2026 年 03 月 01 日 成交量快速反弹，市场情绪细分指 标出现回升 ——量化择时周报 20260301 相关研究 证券分析师 沈思逸 A0230521070001 shensy@swsresearch.com 邓虎 A0230520070003 denghu@swsresearch.com 联系人 沈思逸 A0230521070001 shensy@swsresearch.com 权 益 量 化 研 究 证 券 研 究 报 告 ⚫ 市场情绪维持稳定：截至 2 月 27 日，市场情绪指标数值为 1.85，较节前的 1.9 小幅下 降，情绪指标周内维持稳定，从情绪角度来看，模型观点偏中性。从具体分项来看，本周 部分情绪指标与上周相比出现回升，市场情绪出现一定回暖迹象，但仍需警惕外部政治风 险冲击。 ⚫ 成交量快速反弹，市场情绪细分指标出现回升：本周价量一致性指标快速回升，表明市场 当前价量匹配程度明显提升，资金关注度与标的涨幅关联性显著增强，短期内价格上涨幅 度与市场关注匹配程度形成较强共振，从价量一致性角度反映市场情绪出现明显修复；科 创 50 相对万得全 A 成交占比周内小幅降低，表明市场风险偏好边 ...

证券研究报告/金融工程定期报告 2026 年 03 月 01 日 分析师：吴先兴 执业证书编号：S0740525110003 Email：wuxx02@zts.com.cn 分析师：王鹏飞 执业证书编号：S0740525060001 Email：wangpf@zts.com.cn 量化择时周报：两会来临，短期关注政策驱动 1、《净利润断层策略本周超额收益 1.31%》2026-02-23 2、《量化择时周报：缩量信号如期出 现，聚焦科技与周期》2026-02-23 3.44%》2026-02-08 报告摘要 两会来临，短期关注政策驱动 请务必阅读正文之后的重要声明部分 上周周报（20260223）提示：综合来看，市场整体在海外影响以及日历效应的影响下， 风险偏好有望抬升，同时成交金额也到达我们预期的阶段地量水平，市场有望延续节 前反弹节奏。最终 WIND 全 A 全周上涨 2.75%，并创出新高。市值维度上，本周代表 小市值股票的中证 1000 上涨 4.34%，中盘股中证 500 指数上涨 4.32%，沪深 300 上 涨 1.08%，上证 50 上涨 0.17%;本周中信一级行业中，涨幅靠前的行业包括钢铁 ...

量化择时周报：两会来临，短期关注政策驱动 证券研究报告/金融工程定期报告 2026 年 03 月 01 日 分析师：吴先兴 执业证书编号：S0740525110003 Email：wuxx02@zts.com.cn 分析师：王鹏飞 执业证书编号：S0740525060001 Email：wangpf@zts.com.cn 1、《净利润断层策略本周超额收益 1.31%》2026-02-23 2、《量化择时周报：缩量信号如期出 现，聚焦科技与周期》2026-02-23 3.44%》2026-02-08 报告摘要 | 中泰量化周观点:两会来临,短期关注政策驱动 | | --- | | 风险提示 . | | 市场环境变动风险,模型基于历史数据。 . | | 图表 | 1 | ： | | | --- | --- | --- | --- | | 图表 | 2 | ： | PB 估值水平 4 | 两会来临，短期关注政策驱动 请务必阅读正文之后的重要声明部分 上周周报（20260223）提示：综合来看，市场整体在海外影响以及日历效应的影响下， 风险偏好有望抬升，同时成交金额也到达我们预期的阶段地量水平，市场有望延续节 前反弹 ...

[Table_Page] 金融工程|定期报告 识别风险，发现价值 请务必阅读末页的免责声明 1 / 23 2026 年 3 月 1 日 证券研究报告 数据来源：Wind，广发证券发展研究中心 | [Table_Title] 金融工程：AI 识图关注船舶、电 | | | | | | [分析师： Table_Author]安宁宁 | | --- | --- | --- | --- | --- | --- | --- | | 网、钢铁、机器人 | | | | | | SAC 执证号：S0260512020003 | | | | | | | | SFC CE No. BNW179 | | 股量化择时研究报告 | | A | | | | 0755-23948352 | | | | | | | | anningning@gf.com.cn | | ] [Table_Summary 市场回顾（本期是指 2026 年 年 27 日） | | 2 月 | 13 日—2026 | 2 月 | | 分析师： 张钰东 | | 中证 沪深 结构表现 | | 中证 | 中证 | 中证 | 国证 | SAC 执证号：S0260522070 ...

