Quantitative Models and Construction Methods 1. Model Name: Hotspot Trend ETF Strategy - **Model Construction Idea**: This strategy identifies ETFs with strong upward trends in both their highest and lowest prices, further refining the selection based on the steepness of their 20-day regression coefficients. It aims to capture short-term market attention and construct a risk-parity portfolio[32] - **Model Construction Process**: 1. Select ETFs where both the highest and lowest prices exhibit an upward trend 2. Calculate the regression coefficients of the highest and lowest prices over the past 20 days 3. Construct a support-resistance factor based on the steepness of these coefficients 4. Choose the top 10 ETFs from the factor's long group with the highest 5-day turnover rate/20-day turnover rate ratio 5. Construct a risk-parity portfolio using these ETFs[32] - **Model Evaluation**: The strategy has demonstrated significant excess returns over the benchmark index, indicating its effectiveness in capturing market trends[32] 2. Model Name: All-Weather Strategy - **Model Construction Idea**: This strategy aims to achieve stable returns by avoiding reliance on predictions, leverage, or macroeconomic assumptions. It uses diversified asset allocation, risk adjustment, and structural hedging to smooth volatility[44] - **Model Construction Process**: 1. **Asset Selection**: Diversify across equities, bonds, and commodities 2. **Risk Adjustment**: Balance risk exposure across asset classes 3. **Structural Hedging**: Implement multi-layered hedging to mitigate risks 4. Divide the portfolio into two versions: High-Volatility (High-Vol) and Low-Volatility (Low-Vol) 5. High-Vol employs a four-layer structure focusing on equity, bond, and gold risk parity, while Low-Vol uses a five-layer structure emphasizing risk budgeting[44][53] - **Model Evaluation**: The strategy has shown consistent performance with low drawdowns and high Sharpe ratios, particularly in the Low-Vol version[53] --- Model Backtesting Results 1. Hotspot Trend ETF Strategy - **2025 Performance**: Total return of 58.34%, with an excess return of 38.80% over the CSI 300 Index[32] 2. All-Weather Strategy - **High-Vol Version**: - Annualized Return (2025): 11.8% - Maximum Drawdown: 3.6% - Sharpe Ratio: 1.9 - 2026 YTD Return: 1.9%[53] - **Low-Vol Version**: - Annualized Return (2025): 6.7% - Maximum Drawdown: 2.0% - Sharpe Ratio: 2.4 - 2026 YTD Return: 1.1%[53] --- Quantitative Factors and Construction Methods 1. Factor Name: Return Std 1M - **Factor Construction Idea**: Measures the standard deviation of returns over the past month to capture volatility trends[61] - **Factor Construction Process**: 1. Calculate daily returns for the past 1 month 2. Compute the standard deviation of these returns 3. Normalize the factor by market capitalization and industry[61] - **Factor Evaluation**: Demonstrates strong stock selection ability with stable excess returns[61] 2. Factor Name: Turnover Mean 1M - **Factor Construction Idea**: Uses the average turnover rate over the past month to identify stocks with high liquidity and market attention[61] - **Factor Construction Process**: 1. Calculate daily turnover rates for the past 1 month 2. Compute the average turnover rate 3. Normalize the factor by market capitalization and industry[61] - **Factor Evaluation**: Exhibits robust performance in identifying high-liquidity stocks with consistent excess returns[61] 3. Factor Name: FY1 Net Profit Change (1M) - **Factor Construction Idea**: Tracks changes in consensus forecasts for net profit (FY1) over the past month to gauge market sentiment[63] - **Factor Construction Process**: 1. Obtain consensus net profit forecasts for FY1 from 1 month ago and the current period 2. Calculate the percentage change: $ \text{Change} = \frac{\text{Current FY1 Forecast} - \text{1 Month Ago FY1 Forecast}}{\text{1 Month Ago FY1 Forecast}} $ 3. Normalize the factor by market capitalization and industry[63] - **Factor Evaluation**: Particularly effective in small-cap indices, reflecting market sensitivity to profit changes[63] --- Factor Backtesting Results 1. Return Std 1M - **1-Week Excess Return**: 1.27% - **1-Month Excess Return**: 1.14%[62] 2. Turnover Mean 1M - **1-Week Excess Return**: 1.26% - **1-Month Excess Return**: 1.12%[62] 3. FY1 Net Profit Change (1M) - **Excess Return in CSI 300**: 16.99% - **Excess Return in CSI 500**: 16.98% - **Excess Return in CSI 800**: 25.58% - **Excess Return in CSI 1000**: 9.70%[64]