用扩散模型生成极端市场场景，提升强化学习投资组合的抗风险能力。
Diffusion-Augmented Reinforcement Learning for Robust Portfolio Optimization under Stress Scenarios
- 结合扩散模型生成不同强度的市场崩盘数据，增强训练鲁棒性。
- 在2025年关税危机等突发情境下表现优于传统方法。
- 适合关注金融风控与AI策略稳健性的研究者和从业者。
在不断变化且复杂的金融市场中，投资组合优化仍是投资者和资产管理者的重大挑战。传统方法难以捕捉市场行为的复杂动态，也难以匹配多样化投资者偏好。为此，我们提出一种创新框架——扩散增强强化学习（DARL），将去噪扩散概率模型（DDPMs）与深度强化学习（DRL）相结合，用于投资组合管理。通过利用DDPM生成受不同压力强度条件控制的合成市场崩盘情景，显著提升了训练数据的鲁棒性。实证评估表明，DARL在风险调整收益和应对未知危机（如2025年关税危机）方面均优于传统基线方法。该研究为增强DRL驱动金融应用的抗压能力提供了可靠且实用的方法。
原文摘要 · Abstract (English)
In the ever-changing and intricate landscape of financial markets, portfolio optimisation remains a formidable challenge for investors and asset managers. Conventional methods often struggle to capture the complex dynamics of market behaviour and align with diverse investor preferences. To address this, we propose an innovative framework, termed Diffusion-Augmented Reinforcement Learning (DARL), which synergistically integrates Denoising Diffusion Probabilistic Models (DDPMs) with Deep Reinforcement Learning (DRL) for portfolio management. By leveraging DDPMs to generate synthetic market crash scenarios conditioned on varying stress intensities, our approach significantly enhances the robustness of training data. Empirical evaluations demonstrate that DARL outperforms traditional baselines, delivering superior risk-adjusted returns and resilience against unforeseen crises, such as the 2025 Tariff Crisis. This work offers a robust and practical methodology to bolster stress resilience in DRL-driven financial applications.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。