Quantitative Models and Construction Methods 1. Model Name: Style Rotation Model - **Model Construction Idea**: The model is based on asset pricing theory, incorporating factors that influence profit expectations, discount rates, and investor sentiment. It uses historical data to form a logical, quantifiable, and effective strategy[38]. - **Model Construction Process**: - **Macro Level**: Utilizes an event-driven approach to study the relationship between styles and macroeconomic factors. Six dimensions are considered: economic growth, consumption, monetary policy, interest rates, exchange rates, and real estate. Five event patterns are defined, including historical highs/lows, marginal improvement trends, exceeding expectations, and new highs/lows. The model evaluates the relative returns, information ratios (IR), excess monthly win rates, and correlations of style indices within one month after macro events[38]. - **Market State**: Reflects investor sentiment and risk appetite. Proxy variables include monthly returns, turnover rates, volatility, ERP, BP, DRP, and excess returns of the CSI Dividend Index. Event study methods are used to analyze the relationship between market state and style rotation[38]. - **Micro Features**: Based on multi-factor models, the model incorporates performance changes, capital flows, and trading sentiment of listed companies. It emphasizes the relative position of values rather than absolute values. Backtesting shows momentum effects in performance, capital preference, and trading activity[39]. 2. Model Name: Industry Rotation Model - **Model Construction Idea**: Focuses on micro-level industry rotation due to the difficulty of capturing macro drivers with available data. It adopts a bottom-up perspective to propose effective micro-industry indicators[40]. - **Model Construction Process**: - **Micro Indicators**: Includes fundamental, technical, and analyst-based factors. - **Fundamental**: Historical changes in fundamentals and marginal changes in analyst consensus forecasts. - **Technical**: Adjusted industry momentum and stripped limit-up momentum. - **Analyst**: Analyst-based factors reflecting industry expectations[40][44]. --- Model Backtesting Results 1. Style Rotation Model - **Macro Level**: Evaluates the impact of macro events on style indices' relative returns, IR, and excess monthly win rates[38]. - **Market State**: Uses proxy variables like monthly returns, turnover rates, and volatility to assess the relationship with style rotation[38]. - **Micro Features**: Backtesting confirms momentum effects in performance, capital flows, and trading activity[39]. 2. Industry Rotation Model - **Micro Indicators**: Backtesting results highlight the effectiveness of fundamental, technical, and analyst-based factors in capturing industry rotation signals[40][44]. --- Quantitative Factors and Construction Methods 1. Factor Name: Revenue Surprise (营收超预期) - **Factor Construction Idea**: Measures the degree to which revenue exceeds expectations, reflecting growth potential[12][15]. - **Factor Construction Process**: - **Formula**: Not explicitly provided in the report. - **Factor Evaluation**: Strong performance in recent months, with a positive direction[15]. 2. Factor Name: Annual Momentum (年动量) - **Factor Construction Idea**: Captures price momentum over a one-year horizon, indicating price trends[12][15]. - **Factor Construction Process**: - **Formula**: Not explicitly provided in the report. - **Factor Evaluation**: Positive performance, indicating strong price momentum[15]. 3. Factor Name: Analyst ROE Forecast Change (一致预测ROE环比变化) - **Factor Construction Idea**: Reflects changes in analysts' ROE forecasts over three months, indicating market expectations[12][15]. - **Factor Construction Process**: - **Formula**: Not explicitly provided in the report. - **Factor Evaluation**: Positive performance, showing strong alignment with market sentiment[15]. 4. Factor Name: Quarterly Net Profit YoY Growth (季度净利润同比增速) - **Factor Construction Idea**: Measures year-over-year growth in quarterly net profit, reflecting growth potential[12][15]. - **Factor Construction Process**: - **Formula**: Not explicitly provided in the report. - **Factor Evaluation**: Positive performance, indicating strong growth signals[15]. --- Factor Backtesting Results 1. Revenue Surprise - **1-Month Excess Return**: 4.4% - **3-Month Excess Return**: 3.7% - **6-Month Excess Return**: 6.0% - **12-Month Excess Return**: 7.5%[15] 2. Annual Momentum - **1-Month Excess Return**: 4.4% - **3-Month Excess Return**: 5.1% - **6-Month Excess Return**: 5.9% - **12-Month Excess Return**: 6.5%[15] 3. Analyst ROE Forecast Change - **1-Month Excess Return**: 4.1% - **3-Month Excess Return**: 7.2% - **6-Month Excess Return**: 9.2% - **12-Month Excess Return**: 10.7%[15] 4. Quarterly Net Profit YoY Growth - **1-Month Excess Return**: 3.1% - **3-Month Excess Return**: 6.3% - **6-Month Excess Return**: 8.5% - **12-Month Excess Return**: 12.0%[15]