王武蕾 S1090519080001 wangwulei@cmschina.com.cn 王禹哲 S1090525080001 wangyuzhe@cmschina.com.cn 2. 市场最新观点 风险提示：择时和风格轮动模型结论基于合理假设前提下结合历史数据统计规 律推导而出，市场环境变化下可能导致出现模型失效风险。 定期报告 敬请阅读末页的重要说明 ❑ A 股节后量能如期回暖，但幅度并未超预期，而短期内基本面信号平稳、中 期估值偏高的整体环境并未改变，叠加外部风险增加以及重要会议期间相对 偏弱的日历效应，对 A 股整体维持震荡市的判断。 ❑ 国内维度上，"节后会前"反弹窗口期已近尾声，下周市场将进入"会议 期"交易。虽然 2025 年会议期间市场整体表现亮眼，但从 2014 年以来的统 计上看，会议开幕后 3 个交易日内中证 800 指数平均回撤为 0.82%，日历效 应偏弱，故对会议期间市场行情维持观望。此外，虽然节后量能如期反弹， 但并未出现明显放量，尤其是以沪深 300 为主的大盘风格尤为克制，仅局部 流动性明显回暖，故整体上看量能并未形成明显看多信号。 ❑ 海外维度上，地缘风险、关税扰动、AI ...

Investment Rating - The report indicates a positive outlook for the fixed income + product sector, with a significant increase in market attention and rapid growth in scale, reaching a historical high of 2.78 trillion yuan by December 31, 2025 [1][9]. Core Insights - 2025 is marked by heightened market interest and rapid growth in the fixed income + product sector, with total market size reaching 2.78 trillion yuan, an increase of 0.29 trillion yuan from Q3 2025 and 1.01 trillion yuan from 2024 [1][9]. - The report discusses three main dimensions: asset allocation changes, key industry focus for equity assets, and performance differences among various types of fixed income + products [9][10]. - The overall equity allocation in fixed income + products slightly decreased in Q4 2025 compared to Q3 2025 but remains at a relatively high level since Q4 2024, with a notable focus on technology and cyclical sectors [10][19]. - Performance-wise, the report highlights that the track-focused products outperformed balanced products in 2025, while balanced products demonstrated stronger long-term performance stability [3][26]. - The rapid growth in scale of conservative and aggressive products indicates an increase in investor risk appetite, with conservative products showing the fastest growth in 2025 [3][10]. Summary by Relevant Sections Asset Allocation - In Q4 2025, the median equity allocation for fixed income + products was 16.94%, down from 17.99% in Q3 2025, while bond allocations increased compared to Q3 2025 [18][19]. - The report notes a continuous decline in convertible bond allocations over four consecutive quarters, now below 5% [18][19]. - The report categorizes fixed income + products into conservative, stable, and aggressive types based on equity allocation, with stable and aggressive products showing higher bond duration to hedge against equity risks [16][17]. Performance Analysis - The annualized return for fixed income + products in 2025 was approximately 4.21%, with a Sharpe ratio of 1.54 and a Calmar ratio of 2.37, outperforming medium- and long-term pure bond funds [28]. - Track-focused products achieved a median annualized return exceeding 6% in 2025, while balanced products showed strong performance sustainability over the past five years [26][28]. - The report emphasizes that the performance of conservative products was relatively stable, achieving positive returns even in less favorable market conditions [28]. Scale and Growth - The report highlights that the scale of stable products grew the fastest in 2025, followed by aggressive products, reflecting an increase in investor risk tolerance [3][10]. - The growth in scale for track-focused products was significant, with the track rotation category increasing by over 300 billion yuan and the track concentration category exceeding 190 billion yuan [3][10]. - The report indicates that the long-term performance stability remains a core focus for investors, particularly for balanced products, which have the highest existing scale among the four types [3][10].

