Quantitative Models and Construction Methods 1. Model Name: Industry Mainline Model (Relative Strength Index, RSI) - Model Construction Idea: The model aims to identify leading industries by calculating their relative strength index (RSI) over different time frames[1][9] - Model Construction Process: 1. Use primary industry indices as configuration targets, totaling 31 primary industries 2. Calculate the price changes over the past 20, 40, and 60 trading days for all industries, obtaining the cross-sectional rankings of these changes, then normalize all rankings to get RS_20, RS_40, and RS_60 3. Calculate the average of these three rankings to get the final industry relative strength index: 4. If an industry shows an RS signal greater than 90% before the end of April, it is likely to be a leading industry for the year[9] - Model Evaluation: The model effectively identified leading industries in 2024, such as coal, power and utilities, home appliances, banks, oil and petrochemicals, communications, non-ferrous metals, agriculture, forestry, animal husbandry, and fishery, and automobiles[1][9] 2. Model Name: Industry Rotation Model (Prosperity-Trend-Crowding Framework) - Model Construction Idea: The model uses a three-dimensional framework of prosperity, trend, and crowding to recommend industry allocations[1][2][6] - Model Construction Process: 1. Define two industry rotation schemes: "strong trend-low crowding" and "high prosperity-strong trend" 2. Allocate industry weights based on the framework: Media 16%, Agriculture, Forestry, Animal Husbandry, and Fishery 15%, Non-bank Financials 12%, Computers 12%, Home Appliances 9%, Coal 9%, Building Materials 7%, Banks 7%, Light Industry Manufacturing 7%, Retail 6% 3. Recommend ETFs tracking indices such as CSI Steel, CSI Agriculture, Securities Companies, Communication Equipment, CSI Media, Sub-segment Chemicals, CS Artificial Intelligence, Animation Games, Sub-segment Machinery, All Information, Building Materials, etc.[2][6][15] - Model Evaluation: The model performed well in 2025, with an excess return of 16.4% relative to the CSI 800 index and 4.2% relative to the Wind All A index[2][6][18] 3. Model Name: Left-Side Inventory Reversal Model - Model Construction Idea: The model aims to capture the reversal of industries in distress by analyzing sectors with low inventory pressure and long-term analyst optimism[24] - Model Construction Process: 1. Identify sectors currently or previously in distress with potential for inventory replenishment 2. Analyze sectors with low inventory pressure and long-term analyst optimism 3. Recommend sub-sectors such as cloud services, other light industries, oil service engineering, components, agricultural chemicals, animal husbandry, consumer electronics, special materials, and biomedicine[24][25] - Model Evaluation: The model achieved an absolute return of 25.4% in 2025, with an excess return of 5.4% relative to the industry equal weight index[24][27] Model Backtest Results 1. Industry Mainline Model (RSI) - Absolute Return: Various industries showed significant returns after the RSI signal appeared, such as banks (32.1%), communications (24.0%), home appliances (25.8%), and automobiles (12.8%)[10][12] 2. Industry Rotation Model (Prosperity-Trend-Crowding Framework) - Annualized Return: 21.7% - Excess Annualized Return: 13.8% - Information Ratio (IR): 1.5 - Maximum Drawdown: -8.0% - Monthly Win Rate: 67% - Excess Return in 2023: 7.3% - Excess Return in 2024: 5.7% - Excess Return in 2025: 4.2%[13][14] 3. Left-Side Inventory Reversal Model - Absolute Return in 2023: 13.4% - Excess Return in 2023: 17.0% - Absolute Return in 2024: 26.5% - Excess Return in 2024: 15.4% - Absolute Return in 2025: 25.4% - Excess Return in 2025: 5.4%[24][27]
行业ETF配置模型2025年超额16.4%
GOLDEN SUN SECURITIES·2025-12-07 10:20