Deep Deterministic Portfolio Optimization
This work addresses portfolio optimization for traders, but it is incremental as it applies an existing method to known environments.
The authors tackled the problem of using deep reinforcement learning for optimal trading strategies by testing the deep deterministic policy gradient algorithm on simple but non-trivial trading environments, achieving close-to-optimal rewards.
Can deep reinforcement learning algorithms be exploited as solvers for optimal trading strategies? The aim of this work is to test reinforcement learning algorithms on conceptually simple, but mathematically non-trivial, trading environments. The environments are chosen such that an optimal or close-to-optimal trading strategy is known. We study the deep deterministic policy gradient algorithm and show that such a reinforcement learning agent can successfully recover the essential features of the optimal trading strategies and achieve close-to-optimal rewards.