高频选股因子周报：本周多数高频因子反弹，AI增强组合继续强势表现-2025-03-31

Core Insights - The report highlights a rebound in most high-frequency factors, with AI-enhanced portfolios continuing to perform strongly during the specified period [4][9][10]. High-Frequency Factors Performance - The multi-granularity model for deep learning factors showed a long-short return of 0.41% for the week, 1.34% for March, and 2.77% year-to-date in 2025 [4][10]. - The high-frequency skewness factor recorded long-short returns of 0.53% for the week, 1.24% for March, and 3.79% year-to-date in 2025 [4][10]. - The downside volatility proportion factor achieved long-short returns of 0.48% for the week, 1.64% for March, and 4.12% year-to-date in 2025 [4][10]. - The opening buy intention ratio factor had long-short returns of 0.02% for the week, -0.05% for March, and 0.65% year-to-date in 2025 [4][10]. - The opening large order net buy ratio factor reported long-short returns of 0.41% for the week, 0.86% for March, and 1.88% year-to-date in 2025 [4][10]. AI Enhanced Portfolios - The weekly rebalanced CSI 500 AI-enhanced portfolio yielded long-short returns of 1.06% for the week, 2.61% for March, and 0.69% year-to-date in 2025 [4][11]. - The weekly rebalanced CSI 1000 AI-enhanced portfolio achieved long-short returns of 1.61% for the week, 4.75% for March, and 4.00% year-to-date in 2025 [4][11]. Factor Analysis - The report includes detailed performance metrics for various high-frequency factors, indicating their effectiveness in generating returns [10][11]. - The analysis of the high-frequency skewness factor and downside volatility proportion factor shows their relative strength and profitability over time [4][10][11]. Conclusion - The report emphasizes the continued strength of AI-enhanced investment strategies and high-frequency factors, suggesting potential investment opportunities in these areas [4][10][11].