Deep Reinforcement Learning Optimizes Portfolio Risk-Return

Researchers have developed a novel deep reinforcement learning framework, MORP-DRL, designed to optimize investment portfolios by considering both expected return and downside risk. This framework integrates variance, Conditional Value-at-Risk (CVaR), and Entropic Value-at-Risk (EVaR) to model complex market dynamics, including heavy-tailed behavior and transaction costs. Experiments across various market regimes indicate that MORP-DRL offers competitive risk-return performance and enhanced stability during stressful market conditions, demonstrating scalability for high-dimensional portfolios. AI

IMPACT This framework could enhance financial modeling by providing more robust risk management and return optimization strategies.

RANK_REASON The item is an academic paper detailing a new methodology for portfolio optimization using deep reinforcement learning. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.AI →

• CVaR
• deep reinforcement learning
• GARCH(1,1) Model of the Financial Market with the Minkowski Metric
• MORP-DRL
• Proximal Policy Optimization

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →