小市值低波50策略

Search documents
微盘股指数周报:微盘股将再次迎来高胜率区间-20250804
China Post Securities· 2025-08-04 07:48
Quantitative Models and Construction Methods 1. Model Name: Diffusion Index Model - **Model Construction Idea**: The model is used to monitor the critical points of trend changes in the micro-cap stock index by analyzing the distribution of stock price movements over a specific time window [5][39] - **Model Construction Process**: The diffusion index is calculated based on the relative price changes of constituent stocks over a retrospective or forward-looking window. For example, if the horizontal axis is 0.95 and the vertical axis is 15 days, the value of 0.31 indicates that if all constituent stocks drop by 5% after 5 days, the diffusion index value is 0.31. The model uses thresholds to signal trading actions: - **First Threshold Method (Left-Side Trading)**: Triggered a sell signal on May 8, 2025, when the index reached 0.9850 [43] - **Delayed Threshold Method (Right-Side Trading)**: Triggered a sell signal on May 15, 2025, at 0.8975 [47] - **Double Moving Average Method (Adaptive Trading)**: Triggered a buy signal on July 3, 2025 [48] - **Model Evaluation**: The model effectively identifies trend changes but may be influenced by the distribution of constituent stocks and their updates [39][40] 2. Model Name: Small-Cap Low-Volatility 50 Strategy - **Model Construction Idea**: This strategy selects 50 stocks with small market capitalization and low volatility from the micro-cap stock index, rebalancing every two weeks [7][36] - **Model Construction Process**: - Select stocks with the smallest market capitalization and lowest volatility from the micro-cap index - Rebalance the portfolio bi-weekly - Benchmark: Wind Micro-Cap Stock Index (8841431.WI) - Transaction cost: 0.3% on both sides [7][36] - **Model Evaluation**: The strategy demonstrates strong performance in 2025 but underperformed in 2024, indicating sensitivity to market conditions [7][36] --- Model Backtesting Results 1. Diffusion Index Model - **First Threshold Method**: Triggered sell signal at 0.9850 on May 8, 2025 [43] - **Delayed Threshold Method**: Triggered sell signal at 0.8975 on May 15, 2025 [47] - **Double Moving Average Method**: Triggered buy signal on July 3, 2025 [48] 2. Small-Cap Low-Volatility 50 Strategy - **2024 Return**: 7.07%, underperformed by -2.93% relative to the benchmark [7][36] - **2025 YTD Return**: 69.79%, underperformed by -1.88% relative to the benchmark [7][36] --- Quantitative Factors and Construction Methods 1. Factor Name: Unadjusted Stock Price Factor - **Factor Construction Idea**: Measures the rank IC of unadjusted stock prices within the micro-cap stock index [4][17] - **Factor Construction Process**: - Calculate the rank IC of unadjusted stock prices weekly - Compare with historical averages for evaluation [4][17] - **Factor Evaluation**: Demonstrated strong performance this week with a rank IC of 0.177, significantly above the historical average of -0.015 [4][17] 2. Factor Name: Beta Factor - **Factor Construction Idea**: Measures the systematic risk of stocks within the micro-cap stock index [4][17] - **Factor Construction Process**: - Calculate the beta of each stock relative to the market - Evaluate weekly rank IC and compare with historical averages [4][17] - **Factor Evaluation**: Performed well this week with a rank IC of 0.15, above the historical average of 0.006 [4][17] 3. Factor Name: Illiquidity Factor - **Factor Construction Idea**: Captures the illiquidity of stocks within the micro-cap stock index [4][17] - **Factor Construction Process**: - Measure illiquidity based on trading volume and price impact - Evaluate weekly rank IC and compare with historical averages [4][17] - **Factor Evaluation**: Strong performance this week with a rank IC of 0.143, above