金工定期报告20250507：信息分布均匀度UID选股因子绩效月报-20250507

Quantitative Factors and Construction Methods 1. Factor Name: Information Distribution Uniformity (UID) Factor - Factor Construction Idea: The UID factor is constructed based on high-frequency volatility data, aiming to improve the traditional volatility factors by capturing the uniformity of information distribution across stocks. It significantly outperforms traditional volatility factors in stock selection[6][7]. - Factor Construction Process: 1. Use minute-level stock data to calculate daily high-frequency volatility. 2. Construct the UID factor by measuring the uniformity of information distribution, leveraging the minute-level price fluctuation data. 3. Exclude the interference of common market styles and industry factors to ensure the purity of the UID factor[6][10]. - Factor Evaluation: The UID factor demonstrates strong stock selection ability, even after removing the influence of market and industry factors. It provides a significant improvement over traditional volatility factors[6][7]. --- Factor Backtesting Results 1. Information Distribution Uniformity (UID) Factor - Annualized Return: 26.99%[7][10] - Annualized Volatility: 9.98%[7][10] - Information Ratio (IR): 2.70[7][10] - Monthly Win Rate: 79.26%[7][10] - Maximum Drawdown: 6.05%[7][10] - April 2025 Performance: - Long Portfolio Return: -0.07% - Short Portfolio Return: -3.78% - Long-Short Portfolio Return: 3.72%[10]