Group 1 - The core viewpoint of the article indicates that the market is expected to continue its oscillation in the upcoming week, based on various technical indicators and market sentiment models [1][2]. - The liquidity shock indicator for the CSI 300 index was reported at 6.21, which is higher than the previous week's 5.07, suggesting that current market liquidity is significantly above the average level over the past year [2]. - The PUT-CALL ratio for the SSE 50 ETF increased to 0.96 from 0.89, indicating a rising caution among investors regarding the short-term performance of the SSE 50 ETF [2]. Group 2 - The Shanghai Composite Index and Wind All A five-day average turnover rates were recorded at 1.34% and 1.97%, respectively, indicating a decrease in trading activity, positioned at the 77.24% and 82.76% percentiles since 2005 [2]. - The official manufacturing PMI for China in January was reported at 49.3, lower than the previous value of 50.1 and below the consensus expectation of 50.18, while the S&P Global China Manufacturing PMI was at 50.3, slightly above the previous value [2]. - The SAR indicator showed that the Wind All A index broke below the reversal indicator on February 2, indicating a potential downward trend [2]. Group 3 - The A-share market experienced fluctuations last week, with the SSE 50 index down by 0.93%, the CSI 300 index down by 1.33%, the CSI 500 index down by 2.68%, and the ChiNext index down by 3.28% [3]. - The current overall market PE (TTM) stands at 23.0 times, which is at the 81.0% percentile since 2005, indicating a relatively high valuation level [3]. - Observations on factor crowding indicate a decrease in high profitability factor crowding, with small-cap factor crowding at 0.06 and low valuation factor crowding at -0.31 [3].

Quantitative Models and Construction Methods 1. Growth-Value Rotation Model - **Model Name**: Growth-Value Rotation Model - **Model Construction Idea**: The model suggests overweighting growth stocks based on the current market environment and historical data analysis[4] - **Model Construction Process**: - The model evaluates the macroeconomic environment, valuation signals, short-term momentum signals, style breadth signals, and style congestion signals to determine the allocation between growth and value stocks[22] - The model uses the following signals: - Dynamic macro signal: 0% - Valuation reversion signal: 100% - Short-term momentum signal: 0% - Style breadth signal: 100% - Style congestion signal: 100%[23] - **Model Evaluation**: The model has shown a significant annualized return of 14.47% since 2011, with an annualized excess return of 7.90% over the benchmark[22] - **Model Test Results**: - Annualized return: 14.47% - Annualized volatility: 21.44% - Maximum drawdown: 40.08% - Sharpe ratio: 0.64 - Return-drawdown ratio: 0.36[23] 2. Small-Cap vs. Large-Cap Rotation Model - **Model Name**: Small-Cap vs. Large-Cap Rotation Model - **Model Construction Idea**: The model suggests overweighting large-cap stocks based on liquidity conditions and market trends[4] - **Model Construction Process**: - The model uses 11 effective rotation indicators to construct a comprehensive signal for rotating between small-cap and large-cap stocks[25] - The model evaluates the following indicators: - A-share Dragon Tiger List buying intensity: 0% - R007: 0% - Financing buy balance change: 0% - Thematic investment trading sentiment: 0% - Grade spread: 100% - Option volatility risk premium: 100% - Beta dispersion: 0% - PB differentiation: 0% - Block trading discount/premium rate: 100% - CSI 1000 MACD (10,20,10): 0% - CSI 1000 trading volume: 0%[27] - **Model Evaluation**: The model has consistently generated positive excess returns annually since 2014[26] - **Model Test Results**: - Annualized return: 20.61% - Annualized excess return: 13.18% - Maximum drawdown: 40.70% - Average turnover interval (trading days): 20 - Win rate (by trade): 50.00%[27] Model Backtest Results Growth-Value Rotation Model - Annualized return: 14.47% - Annualized volatility: 21.44% - Maximum drawdown: 40.08% - Sharpe ratio: 0.64 - Return-drawdown ratio: 0.36[23] Small-Cap vs. Large-Cap Rotation Model - Annualized return: 20.61% - Annualized excess return: 13.18% - Maximum drawdown: 40.70% - Average turnover interval (trading days): 20 - Win rate (by trade): 50.00%[27]

