金融工程研究 跟踪周报 证券研究报告 2025 年 11 月 23 日 湘财证券研究所 核心要点： ❑ 市场表现 本周（2025.11.17-2025.11.21），根据 Wind，主要指数中，收益排名靠前的 为上证 50 和中证红利指数，收益分别为-2.72%和-3.69%；收益排名靠后的 为微盘指数和创业板指数，收益分别为-7.80%和-6.15%。 本年，根据 Wind，主要指数中，收益排名靠前的为微盘指数和创业板指， 收益分别为 66.12%和 36.35%；收益排名靠后的为中证红利和上证 50 指数， 收益分别为-0.48%和 10.10%。 指数增强策略跟踪周报 相关研究： 1. 《多因子量化选股系列之八： 中证1000指数增强策略改进》 2024.03.28 分析师：别璐莎 证书编号：S0500524010001 Tel：(021) 50293663 Email：bls06644 @xcsc.com 地址：上海市浦东新区银城路88号 中国人寿金融中心10楼 ❑ 策略收益 本周，根据 Wind，中证 1000 指数增强策略收益为-5.89%，同期指数收益 为-5.80%，相较于基准，策略超额收益为 ...

Core Insights - The report highlights the strong performance of the CSI 1000 Index in 2025, driven by its focus on small-cap companies in sectors such as new energy, semiconductors, and medical devices [5][20] - The report indicates that the CSI 1000 Index has shown significant returns, ranking in the middle among major indices for the year, with a year-to-date return of 31.03%, outperforming the benchmark by 6.50% [4][16] Market Performance - For the week of October 20-24, 2025, the top-performing indices were the ChiNext Index and the Sci-Tech 50 Index, with returns of 8.05% and 7.27% respectively, while the lowest were the CSI Dividend and SSE 50 indices, with returns of 1.05% and 2.63% [3][7] - Year-to-date, the Micro-Cap Index and ChiNext Index led with returns of 66.54% and 48.09%, while the CSI Dividend and SSE 50 indices lagged with returns of 1.32% and 13.45% [8] Strategy Performance - The CSI 1000 Index enhancement strategy yielded a return of 3.55% for the week, surpassing the index return of 3.25% by 0.30% [4][13] - For the month, the strategy achieved a return of 0.18%, while the index returned -2.06%, resulting in an excess return of 2.24% [15] - Year-to-date, the strategy's return was 31.03%, compared to the index's 24.53%, leading to an excess return of 6.50% [16] Investment Recommendations - The report suggests that the CSI 1000 Index remains a strong investment opportunity due to its strategic positioning in high-growth sectors and favorable policy signals following the recent political meetings [5][20] - The report emphasizes the importance of adjusting asset allocations towards lower volatility assets as the year-end approaches, while remaining cautious of the inherent volatility in the CSI 1000 Index [5][20]

Core Insights - The report emphasizes the development of a deep learning factor selection system for stock prediction and portfolio optimization, shifting from traditional logic-driven methods to data-driven approaches [7][10]. - The system integrates diverse data sources, including daily and minute market data, to enhance the performance of alpha signals [7][10]. - The report outlines the construction of multiple models that utilize different network architectures to extract unique alpha signals, demonstrating low correlation among them [8][54]. Data and Network - The input data consists of three categories: daily market data, minute market data, and manually crafted features, with neural networks independently extracting alpha features from each dataset [11]. - The report describes the use of Long Short-Term Memory (LSTM) networks combined with self-attention mechanisms to capture long-term dependencies in time series data [19]. - A Graph Attention Network (GAT) is employed to model the complex relationships between stocks, providing a global analysis perspective [20]. Alpha Models - The report presents various alpha models, including simple equal-weight, tree model weighting, and network weighting, with a focus on combining multiple signals to enhance robustness [3][3.1][3.2]. - The average Information Coefficient (IC) for the combined factors since 2019 is reported as 11.3% for 5-day IC and 12.4% for 10-day IC, indicating strong predictive power [31][32]. Risk Models - The report highlights the use of neural networks to identify high-dimensional non-linear risk patterns directly from raw price and volume data, enhancing risk control in portfolio construction [9]. Index Enhancement Strategies - The report details the performance of enhanced index strategies based on deep learning alpha signals, with annualized returns reported as follows: - CSI 300 enhanced portfolio: 18.2% annualized return, 14.2% excess return over the index [3][5.1]. - CSI 500 enhanced portfolio: 22.4% annualized return, 17.2% excess return over the index [3][5.2]. - CSI 1000 enhanced portfolio: 29.8% annualized return, 24.5% excess return over the index [3][5.3].