Portfolio Management of SET50 Stocks Using Deep Reinforcement Learning Methods

In this article, we propose a reinforcement learning method for developing a stock trading strategy while optimizing investment return. Using the Advantage Actor Critic (A2C) algorithms to check all SET 50 stocks, to train a deep reinforcement learning agent to get this trading technique. Our empirical findings demonstrate that, in terms of Sharpe ratio and cumulative returns, the proposed reinforcement learning technique beats both the SET 50 Average index and the conventional min-variance portfolio allocation strategy.