金融工程定期：开源交易行为因子绩效月报（2025年12月）-20251231

Quantitative Models and Construction Methods Barra Style Factors - Model Name: Barra Style Factors - Construction Idea: The model tracks the performance of common style factors such as size, value, growth, and profitability in the market - Construction Process: The factors are calculated based on specific financial metrics. For example: - Size factor is based on market capitalization - Book-to-market ratio is used for the value factor - Growth factor is derived from growth-related metrics - Profitability factor is based on earnings expectations[3][13] - Evaluation: The model provides insights into the performance of different market styles, helping investors understand factor contributions to returns[3][13] Kaiyuan Behavioral Factors - Model Name: Kaiyuan Behavioral Factors - Construction Idea: These factors are based on trading behaviors, aiming to capture alpha signals from microstructure patterns in the market - Construction Process: - Ideal Reversal Factor: Measures the reversal strength of trading days by analyzing the average transaction size. It identifies days with the strongest reversal attributes[4][13] - Smart Money Factor: Tracks institutional trading activity using minute-level price and volume data. It identifies "smart money" trades by sorting minute data based on a constructed indicator and calculating the volume-weighted average price (VWAP) of these trades[4][40][42] - APM Factor: Measures the difference in trading behavior between morning and afternoon sessions. It uses regression analysis on overnight and afternoon returns to calculate residuals, which are then used to construct the factor[4][41][43][44] - Ideal Amplitude Factor: Captures the structural differences in stock price amplitude under high and low price states. It calculates the difference between the average amplitude of high-price and low-price days[4][46] - Evaluation: These factors are recognized for their ability to capture unique trading behavior patterns and provide alpha signals[4][13] Kaiyuan Behavioral Composite Factor - Model Name: Kaiyuan Behavioral Composite Factor - Construction Idea: Combines the above behavioral factors into a single composite factor to enhance performance - Construction Process: - Standardizes and winsorizes individual factors within industries - Uses the past 12 periods' ICIR values as weights to calculate the composite factor - Applies industry and market capitalization neutrality adjustments[30][34] - Evaluation: The composite factor demonstrates superior performance in small and mid-cap stock pools compared to large-cap pools[30][35] --- Backtesting Results of Models Barra Style Factors - Size Factor: Return of 1.06% in December 2025[3][13] - Book-to-Market Ratio Factor: Return of -0.18% in December 2025[3][13] - Growth Factor: Return of 0.20% in December 2025[3][13] - Profitability Factor: Return of 0.94% in December 2025[3][13] Kaiyuan Behavioral Factors - Ideal Reversal Factor: - IC: -0.049 - RankIC: -0.060 - IR: 2.42 - Long-Short Monthly Win Rate: 77.8% - December 2025 Long-Short Return: 0.14% - 12-Month Long-Short Monthly Win Rate: 58.3%[5][14][17] - Smart Money Factor: - IC: -0.037 - RankIC: -0.061 - IR: 2.69 - Long-Short Monthly Win Rate: 80.1% - December 2025 Long-Short Return: -0.24% - 12-Month Long-Short Monthly Win Rate: 75.0%[5][19][22] - APM Factor: - IC: 0.028 - RankIC: 0.034 - IR: 2.25 - Long-Short Monthly Win Rate: 76.2% - December 2025 Long-Short Return: 1.08% - 12-Month Long-Short Monthly Win Rate: 41.7%[5][23][26] - Ideal Amplitude Factor: - IC: -0.053 - RankIC: -0.073 - IR: 2.99 - Long-Short Monthly Win Rate: 83.0% - December 2025 Long-Short Return: -0.63% - 12-Month Long-Short Monthly Win Rate: 58.3%[5][26][29] Kaiyuan Behavioral Composite Factor - Composite Factor: - IC: 0.066 - RankIC: 0.093 - IR: 3.25 - Long-Short Monthly Win Rate: 78.8% - December 2025 Long-Short Return: -0.04% - 12-Month Long-Short Monthly Win Rate: 58.3% - Annualized Return of Long-Short Portfolio: 8.09% - Sharpe Ratio: 2.56 - Monthly Win Rate: 77.4% - IR in Small and Mid-Cap Pools: 2.83 (China Securities 2000), 2.62 (China Securities 1000), 1.00 (China Securities 800)[5][30][34][35]