the historical average of 0.04 [4][17] 4. Factor Name: 10-Day Return Factor - **Factor Construction Idea**: Tracks the short-term momentum of stocks within the micro-cap stock index [4][17] - **Factor Construction Process**: - Calculate the 10-day return for each stock - Evaluate weekly rank IC and compare with historical averages [4][17] - **Factor Evaluation**: Positive performance this week with a rank IC of 0.105, above the historical average of -0.061 [4][17] 5. Factor Name: PE_TTM Reciprocal Factor - **Factor Construction Idea**: Measures valuation based on the reciprocal of the trailing twelve-month price-to-earnings ratio [4][17] - **Factor Construction Process**: - Calculate the reciprocal of PE_TTM for each stock - Evaluate weekly rank IC and compare with historical averages [4][17] - **Factor Evaluation**: Moderate performance this week with a rank IC of 0.041, above the historical average of 0.017 [4][17] --- Factor Backtesting Results Top 5 Factors by Weekly Rank IC 1. **Unadjusted Stock Price Factor**: Weekly rank IC = 0.177, Historical Average = -0.015 [4][17] 2. **Beta Factor**: Weekly rank IC = 0.15, Historical Average = 0.006 [4][17] 3. **Illiquidity Factor**: Weekly rank IC = 0.143, Historical Average = 0.04 [4][17] 4. **10-Day Return Factor**: Weekly rank IC = 0.105, Historical Average = -0.061 [4][17] 5. **PE_TTM Reciprocal Factor**: Weekly rank IC = 0.041, Historical Average = 0.017 [4][17] Bottom 5 Factors by Weekly Rank IC 1. **Turnover Factor**: Weekly rank IC = -0.189, Historical Average = -0.082 [4][17] 2. **Momentum Factor**: Weekly rank IC = -0.132, Historical Average = -0.005 [4][17] 3. **Residual Volatility Factor**: Weekly rank IC = -0.13, Historical Average = -0.04 [4][17] 4. **10-Day Free Float Turnover Factor**: Weekly rank IC = -0.12, Historical Average = -0.062 [4][17] 5. **Liquidity Factor**: Weekly rank IC = -0.118, Historical Average = -0.041 [4][17]
微盘股指数周报:调整仍不充分-20250623
China Post Securities· 2025-06-23 07:10
Quantitative Models and Construction Methods Diffusion Index Model - Model Name: Diffusion Index Model - Model Construction Idea: The model monitors the critical point of future diffusion index changes to predict market trends. - Model Construction Process: - The horizontal axis represents the relative price change of stocks in the future, ranging from 1.1 to 0.9, indicating a 10% rise to a 10% fall. - The vertical axis represents the length of the review period or future days, ranging from 20 to 10 days. - Example: A value of 0.07 at the horizontal axis 0.95 and vertical axis 15 days indicates that if all stocks in the micro-cap index fall by 5% after 5 days, the diffusion index value is 0.07. - Formula: $ \text{Diffusion Index} = \frac{\text{Number of stocks rising}}{\text{Total number of stocks}} $ - Model Evaluation: The model is useful for monitoring the critical point of future diffusion index changes and predicting market trends.[6][17][40] First Threshold Method (Left-side Trading) - Model Name: First Threshold Method - Model Construction Idea: The model triggers a signal based on the first threshold value to indicate trading actions. - Model Construction Process: - The model triggered a no-position signal at the closing value of 0.9850 on May 8, 2025. - Formula: $ \text{Threshold Value} = \text{Current Index Value} $ - Model Evaluation: The model provides early signals for trading actions based on threshold values.[6][43][44] Delayed Threshold Method (Right-side Trading) - Model Name: Delayed Threshold Method - Model Construction Idea: The model triggers a signal based on the delayed threshold value to indicate trading actions. - Model Construction Process: - The model triggered a no-position signal at the closing value of 0.8975 on May 15, 2025. - Formula: $ \text{Delayed Threshold Value} = \text{Current Index Value} $ - Model Evaluation: The model provides delayed signals for trading actions based on threshold values.