Enhancing Profit in TAIEX Futures Trading with a PPO-Based Framework

The financial markets, particularly Taiwan futures trading, present a challenging environment in which retail investors often struggle to achieve consistent profit. These markets offer opportunities for growth and success with the right strategies and approach. Traditional trading tactics rely on technical indicators like average movements and random oscillators, frequently fail to deliver sustainable returns. To address these limitations, we propose a deep reinforcement learning (DRL)-based trading agent that autonomously learns optimal trading strategies without relying on predefined rules. The primary objective is to evaluate the effectiveness of DRL algorithms in making informed trading decisions and achieving superior financial performance relative to a passive buy-and-hold strategy. Our model leverages the Proximal Policy Optimization (PPO) algorithm to dynamically adapt to evolving market conditions with the goal of maximizing long-term profits. Experimental results on TAIEX futures demonstrate that the PPO-based agent significantly outperforms the conventional buy-and-hold approach, highlighting its potential to enhance decision-making and profitability in futures trading.