Quantitative Models and Construction Methods - Model Name: Industry Profitability Tracking Strategy Model Construction Idea: This model aims to identify industries with upward profitability expectations based on rolling industry data from Wind analysts' consensus forecasts [14] Model Construction Process: 1. Collect rolling industry data from Wind analysts' consensus forecasts 2. Build a multi-factor model to track profitability expectations across industries 3. Rank industries based on profitability expectations [14] Model Evaluation: The model effectively identifies industries with high profitability expectations, such as non-cyclical sectors [14] - Model Name: Unverified Sentiment Tracking Strategy Model Construction Idea: This model addresses the challenge of market valuations often preceding sell-side profitability forecasts by constructing a sentiment momentum strategy that anticipates profitability data releases [17] Model Construction Process: 1. Analyze market sentiment and valuation trends 2. Build a momentum strategy based on implicit sentiment indicators 3. Rank industries based on sentiment momentum [17] Model Evaluation: The model captures market sentiment shifts effectively, providing early signals for industry rotation [17] - Model Name: Macro Style Industry Rotation Strategy Model Construction Idea: This model predicts the long-short positions of four industry styles (high beta, high valuation, 12-month momentum, high volatility) based on macro indicators and ranks industries by their exposure to these styles [22] Model Construction Process: 1. Analyze macro indicators and their correlation with industry style returns 2. Predict long-short positions for the four styles 3. Rank industries based on their alignment with macro style predictions [22] Model Evaluation: The model provides a systematic approach to align industry rotation with macroeconomic trends [22] Model Backtesting Results - Industry Profitability Tracking Strategy: Year-to-date excess return of 10.00% compared to the equal-weighted benchmark of CICC primary industries [33] - Unverified Sentiment Tracking Strategy: Year-to-date excess return of 6.22% compared to the equal-weighted benchmark of CICC primary industries [33] - Macro Style Industry Rotation Strategy: Year-to-date excess return of -9.45% compared to the equal-weighted benchmark of CICC primary industries [33] Quantitative Factors and Construction Methods - Factor Name: Rolling 6-Year PB Valuation Percentile Factor Construction Idea: This factor aims to avoid high valuation risks by calculating the percentile of current PB valuations within a rolling 6-year window, excluding extreme values [9] Factor Construction Process: 1. Exclude the top 10% of PB values in the rolling 6-year window to ensure robust estimation 2. Calculate the current PB valuation percentile within the adjusted window 3. Exclude industries with PB percentiles above 95% from strategy calculations [9][10] Factor Evaluation: The factor effectively identifies industries with extreme valuation risks, such as oil and petrochemical sectors [10] Factor Backtesting Results - Rolling 6-Year PB Valuation Percentile: - Oil and Petrochemical: 96.8% (triggered high valuation warning) - Coal: 92.0% - Electric Power and Utilities: 91.9% [10][11]
中银国际:中银量化多策略行业轮动周报-20240812
Bank of China Securities·2024-08-11 09:27