行业轮动组合月报：量价行业轮动组合2025年前4个月皆跑赢基准-20250503

Quantitative Models and Construction Methods 1. Model Name: Volume-Price Industry Rotation Strategy - Model Construction Idea: The strategy is based on six dimensions of volume-price factors, including momentum, trading volatility, turnover rate, long-short comparison, volume-price divergence, and volume-amplitude alignment. These factors are tested on a single-factor basis at the monthly frequency for the CSI Level-1 industries, resulting in 11 effective and logically strong industry factors[6] - Model Construction Process: 1. Construct 11 volume-price factors based on the six dimensions mentioned above 2. At the end of each month, select the top five industries with the highest composite factor scores from the CSI Level-1 industries (excluding "Comprehensive" and "Comprehensive Finance") 3. Apply equal weighting within factors and equal weighting across industries to form the final strategy[7] - Model Evaluation: The model demonstrates strong logical consistency and effectiveness in identifying outperforming industries[6] --- Quantitative Factors and Construction Methods 1. Factor Name: Second-Order Momentum - Factor Construction Idea: Measures the exponential weighted moving average (EWMA) of the closing price relative to its historical mean[7] - Factor Construction Process: $ \text{Second-Order Momentum} = \text{Close}t \cdot \text{EWMA}(\text{Close}{t-\text{window1}:t}) - \text{mean}(\text{Close}{t-\text{window1}:t}) $ - Parameters: "Close" represents the closing price, "window1" defines the lookback period[7] 2. Factor Name: Momentum Term Spread - Factor Construction Idea: Captures the difference in momentum over two different time windows[7] - Factor Construction Process: $ \text{Momentum Term Spread} = \frac{\text{Close}t - \text{Close}{t-\text{window1}}}{\text{Close}{t-\text{window1}}} - \frac{\text{Close}t - \text{Close}{t-\text{window2}}}{\text{Close}{t-\text{window2}}} $ - Parameters: "window1" and "window2" represent two different lookback periods[7] 3. Factor Name: Trading Amount Volatility - Factor Construction Idea: Measures the standard deviation of trading amounts over a specific window[7] - Factor Construction Process: $ \text{Trading Amount Volatility} = -\text{STD}(\text{Amount}) $ - Parameters: "Amount" refers to the trading amount, and "STD" is the standard deviation operator[7] 4. Factor Name: Volume-Price Divergence Covariance - Factor Construction Idea: Measures the covariance between ranked closing prices and ranked volumes over a specific window[7] - Factor Construction Process: $ \text{Volume-Price Divergence Covariance} = \text{rank}(\text{covariance}[\text{rank}(\text{Close}), \text{rank}(\text{Volume}), \text{window}]) $ - Parameters: "Close" represents the closing price, "Volume" represents the trading volume, and "window" defines the lookback period[7] 5. Factor Name: Volume-Amplitude Alignment - Factor Construction Idea: Measures the correlation between ranked volumes and ranked price ranges over a specific window[7] - Factor Construction Process: $ \text{Volume-Amplitude Alignment} = \text{correlation}[\text{rank}(\text{Volume}{i-1}), \text{rank}(\text{High}_i - \text{Low}_i), \text{window}] $ - Parameters: "High" and "Low" represent the highest and lowest prices, respectively, and "window" defines the lookback period[7] --- Backtesting Results of the Model 1. Volume-Price Industry Rotation Strategy - Cumulative Return (2010-2025): 694.50%[9] - Cumulative Excess Return over Equal-Weighted Industry Portfolio: 605.20%[9] - April 2025 Monthly Return: -1.59%[9] - April 2025 Excess Return over Equal-Weighted Industry Portfolio: 0.81%[9]