchmark data (Jan 2023–Feb 2025)
    • Conceptualization
      Loan Thi Vu, Minh Phuong Nguyen, Quoc Anh Hoang
    • Formal Analysis
      Loan Thi Vu, Minh Phuong Nguyen
    • Investigation
      Loan Thi Vu, Quoc Anh Hoang
    • Methodology
      Loan Thi Vu, Minh Phuong Nguyen, Quoc Anh Hoang
    • Supervision
      Loan Thi Vu
    • Validation
      Loan Thi Vu
    • Writing – review & editing
      Loan Thi Vu
    • Data curation
      Minh Phuong Nguyen
    • Software
      Minh Phuong Nguyen
    • Writing – original draft
      Minh Phuong Nguyen, Quoc Anh Hoang
    • Visualization
      Quoc Anh Hoang