金工定期报告20251129：“重拾自信2.0”RCP因子绩效月报20251128-20251129

Quantitative Models and Construction Methods 1. Model Name: "Rediscover Confidence 2.0" RCP Factor - Model Construction Idea: The model is based on a common expectation bias in behavioral finance—overconfidence. It innovatively uses high-frequency minute sequence data to construct the overconfidence factor CP by calculating the time gap between favorable price surges and price corrections. The second-generation Rediscover Confidence Factor (RCP) is derived by orthogonalizing the first-generation CP factor with intraday returns and using the residuals as the RCP factor[1][6]. - Model Construction Process: - Step 1: Calculate the time gap between favorable price surges and price corrections to construct the overconfidence factor CP. - Step 2: Orthogonalize the CP factor with intraday returns. - Step 3: Use the residuals from the orthogonalization process as the second-generation Rediscover Confidence Factor (RCP)[6]. - Model Evaluation: The RCP factor constructed based on the Rediscover Confidence idea performs significantly better than traditional combination methods[6]. Model Backtesting Results 1. "Rediscover Confidence 2.0" RCP Factor: - Annualized Return: 17.68%[1][7][12] - Annualized Volatility: 7.83%[1][7][12] - Information Ratio (IR): 2.26[1][7][12] - Monthly Win Rate: 77.46%[1][7][12] - Maximum Drawdown: 7.46%[1][7][12] Quantitative Factors and Construction Methods 1. Factor Name: Overconfidence CP Factor - Factor Construction Idea: The factor is based on the degree of investor overconfidence affecting stock prices, using the time difference between rapid price increases and decreases as a proxy variable[6]. - Factor Construction Process: - Step 1: Calculate the time difference between rapid price increases and decreases to construct the overconfidence factor CP[6]. - Factor Evaluation: The CP factor innovatively captures the overconfidence bias in investor behavior[6]. 2. Factor Name: Rediscover Confidence RCP Factor - Factor Construction Idea: Considering that investors may become overly pessimistic during price corrections, leading to excessive corrections, but due to favorable news, such stocks will eventually rebound. The RCP factor is derived by orthogonalizing the CP factor with intraday returns and using the residuals[6]. - Factor Construction Process: - Step 1: Orthogonalize the CP factor with intraday returns. - Step 2: Use the residuals from the orthogonalization process as the Rediscover Confidence Factor (RCP)[6]. - Factor Evaluation: The RCP factor, after purification, shows significantly improved performance[7]. Factor Backtesting Results 1. Rediscover Confidence RCP Factor: - IC Mean: 0.04[1] - Annualized ICIR: 3.27[1] - Annualized Return: 20.69%[1] - Information Ratio (IR): 2.91[1] - Monthly Win Rate: 81.55%[1]