Quantitative Models and Construction Methods 1. Model Name: Volume Timing Signal - **Model Construction Idea**: The model uses volume-based signals to determine market timing, identifying bullish or cautious market views based on volume trends[23][24] - **Model Construction Process**: 1. Analyze the volume trends of major broad-based indices 2. Assign a "bullish" or "cautious" signal based on the volume dynamics 3. For example, as of April 11, 2025, the Beixin 50 index showed a "bullish" signal, while other indices like the Shanghai Composite and CSI 300 showed "cautious" signals[23][24] - **Model Evaluation**: The model provides a straightforward and intuitive approach to market timing but may lack robustness in highly volatile markets[23][24] 2. Model Name: Momentum Sentiment Indicator - **Model Construction Idea**: This model captures market sentiment by analyzing the proportion of stocks with positive returns in the CSI 300 index over a specific period[24][25] - **Model Construction Process**: 1. Calculate the proportion of CSI 300 constituent stocks with positive returns over the past N days $ \text{CSI 300 N-day Upward Proportion} = \frac{\text{Number of stocks with positive returns in N days}}{\text{Total number of stocks in CSI 300}} $ 2. Smooth the indicator using two moving averages with different windows (N1 and N2, where N1 > N2) 3. Generate signals: - If the short-term moving average (fast line) exceeds the long-term moving average (slow line), the market is considered bullish - If the fast line falls below the slow line, the market sentiment is turning cautious[27] - **Model Evaluation**: The indicator is effective in capturing upward opportunities but may fail to avoid risks in declining markets[25][27] 3. Model Name: Moving Average Sentiment Indicator - **Model Construction Idea**: This model uses an eight-moving-average system to assess the trend state of the CSI 300 index[32][33] - **Model Construction Process**: 1. Calculate the closing prices of the CSI 300 index for eight moving averages (parameters: 8, 13, 21, 34, 55, 89, 144, 233) 2. Assign values to the indicator based on the number of moving averages above or below the current price: - If the current price exceeds five or more moving averages, the market is considered bullish - Otherwise, the market is neutral or bearish[32][33] - **Model Evaluation**: The model provides a clear trend-following signal but may lag in rapidly changing markets[35] --- Model Backtesting Results 1. Volume Timing Signal - Beixin 50 Index: Bullish signal[23][24] - Other indices (e.g., Shanghai Composite, CSI 300): Cautious signal[23][24] 2. Momentum Sentiment Indicator - CSI 300 Upward Proportion: Approximately 53% in the most recent week[25] 3. Moving Average Sentiment Indicator - CSI 300 Index: Currently in a non-bullish sentiment zone[35] --- Quantitative Factors and Construction Methods 1. Factor Name: Cross-Sectional Volatility - **Factor Construction Idea**: Measures the dispersion of stock returns within an index to assess the alpha environment[37] - **Factor Construction Process**: 1. Calculate the cross-sectional volatility of constituent stocks in indices like CSI 300, CSI 500, and CSI 1000 2. Compare the recent quarter's average volatility with historical periods to determine the alpha environment[41] - **Factor Evaluation**: Higher cross-sectional volatility indicates a better alpha environment for stock selection[37][41] 2. Factor Name: Time-Series Volatility - **Factor Construction Idea**: Measures the volatility of index returns over time to assess the alpha environment[41][44] - **Factor Construction Process**: 1. Calculate the time-series volatility of indices like CSI 300, CSI 500, and CSI 1000 2. Compare the recent quarter's average volatility with historical periods to evaluate the alpha environment[44] - **Factor Evaluation**: Higher time-series volatility suggests a favorable alpha environment for active strategies[41][44] --- Factor Backtesting Results 1. Cross-Sectional Volatility - CSI 300: 1.90% (recent quarter), 77.80% of the past six months' percentile[41] - CSI 500: 2.16% (recent quarter), 42.06% of the past six months' percentile[41] - CSI 1000: 2.52% (recent quarter), 64.54% of the past six months' percentile[41] 2. Time-Series Volatility - CSI 300: 0.63% (recent quarter), 78.84% of the past six months' percentile[44] - CSI 500: 0.48% (recent quarter), 55.56% of the past six months' percentile[44] - CSI 1000: 0.29% (recent quarter), 70.12% of the past six months' percentile[44]