Quantitative Factors and Construction 1. Factor Name: Valuation Deviation (EPD) Factor - **Construction Idea**: The EPD factor is constructed by combining the mean-reversion strategy commonly used in the CTA domain with the valuation repair logic based on fundamentals. It leverages the mean-reversion characteristics of the PE valuation metric[7][2] - **Construction Process**: 1. Identify the mean-reversion property of the PE ratio as the core valuation metric 2. Construct the EPD factor to capture valuation deviations based on this mean-reversion logic[7][2] - **Evaluation**: The factor effectively captures valuation deviations and demonstrates strong performance in backtesting[7] 2. Factor Name: Slow Deviation (EPDS) Factor - **Construction Idea**: The EPDS factor is derived from the EPD factor by removing the probability of changes in individual stock valuation logic, which is proxied by the stock's information ratio[7][2] - **Construction Process**: 1. Start with the EPD factor 2. Use the information ratio of individual stocks to filter out those with altered valuation logic 3. Construct the EPDS factor to focus on slow valuation deviations[7][2] - **Evaluation**: The factor further refines the valuation deviation logic and improves performance metrics compared to the EPD factor[7] 3. Factor Name: Valuation Anomaly (EPA) Factor - **Construction Idea**: The EPA factor is constructed by further refining the EPDS factor. It removes the influence of beta, growth, and value styles to isolate the "valuation anomaly" logic[7][2] - **Construction Process**: 1. Start with the EPDS factor 2. Remove the impact of beta, growth, and value styles 3. Construct the EPA factor to focus solely on valuation anomalies[7][2] - **Evaluation**: The EPA factor demonstrates the highest performance among the three factors, with strong backtesting results and robustness[7] --- Factor Backtesting Results 1. Valuation Deviation (EPD) Factor - Annualized Return: 17.52%[2][8][12] - Annualized Volatility: 10.00%[2][8][12] - IR: 1.75[2][8][12] - Monthly Win Rate: 70.65%[2][8][12] - Maximum Drawdown: 8.93%[2][8][12] 2. Slow Deviation (EPDS) Factor - Annualized Return: 16.25%[2][8][12] - Annualized Volatility: 5.72%[2][8][12] - IR: 2.84[2][8][12] - Monthly Win Rate: 78.80%[2][8][12] - Maximum Drawdown: 3.10%[2][8][12] 3. Valuation Anomaly (EPA) Factor - Annualized Return: 17.23%[2][8][12] - Annualized Volatility: 5.11%[2][8][12] - IR: 3.37[2][8][12] - Monthly Win Rate: 80.98%[2][8][12] - Maximum Drawdown: 3.12%[2][8][12] --- Additional Insights - In May 2025, the EPA factor's 5-group long portfolio achieved a return of 3.88%, while the short portfolio returned 3.27%, resulting in a long-short spread of 0.61%[2][15] - The EPA factor's monthly RankIC mean during the backtesting period (2010/01–2022/05) was 0.061, with a RankICIR of 4.75[2][7] - The EPA factor's 5-group long-short portfolio achieved an annualized return of 18.29%, an IR of 3.76, a win rate of 86.99%, and a maximum drawdown of 1.53% during the backtesting period[7]

证券研究报告·金融工程·金工定期报告 金工定期报告 20250507 估值异常因子绩效月报 20250430 2025 年 05 月 07 日 [Table_Tag] [Table_Summary] 报告要点 ◼ 估值偏离 EPD 因子多空对冲绩效（全市场）：2010 年 2 月至 2025 年 4 月，估值偏离 EPD 因子在全体 A 股（剔除北交所股票）中，5 分组多 空对冲的年化收益为 17.65%，年化波动为 10.02%，信息比率为 1.76， 月度胜率为 71.04%，月度最大回撤为 8.93%。 ◼ 缓慢偏离 EPDS 因子多空对冲绩效（全市场）：2010 年 2 月至 2025 年 4 月，缓慢偏离 EPDS 因子在全体 A 股（剔除北交所股票）中，5 分组 多空对冲的年化收益为 16.31%，年化波动为 5.73%，信息比率为 2.85， 月度胜率为 79.23%，月度最大回撤为 3.10%。 ◼ 估值异常 EPA 因子多空对冲绩效（全市场）：2010 年 2 月至 2025 年 4 月，估值异常 EPA 因子在全体 A 股（剔除北交所股票）中，5 分组多空 对冲的年化收益为 17.30%， ...