风格因子表现跟踪
Search documents
国泰海通|金工:3月建议超配小盘和价值风格,中长期继续看好小盘、成长风格
国泰海通证券研究· 2026-03-05 14:13
Group 1: Core Views - The report suggests an overweight position in small-cap and value styles for March, based on quantitative model signals indicating a preference for small-cap stocks [1] - The long-term outlook remains positive for small-cap and growth styles over the next year [1] Group 2: Small-Cap Style Rotation Strategy - As of the end of February, the quantitative model signal was 0.83, favoring small-cap stocks, with historical data showing small-cap outperformance in March [1] - The current market capitalization factor valuation spread is 0.86, indicating room for growth as it is below historical peaks of 1.7 to 2.6 [1] - The model has achieved a year-to-date return of 13.35%, outperforming the equal-weight benchmark return of 7.47% by 5.88% [1] Group 3: Value-Growth Style Rotation Strategy - The quantitative model signal for the end of February was -0.67, recommending an overweight in value style for March [2] - The long-term outlook favors growth style for the upcoming year [2] - The model's return as of the end of February was 5.22%, with no excess return compared to the equal-weight benchmark [2] Group 4: Factor Performance Tracking - Among eight major factors, liquidity and momentum factors showed high positive returns, while large-cap and quality factors exhibited negative returns [2] - Year-to-date, value and volatility factors have performed positively, while large-cap and quality factors have shown negative returns [2] - In a broader analysis of 24 style factors, liquidity, momentum, and long-term reversal factors had high positive returns, while residual volatility, large-cap, and book-to-market ratio factors had negative returns [2] Group 5: Factor Covariance Matrix Update - The report updates the factor covariance matrix as of February 27, 2026, which is crucial for predicting stock portfolio risks [3] - The multi-factor model allows for a breakdown of the stock covariance matrix into factor covariance and stock-specific risk matrices for more accurate estimations [3]
国泰海通|金工:综合量化模型信号和日历效应,12月建议超配大盘风格、价值风格
国泰海通证券研究· 2025-12-05 10:48
Core Insights - The report suggests an overweight allocation to large-cap and value styles for December based on quantitative model signals and calendar effects [1][2]. Size and Style Rotation Monthly Strategy - The latest quantitative model signal for the end of November is -0.17, indicating a preference for large-cap stocks. Historically, large-cap stocks have outperformed in December, leading to a recommendation for an overweight allocation in December [1]. - The year-to-date return for the size rotation quantitative model is 24.71%, with an excess return of 1.5% compared to an equal-weight benchmark of 23.21% [1]. - The combined strategy, incorporating subjective views, has yielded a return of 26.1%, with an excess return of 2.89% [1]. Value and Growth Style Rotation Monthly Strategy - The monthly quantitative model signal is -0.33, indicating a preference for value stocks. Historically, value style has slightly outperformed in December, leading to a recommendation for an overweight allocation in December [2]. - The year-to-date return for the value-growth style rotation model is 20.37%, with an excess return of 2.99% compared to an equal-weight benchmark of 16.88% [2]. Style Factor Performance Tracking - Among eight major factors, dividend and quality factors showed high positive returns in November, while large-cap and momentum factors exhibited high negative returns [2]. - For the year, volatility and growth factors had high positive returns, while liquidity and large-cap factors had high negative returns [2]. - In November, residual volatility, short-term reversal, and earnings quality factors had high positive returns, while momentum, profitability, and large-cap factors had high negative returns [2]. Factor Covariance Matrix Update - The report updates the latest factor covariance matrix as of November 28, 2025, which is essential for predicting stock portfolio risks using a multi-factor model [3].