Quantitative Models and Construction Methods Model 1: Volume Timing Signal - **Model Name**: Volume Timing Signal - **Model Construction Idea**: The model uses volume data to generate timing signals for broad-based indices. - **Model Construction Process**: The model evaluates the volume timing signals for various broad-based indices and assigns a cautious view if the signals indicate a potential downturn. - **Model Evaluation**: The model provides a cautious outlook for all major broad-based indices as of April 3, 2025[23][24]. Model 2: Momentum Sentiment Indicator - **Model Name**: Momentum Sentiment Indicator - **Model Construction Idea**: The model uses the proportion of stocks with positive returns within an index to gauge market sentiment. - **Model Construction Process**: - Calculate the proportion of stocks in the CSI 300 index with positive returns over a specified period. - Apply smoothing with two different window periods to capture the trend. - Generate signals based on the relationship between the short-term and long-term smoothed lines. - Formula: $$ \text{Proportion of Rising Stocks in CSI 300} = \frac{\text{Number of Stocks with Positive Returns in N Days}}{\text{Total Number of Stocks in CSI 300}} $$ - When the short-term smoothed line is above the long-term smoothed line, it indicates a bullish market sentiment[24][25][27]. - **Model Evaluation**: The model can quickly capture upward opportunities but may miss out on gains during prolonged market exuberance. It also has limitations in predicting downturns[25]. Model 3: Moving Average Sentiment Indicator - **Model Name**: Moving Average Sentiment Indicator - **Model Construction Idea**: The model uses the eight moving average system to determine the trend state of the CSI 300 index. - **Model Construction Process**: - Calculate the eight moving averages for the CSI 300 index with parameters 8, 13, 21, 34, 55, 89, 144, and 233. - Assign values based on the number of moving averages the current price exceeds. - Generate signals based on the number of moving averages the current price exceeds. - Formula: $$ \text{Number of Moving Averages Exceeded} = \sum_{i=1}^{8} \mathbb{1}(\text{Price} > \text{MA}_i) $$ - If the number exceeds 5, it indicates a bullish sentiment[32][33]. - **Model Evaluation**: The model indicates that the CSI 300 index is currently in a non-optimistic sentiment zone[35]. Model Backtesting Results 1. **Volume Timing Signal**: All broad-based indices have a cautious view as of April 3, 2025[24]. 2. **Momentum Sentiment Indicator**: The proportion of rising stocks in the CSI 300 index is around 61%[25]. 3. **Moving Average Sentiment Indicator**: The CSI 300 index is in a non-optimistic sentiment zone[35]. Quantitative Factors and Construction Methods Factor 1: Cross-sectional Volatility - **Factor Name**: Cross-sectional Volatility - **Factor Construction Idea**: The factor measures the dispersion of returns among index constituents to gauge the alpha environment. - **Factor Construction Process**: - Calculate the cross-sectional volatility of the constituents of the CSI 300, CSI 500, and CSI 1000 indices. - Compare the recent values with historical averages to determine the alpha environment. - Formula: $$ \text{Cross-sectional Volatility} = \sqrt{\frac{1}{N-1} \sum_{i=1}^{N} (R_i - \bar{R})^2} $$ - Where \( R_i \) is the return of stock \( i \) and \( \bar{R} \) is the average return[41]. - **Factor Evaluation**: The recent cross-sectional volatility indicates a better short-term alpha environment for the CSI 300 and CSI 500 indices, while the CSI 1000 index shows a deteriorating alpha environment[40]. Factor 2: Time-series Volatility - **Factor Name**: Time-series Volatility - **Factor Construction Idea**: The factor measures the volatility of index constituents over time to gauge the alpha environment. - **Factor Construction Process**: - Calculate the time-series volatility of the constituents of the CSI 300, CSI 500, and CSI 1000 indices. - Compare the recent values with historical averages to determine the alpha environment. - Formula: $$ \text{Time-series Volatility} = \sqrt{\frac{1}{T-1} \sum_{t=1}^{T} (R_t - \bar{R})^2} $$ - Where \( R_t \) is the return at time \( t \) and \( \bar{R} \) is the average return[43]. - **Factor Evaluation**: The recent time-series volatility indicates a better short-term alpha environment for the CSI 300 index, while the CSI 500 and CSI 1000 indices show a mixed alpha environment[42]. Factor Backtesting Results 1. **Cross-sectional Volatility**: - CSI 300: Recent quarterly average 1.83%, 51.24% of the two-year range[41]. - CSI 500: Recent quarterly average 2.10%, 47.62% of the two-year range[41]. - CSI 1000: Recent quarterly average 2.48%, 51.79% of the two-year range[41]. 2. **Time-series Volatility**: - CSI 300: Recent quarterly average 0.57%, 49.38% of the two-year range[43]. - CSI 500: Recent quarterly average 0.43%, 40.48% of the two-year range[43]. - CSI 1000: Recent quarterly average 0.26%, 50.20% of the two-year range[43].