量化点评报告：行业ETF轮动模型2025年超额9.3%

Quantitative Models and Construction Methods 1. Model Name: Industry Mainline Model (Relative Strength Index, RSI) - Model Construction Idea: The model identifies leading industries by calculating their relative strength (RS) and uses a threshold (RS > 90%) to signal potential outperforming sectors for the year [13] - Model Construction Process: 1. Use 29 first-level industry indices as the configuration targets [13] 2. Calculate the price change rates over the past 20, 40, and 60 trading days for each industry, and rank them cross-sectionally [13] 3. Normalize the rankings to obtain RS_20, RS_40, and RS_60 [13] 4. Compute the average of the three normalized rankings to derive the final RS index: $ RS = (RS_{20} + RS_{40} + RS_{60}) / 3 $ where RS_20, RS_40, and RS_60 represent the normalized rankings for 20, 40, and 60-day returns, respectively [13] 5. Industries with RS > 90% before the end of April are considered likely to lead the market for the year [13] - Model Evaluation: The model successfully identified key themes such as high dividends, resource products, overseas markets, and AI in 2024, which aligned with market trends during the year [13] 2. Model Name: Industry Rotation Model (Prosperity-Trend-Crowdedness Framework) - Model Construction Idea: This framework combines three dimensions—industry prosperity, trend strength, and crowdedness—to optimize sector allocation [2][6] - Model Construction Process: 1. Prosperity Model: Focuses on high prosperity and strong trends while avoiding highly crowded sectors [17] 2. Trend Model: Prioritizes strong trends and low crowdedness while avoiding low-prosperity sectors [17] 3. Combine the two models to form a comprehensive allocation strategy [17] - Model Evaluation: The combined framework adapts well to different market conditions, with strong historical performance metrics [17] 3. Model Name: Left-Side Inventory Reversal Model - Model Construction Idea: This model identifies industries in a turnaround phase by analyzing sectors with low inventory pressure and high analyst optimism, aiming to capture rebound opportunities during restocking cycles [26] - Model Construction Process: 1. Identify sectors currently or historically in distress but showing signs of recovery [26] 2. Focus on industries with low inventory pressure and favorable restocking conditions [26] 3. Incorporate analyst views to select sectors with long-term growth potential [26] - Model Evaluation: The model complements right-side models by targeting contrarian opportunities, achieving strong absolute and relative returns in recent years [26] --- Model Backtesting Results 1. Industry Mainline Model (RSI) - Annualized Excess Return: Not explicitly provided, but the model identified leading sectors with significant subsequent outperformance (e.g., banking: +32.1% absolute return after signal) [14][16] - Information Ratio (IR): Not explicitly provided [13][14] - Maximum Drawdown: Not explicitly provided [13][14] - Monthly Win Rate: Not explicitly provided [13][14] 2. Industry Rotation Model (Prosperity-Trend-Crowdedness Framework) - Annualized Excess Return: +14.0% relative to Wind All-A Index [17] - Information Ratio (IR): 1.52 [17] - Maximum Drawdown: -8.0% [17] - Monthly Win Rate: 68% [17] 3. Left-Side Inventory Reversal Model - Annualized Excess Return: +14.8% relative to equal-weighted industry benchmark in 2024; +5.8% in 2025 YTD [26] - Information Ratio (IR): Not explicitly provided [26] - Maximum Drawdown: Not explicitly provided [26] - Monthly Win Rate: Not explicitly provided [26] --- Quantitative Factors and Construction Methods 1. Factor Name: Prosperity Factor - Factor Construction Idea: Measures industry prosperity based on macro and micro indicators to identify high-growth sectors [17] - Factor Construction Process: 1. Aggregate macroeconomic and industry-specific indicators to assess prosperity levels [17] 2. Rank industries based on prosperity scores and select top-performing sectors [17] - Factor Evaluation: The factor effectively captures growth opportunities, contributing to the overall model's success [17] 2. Factor Name: Trend Factor - Factor Construction Idea: Captures momentum by identifying industries with strong upward trends [17] - Factor Construction Process: 1. Analyze price trends and volume data to assess momentum strength [17] 2. Rank industries based on trend scores and prioritize those with strong momentum [17] - Factor Evaluation: The factor enhances the model's ability to align with prevailing market trends [17] 3. Factor Name: Crowdedness Factor - Factor Construction Idea: Measures the level of investor positioning in industries to avoid overcrowded trades [17] - Factor Construction Process: 1. Use metrics such as fund flows and trading volumes to assess crowdedness [17] 2. Avoid industries with high crowdedness scores to mitigate risk [17] - Factor Evaluation: The factor provides a risk-control mechanism, improving the model's robustness [17] --- Factor Backtesting Results 1. Prosperity Factor - Annualized Excess Return: Not explicitly provided [17] - Information Ratio (IR): Not explicitly provided [17] - Maximum Drawdown: Not explicitly provided [17] - Monthly Win Rate: Not explicitly provided [17] 2. Trend Factor - Annualized Excess Return: Not explicitly provided [17] - Information Ratio (IR): Not explicitly provided [17] - Maximum Drawdown: Not explicitly provided [17] - Monthly Win Rate: Not explicitly provided [17] 3. Crowdedness Factor - Annualized Excess Return: Not explicitly provided [17] - Information Ratio (IR): Not explicitly provided [17] - Maximum Drawdown: Not explicitly provided [17] - Monthly Win Rate: Not explicitly provided [17]