End-to-end data collection

This research data is utilized in the article 'An End-to-End Direct Reinforcement Learning Approach for Multi-factor Based Portfolio Management.' It consists of stock market data from China and the United States, and includes macroeconomic data, financial indicators of the index, and trading data of the index. The article integrates multi-factor models with direct reinforcement learning to formulate an end-to-end portfolio management algorithm, which diverges from traditional prediction-then-optimize approaches. We employ this algorithm for empirical testing in both the Chinese and U.S. stock markets. The dataset encompasses the original data of asset prices and selected factors selected in the article.

本研究数据集应用于论文《面向多因子组合管理的端到端直接强化学习（End-to-End Direct Reinforcement Learning）方法》（An End-to-End Direct Reinforcement Learning Approach for Multi-factor Based Portfolio Management）。该数据集涵盖中美两国股票市场相关数据，包含宏观经济数据、指数金融指标以及指数交易数据。本论文将多因子模型（Multi-factor Model）与直接强化学习（Direct Reinforcement Learning）相结合，构建了端到端的组合管理算法，该算法区别于传统的“先预测后优化”研究范式。我们将此算法应用于中美股票市场开展实证检验。本数据集包含论文中选用的资产价格原始数据与筛选后的因子数据。