金工专题：微盘股还能配置吗？兼谈三季报小微盘拥挤度

Quantitative Models and Construction Methods - Model Name: Diffusion Index Timing Model (First Threshold Method) Model Construction Idea: The model uses the proportion of stocks in the Wind Micro-Cap Index that have been in an upward trend over the past 20 trading days to measure market sentiment for short-term timing decisions[66][67] Model Construction Process: - When the diffusion index exceeds 0.9, the portfolio is emptied; when it falls below 0.1, the portfolio is fully invested. - Signals are filled based on the previous period unless the portfolio transitions from fully invested to empty or vice versa. - Formula: Diffusion Index = (Number of stocks in an upward trend over the past 20 days) / (Total number of stocks in the index)[66][67] Model Evaluation: The model struggles to outperform the Wind Micro-Cap Index in sample data, with controlled drawdowns. However, it fails to identify major bear markets and prematurely exits during bull markets, leading to suboptimal performance[67][68] - Model Name: Diffusion Index Timing Model (Delayed Threshold Method) Model Construction Idea: This model refines the first threshold method by introducing a delay mechanism to improve timing accuracy[66][69] Model Construction Process: - When the diffusion index drops below 0.9 after exceeding it the previous day, the portfolio is emptied. Conversely, when it rises above 0.1 after being below it the previous day, the portfolio is fully invested. - Signals are filled based on the previous period unless the portfolio transitions from fully invested to empty or vice versa. - Formula: Diffusion Index = (Number of stocks in an upward trend over the past 20 days) / (Total number of stocks in the index)[66][69] Model Evaluation: The model performs better than the first threshold method, avoiding premature exits during bull markets and achieving new highs in strategy net value. However, it still struggles to identify major bear markets[69][71] - Model Name: Diffusion Index Timing Model (Dual Moving Average Method) Model Construction Idea: This model uses short-term and long-term moving averages of the diffusion index to adapt to market trends and reduce reliance on fixed thresholds[66][70] Model Construction Process: - Short-term moving average: 10-day MA of the diffusion index - Long-term moving average: 20-day MA of the diffusion index - When the short-term MA crosses above the long-term MA, the portfolio is fully invested; when it crosses below, the portfolio is emptied. - Formula: Diffusion Index = (Number of stocks in an upward trend over the past 20 days) / (Total number of stocks in the index)[66][72] Model Evaluation: The model effectively controls drawdowns during bear markets and retains gains during bull markets, making it the most satisfactory among the three methods[72][73] --- Model Backtesting Results - First Threshold Method: - Sample data performance: Unable to outperform the Wind Micro-Cap Index[67][68] - Drawdown control: Moderate[67][68] - Major bear market identification: Failed[67][68] - Bull market performance: Premature exit during September 2024 bull market[67][68] - Delayed Threshold Method: - Sample data performance: Unable to outperform the Wind Micro-Cap Index[69][71] - Drawdown control: Moderate[69][71] - Major bear market identification: Failed[69][71] - Bull market performance: Retained gains during September 2024 bull market, achieving new highs in strategy net value[69][71] - Dual Moving Average Method: - Sample data performance: Unable to outperform the Wind Micro-Cap Index[72][73] - Drawdown control: Excellent[72][73] - Major bear market identification: Successfully avoided significant losses during bear markets[72][73] - Bull market performance: Retained gains during September 2024 bull market, achieving satisfactory results[72][73] --- Quantitative Factors and Construction Methods - Factor Name: Micro-Cap Stock Diffusion Index Factor Construction Idea: Measures the proportion of stocks in the Wind Micro-Cap Index that have been in an upward trend over the past 20 trading days to gauge market sentiment[66][67] Factor Construction Process: - Formula: Diffusion Index = (Number of stocks in an upward trend over the past 20 days) / (Total number of stocks in the index)[66][67] Factor Evaluation: Provides a reliable measure of short-term market sentiment but requires additional mechanisms for effective timing decisions[66][67] --- Factor Backtesting Results - Micro-Cap Stock Diffusion Index: - Sample data performance: Used as the basis for three timing models, with varying degrees of success[66][67][69][72] - Drawdown control: Dependent on the timing model applied[66][67][69][72] - Major bear market identification: Limited effectiveness without additional mechanisms[66][67][69][72] - Bull market performance: Retained gains during September 2024 bull market when paired with the dual moving average method[66][72][73]