Quantitative Models and Construction Methods Model Name: Industry Trend Allocation Model - **Model Construction Idea**: This model aims to identify industry trends and allocate investments accordingly[5][8][10] - **Model Construction Process**: - The model uses various indicators to assess industry trends, including market performance, valuation levels, and risk appetite. - It incorporates signals from different sub-models such as the Mid-term Distress Reversal Expectation Model, TWO BETA Model, and Performance Trend Model. - The Mid-term Distress Reversal Expectation Model waits for reversal signals in industries like liquor and real estate. - The TWO BETA Model recommends the technology sector and monitors opportunities in commercial aerospace. - The Performance Trend Model focuses on the computing power industry chain and oversold sectors like non-ferrous metals and chemicals. - **Model Evaluation**: The model is effective in identifying industry trends and making allocation recommendations based on various market signals[5][8][10] Model Name: Timing System - **Model Construction Idea**: This model aims to distinguish the overall market environment and provide timing signals for investment decisions[5][8][9] - **Model Construction Process**: - The model uses the distance between the long-term moving average (120 days) and the short-term moving average (20 days) of the WIND All A Index. - The latest data shows the 20-day moving average at 6787 and the 120-day moving average at 6338, with a difference of 7.08%. - The model also considers the market trend line, which is currently around 6780 points, and the profitability effect, which is -1.44%. - The model suggests that the market is in a shock pattern and monitors short-term risk appetite changes. - **Model Evaluation**: The model provides clear signals for market timing based on moving averages and other indicators[5][8][9] Model Backtesting Results - **Industry Trend Allocation Model**: - **PE Valuation Level**: 90th percentile, indicating a high level[8][10] - **PB Valuation Level**: 50th percentile, indicating a medium level[8][10] - **Position Recommendation**: 70% for absolute return products with WIND All A as the stock allocation subject[8][10] - **Timing System**: - **Moving Average Distance**: 7.08%, greater than the absolute value of 3%[5][8][9] - **Market Trend Line**: Around 6780 points[5][8][9] - **Profitability Effect**: -1.44%, indicating a temporary end to the upward trend[5][8][9] Quantitative Factors and Construction Methods Factor Name: Mid-term Distress Reversal Expectation Model - **Factor Construction Idea**: This factor aims to identify potential reversal signals in distressed industries[5][8][10] - **Factor Construction Process**: - The model monitors industries like liquor and real estate for reversal signals. - It uses various market indicators to assess the likelihood of a reversal. - **Factor Evaluation**: The factor is useful for identifying potential investment opportunities in distressed industries[5][8][10] Factor Name: TWO BETA Model - **Factor Construction Idea**: This factor aims to recommend sectors with high growth potential, such as technology[5][8][10] - **Factor Construction Process**: - The model focuses on the technology sector and monitors opportunities in commercial aerospace. - It uses market performance and other indicators to make recommendations. - **Factor Evaluation**: The factor is effective in identifying high-growth sectors and making investment recommendations[5][8][10] Factor Name: Performance Trend Model - **Factor Construction Idea**: This factor aims to identify sectors with strong performance trends[5][8][10] - **Factor Construction Process**: - The model focuses on the computing power industry chain and oversold sectors like non-ferrous metals and chemicals. - It uses performance indicators to make recommendations. - **Factor Evaluation**: The factor is useful for identifying sectors with strong performance trends and making investment recommendations[5][8][10] Factor Backtesting Results - **Mid-term Distress Reversal Expectation Model**: - **PE Valuation Level**: 90th percentile, indicating a high level[8][10] - **PB Valuation Level**: 50th percentile, indicating a medium level[8][10] - **TWO BETA Model**: - **PE Valuation Level**: 90th percentile, indicating a high level[8][10] - **PB Valuation Level**: 50th percentile, indicating a medium level[8][10] - **Performance Trend Model**: - **PE Valuation Level**: 90th percentile, indicating a high level[8][10] - **PB Valuation Level**: 50th percentile, indicating a medium level[8][10]

- The report discusses multiple quantitative models for market timing, including short-term, medium-term, and long-term models. These models are constructed based on price-volume, acceleration and trend, momentum, and limit-up/down perspectives. The report emphasizes the importance of combining signals from different models and periods to achieve a balanced strategy[9][11][12] - The short-term models include the "Volume Model" (neutral), "Feature Institutional Model" (neutral), "Feature Volume Model" (bearish), "Smart Algorithm CSI 300 Model" (bullish), and "Smart Algorithm CSI 500 Model" (bearish)[11][70] - Medium-term models include the "Limit-Up/Down Model" (neutral), "Up-Down Return Difference Model" (bullish for some broad-based indices), and "Calendar Effect Model" (bullish)[12][71] - The long-term model is the "Long-Term Momentum Model," which is neutral[72] - Comprehensive models such as the "A-Share Comprehensive Weapon V3 Model" and "A-Share Comprehensive CSI 2000 Model" are neutral[73] - For Hong Kong stocks, the medium-term models include the "Turnover-to-Volatility Model" (bearish), "Hang Seng Index Up-Down Return Difference Model" (neutral), and "Up-Down Return Similarity Model" (bullish)[13][74] - Backtesting results for the "Cup-and-Handle Pattern" show a weekly decline of -0.44%, outperforming the Shanghai Composite Index by 0.83%. Since December 31, 2020, the cumulative return of this pattern is 19.67%, exceeding the Shanghai Composite Index by 2.61%[43][44] - Backtesting results for the "Double-Bottom Pattern" show a weekly decline of -0.88%, outperforming the Shanghai Composite Index by 0.39%. Since December 31, 2020, the cumulative return of this pattern is 23.45%, exceeding the Shanghai Composite Index by 6.39%[43][50]