Financial Zen: Achieving Market Harmony through Reinforcement Learning and Algorithmic Trading

Algorithmic trading, utilizing reinforcement learning, has attracted significant attention for its adaptability and potential in present financial markets. This research combines Deep Deterministic Policy Gradients (DDPG), Advantage Actor-Critic (AAC), and Proximal Policy Optimization (PPO) algorithms for a comprehensive trading system. To address memory constraints, a load-on-demand technique manages large datasets. Performance is evaluated across 30 Dow Jones stocks, compared to the Dow Jones Industrial Average (DJIA) and a minimum-variance portfolio. Results show the integrated approach outperforms traditional strategies in annualized return, Sharpe ratio, and maximum drawdown. This study contributes to algo-trading literature by leveraging multiple reinforcement learning algorithms, offering benefits for investors in dynamic markets.