Dynamic Portfolio Optimization with Deep Reinforcement Learning: Empirical Insights from Indian Capital Markets
The Journal of Financial Data Science · 31.01.2026 · 8 (1) · 115-129
DOI: 10.3905/jfds.2025.1.210 ↗Kurz gefasst
Fachartikel in der Zeitschrift The Journal of Financial Data Science, online erschienen am 25.12.2025. Verfasst von Iqbal K. I. und 1 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref).
Abstract (Originalsprache)
Financial markets have grown increasingly complex, rendering traditional portfolio management strategies insufficient to address modern market dynamics. Deep reinforcement learning (DRL) has emerged as a promising approach, combining the adaptive power of reinforcement learning with deep neural networks to optimize stock portfolio performance. This study investigates the practical applicability of DRL models in real-world financial markets, focusing on their ability to optimize portfolio performance under realistic constraints such as transaction costs, liquidity limitations, and market fluctuations. Utilizing stock market data from NIFTY 50 companies and employing the advantage actor–critic (A2C) algorithm, this research evaluates the performance of DRL-based portfolio management systems against traditional benchmarks such as the NIFTY 50 Index. The findings reveal that the DRL strategy consistently outperforms the benchmark across key performance metrics, including cumulative return, Sharpe ratio, and risk-adjusted returns. Additionally, the study highlights the resilience of DRL strategies during market downturns and their capacity for dynamic decision making. These results underscore the potential of DRL models as reliable tools for portfolio management in live trading scenarios.
Zitieren
APA
I., I. K., & Ramachandran, R. (2026). Dynamic Portfolio Optimization with Deep Reinforcement Learning: Empirical Insights from Indian Capital Markets. The Journal of Financial Data Science, 8(1), 115-129. https://doi.org/10.3905/jfds.2025.1.210
Vancouver
I. IK, Ramachandran R. Dynamic Portfolio Optimization with Deep Reinforcement Learning: Empirical Insights from Indian Capital Markets. The Journal of Financial Data Science. 2026;8(1):115-129. doi:10.3905/jfds.2025.1.210
BibTeX
@article{I2026,
author = {I., Iqbal K. and Ramachandran, Remya},
title = {Dynamic Portfolio Optimization with Deep Reinforcement Learning: Empirical Insights from Indian Capital Markets},
journal = {The Journal of Financial Data Science},
year = {2026},
volume = {8},
number = {1},
pages = {115--129},
doi = {10.3905/jfds.2025.1.210}
}
EUVIMED