[6][45][47] Dual Moving Average Method (Adaptive Trading) - Model Name: Dual Moving Average Method - Model Construction Idea: The model uses dual moving averages to trigger trading signals. - Model Construction Process: - The model triggered a no-position signal at the closing value on June 11, 2025. - Formula: $ \text{Signal} = \text{Short-term Moving Average} - \text{Long-term Moving Average} $ - Model Evaluation: The model adapts to market changes using dual moving averages to provide trading signals.[6][48][49] Model Backtesting Results Diffusion Index Model - Diffusion Index Model, Current Value: 0.34[40] First Threshold Method (Left-side Trading) - First Threshold Method, Closing Value: 0.9850[43] Delayed Threshold Method (Right-side Trading) - Delayed Threshold Method, Closing Value: 0.8975[47] Dual Moving Average Method (Adaptive Trading) - Dual Moving Average Method, Closing Value: Not specified[48] Quantitative Factors and Construction Methods Past Year Volatility Factor - Factor Name: Past Year Volatility Factor - Factor Construction Idea: The factor measures the volatility of stocks over the past year. - Factor Construction Process: - Formula: $ \text{Volatility} = \sqrt{\frac{\sum (R_i - \bar{R})^2}{N}} $ - This week's rank IC: 0.171, Historical average: -0.033 - Factor Evaluation: The factor is effective in capturing the volatility of stocks over the past year.[5][16][33] Beta Factor - Factor Name: Beta Factor - Factor Construction Idea: The factor measures the sensitivity of stocks to market movements. - Factor Construction Process: - Formula: $ \beta = \frac{\text{Cov}(R_i, R_m)}{\text{Var}(R_m)} $ - This week's rank IC: 0.145, Historical average: 0.004 - Factor Evaluation: The factor is effective in capturing the sensitivity of stocks to market movements.[5][16][33] Logarithmic Market Value Factor - Factor Name: Logarithmic Market Value Factor - Factor Construction Idea: The factor measures the logarithmic market value of stocks. - Factor Construction Process: - Formula: $ \text{Log Market Value} = \log(\text{Market Value}) $ - This week's rank IC: 0.138, Historical average: -0.033 - Factor Evaluation: The factor is effective in capturing the logarithmic market value of stocks.[5][16][33] Nonlinear Market Value Factor - Factor Name: Nonlinear Market Value Factor - Factor Construction Idea: The factor measures the nonlinear market value of stocks. - Factor Construction Process: - Formula: $ \text{Nonlinear Market Value} = (\text{Market Value})^2 $ - This week's rank IC: 0.138, Historical average: -0.033 - Factor Evaluation: The factor is effective in capturing the nonlinear market value of stocks.[5][16][33] Non-liquidity Factor - Factor Name: Non-liquidity Factor - Factor Construction Idea: The factor measures the non-liquidity of stocks. - Factor Construction Process: - Formula: $ \text{Non-liquidity} = \frac{\text{Number of non-trading days}}{\text{Total number of days}} $ - This week's rank IC: 0.125, Historical average: 0.038 - Factor Evaluation: The factor is effective in capturing the non-liquidity of stocks.[5][16][33] Factor Backtesting Results Past Year Volatility Factor - Past Year Volatility Factor, This week's rank IC: 0.171, Historical average: -0.033[5][16][33] Beta Factor - Beta Factor, This week's rank IC: 0.145, Historical average: 0.004[5][16][33] Logarithmic Market Value Factor - Logarithmic Market Value Factor, This week's rank IC: 0.138, Historical average: -0.033[5][16][33] Nonlinear Market Value Factor - Nonlinear Market Value Factor, This week's rank IC: 0.138, Historical average: -0.033[5][16][33] Non-liquidity Factor - Non-liquidity Factor, This week's rank IC: 0.125, Historical average: 0.038[5][16][33]