微盘股指数周报：微盘股新高，成交占比快速回升-20251027

Quantitative Models and Construction - Model Name: Diffusion Index Model Construction Idea: The model monitors the future diffusion index's critical points to predict market changes[39][40] Construction Process: The diffusion index is calculated based on the relative price changes of constituent stocks over a specific time window. The horizontal axis represents the future price change percentage (from +10% to -10%), while the vertical axis represents the review period length (from 20 days to 10 days). For example, at a horizontal axis value of 0.95 and a vertical axis value of 15 days, the diffusion index is 0.42, indicating that if all constituent stocks drop by 5% after 5 days, the diffusion index value will be 0.42[39][41] Evaluation: The model is effective in identifying potential turning points in the market[40] - Model Name: First Threshold Method (Left-Side Trading) Construction Idea: This method triggers a buy signal when the diffusion index reaches a predefined threshold[43] Construction Process: The first threshold method triggered a buy signal on September 23, 2025, when the diffusion index reached 0.0575[43] Evaluation: The method is suitable for early-stage market entry[43] - Model Name: Delayed Threshold Method (Right-Side Trading) Construction Idea: This method provides a buy signal after the diffusion index stabilizes above the threshold[46] Construction Process: The delayed threshold method triggered a buy signal on September 25, 2025, when the diffusion index reached 0.1825[46] Evaluation: The method is more conservative and reduces false signals compared to the first threshold method[46] - Model Name: Dual Moving Average Method (Adaptive Trading) Construction Idea: This method uses two moving averages to generate trading signals based on their crossover[47] Construction Process: The dual moving average method provided a buy signal on October 13, 2025, at the market close[47] Evaluation: The method adapts to market trends and is effective in capturing medium-term signals[47] Model Backtesting Results - Diffusion Index Model: Current diffusion index value is 0.67, indicating a medium-high level. If the market rises by 4% next week, it will trigger the risk threshold[39][40] - First Threshold Method: Triggered buy signal at diffusion index value of 0.0575[43] - Delayed Threshold Method: Triggered buy signal at diffusion index value of 0.1825[46] - Dual Moving Average Method: Triggered buy signal on October 13, 2025[47] Quantitative Factors and Construction - Factor Name: Non-Liquidity Factor Construction Idea: Measures the illiquidity of stocks to predict their future performance[5][33] Construction Process: The rank IC for this factor is calculated weekly. This week, the rank IC is 0.147, compared to the historical average of 0.04[5][33] Evaluation: The factor shows strong predictive power this week[5][33] - Factor Name: Beta Factor Construction Idea: Captures the sensitivity of stock returns to market movements[5][33] Construction Process: The rank IC for this factor is 0.134 this week, compared to the historical average of 0.004[5][33] Evaluation: The factor demonstrates significant improvement in predictive ability this week[5][33] - Factor Name: Standardized Expected Earnings Factor Construction Idea: Reflects the expected profitability of stocks[5][33] Construction Process: The rank IC for this factor is 0.132 this week, compared to the historical average of 0.014[5][33] Evaluation: The factor is effective in identifying profitable stocks[5][33] - Factor Name: Single-Quarter Net Profit Growth Factor Construction Idea: Measures the growth rate of net profit in a single quarter[5][33] Construction Process: The rank IC for this factor is 0.088 this week, compared to the historical average of 0.02[5][33] Evaluation: The factor is moderately effective in predicting stock performance[5][33] - Factor Name: Single-Quarter ROE Factor Construction Idea: Reflects the return on equity for a single quarter[5][33] Construction Process: The rank IC for this factor is 0.045 this week, compared to the historical average of 0.022[5][33] Evaluation: The factor shows limited predictive power this week[5][33] Factor Backtesting Results - Non-Liquidity Factor: Rank IC 0.147, historical average 0.04[5][33] - Beta Factor: Rank IC 0.134, historical average 0.004[5][33] - Standardized Expected Earnings Factor: Rank IC 0.132, historical average 0.014[5][33] - Single-Quarter Net Profit Growth Factor: Rank IC 0.088, historical average 0.02[5][33] - Single-Quarter ROE Factor: Rank IC 0.045, historical average 0.022[5][33] Composite Strategy and Construction - Strategy Name: Small-Cap Low-Volatility 50 Strategy Construction Idea: Selects 50 stocks with small market capitalization and low volatility from micro-cap stocks[8][17][35] Construction Process: The strategy rebalances every two weeks. In 2024, the strategy achieved a return of 7.07% with an excess return of -2.93%. In 2025, the YTD return is 71.25%, with a weekly excess return of 2.65%. The benchmark is the Wind Micro-Cap Index (8841431.WI), and transaction costs are 0.3% on both sides[8][17][35] Evaluation: The strategy is highly effective in capturing the performance of small-cap stocks in 2025[8][17][35] Strategy Backtesting Results - Small-Cap Low-Volatility 50 Strategy: - 2024 Return: 7.07%, Excess Return: -2.93%[8][17][35] - 2025 YTD Return: 71.25%, Weekly Excess Return: 2.65%[8][17][35]