东方因子周报：Trend风格登顶，预期EPTTM因子表现出色-2025-03-16

Quantitative Factors and Models Summary Quantitative Factors and Their Construction - Factor Name: Trend - Construction Idea: Measures market preference for trend-following strategies, using exponential weighted moving averages (EWMA) with different half-lives [12] - Construction Process: - Trend_120: $ EWMA(halflife=20) / EWMA(halflife=120) $ - Trend_240: $ EWMA(halflife=20) / EWMA(halflife=240) $ [12] - Evaluation: Demonstrates strong performance in short-term market rebounds, indicating increased preference for trend-following strategies [9] - Factor Name: Certainty - Construction Idea: Captures market confidence through institutional holdings and analyst coverage [12] - Construction Process: - Instholder Pct: Proportion of public fund holdings - Cov: Analyst coverage adjusted for market capitalization - Listdays: Number of days since listing [12] - Evaluation: Shows recovery in market confidence during the observed period [9] - Factor Name: Value - Construction Idea: Measures valuation attractiveness using metrics like book-to-price (BP) and earnings yield (EP) [12] - Construction Process: - BP: $ Net\ Assets / Market\ Capitalization $ - EP: $ Earnings / Market\ Capitalization $ [12] - Evaluation: Underperformed during the observed period, reflecting reduced market preference for value stocks [9] - Factor Name: Volatility - Construction Idea: Captures market risk perception through historical and idiosyncratic volatility measures [12] - Construction Process: - Stdvol: Standard deviation of daily returns over 243 days - Ivff: Idiosyncratic volatility from Fama-French 3-factor model over 243 days - Range: $ (High\ Price - Low\ Price) / Low\ Price $ over 243 days [12] - Evaluation: Declined significantly, indicating increased market aversion to high-volatility assets [10] - Factor Name: Size - Construction Idea: Measures the impact of company size on returns using logarithmic market capitalization [12] - Construction Process: $ Log(Market\ Capitalization) $ [12] - Evaluation: Continued to underperform, reflecting negative sentiment towards smaller companies [10] Factor Backtesting Results - Trend Factor - Weekly return: 1.49% [9] - Monthly return: -4.81% [11] - Year-to-date return: -10.89% [11] - Historical annualized return: 13.86% [11] - Certainty Factor - Weekly return: 1.35% [9] - Monthly return: -2.32% [11] - Year-to-date return: -3.44% [11] - Historical annualized return: 3.20% [11] - Value Factor - Weekly return: 0.35% [9] - Monthly return: -2.42% [11] - Year-to-date return: -10.53% [11] - Historical annualized return: 7.28% [11] - Volatility Factor - Weekly return: -0.51% [10] - Monthly return: 5.01% [11] - Year-to-date return: 16.90% [11] - Historical annualized return: -12.84% [11] - Size Factor - Weekly return: -4.61% [10] - Monthly return: -12.65% [11] - Year-to-date return: -22.68% [11] - Historical annualized return: -29.42% [11] Composite Factor Portfolio Construction - Model Name: Maximized Factor Exposure (MFE) Portfolio - Construction Idea: Optimizes portfolio to maximize exposure to a single factor while controlling for industry, style, and turnover constraints [52] - Construction Process: - Objective function: $ max\ f^{T}w $ - Constraints: - Style exposure: $ s_{l} \leq X(w-w_{b}) \leq s_{h} $ - Industry exposure: $ h_{l} \leq H(w-w_{b}) \leq h_{h} $ - Stock weight deviation: $ w_{l} \leq w-w_{b} \leq w_{h} $ - Turnover limit: $ \Sigma|w-w_{0}| \leq to_{h} $ [52][53] - Evaluation: Provides a robust framework for testing factor effectiveness under realistic portfolio constraints [53] MFE Portfolio Backtesting Results - MFE Portfolio for Trend Factor - Weekly return: 1.49% [9] - Monthly return: -4.81% [11] - Year-to-date return: -10.89% [11] - Historical annualized return: 13.86% [11]