2025 年 09 月 22 日 华泰柏瑞量化：系统化投研体系， 追求持续稳定的超额收益 ——基金经理研究系列报告之八十一 相关研究 - 证券分析师 奚佳诚 A0230523070004 xijc@swsresearch.com 蒋辛 A0230521080002 jiangxin@swsresearch.com 邓虎 A0230520070003 denghu@swsresearch.com 联系人 奚佳诚 (8621)23297818× xijc@swsresearch.com 本研究报告仅通过邮件提供给 中庚基金 使用。1 请务必仔细阅读正文之后的各项信息披露与声明 权 益 量 化 研 究 股 票 基 金 证 券 研 究 报 告 股票基金 股票基金 目录 1. 华泰柏瑞量化——系统化投研体系，追求持续稳定的超额 | 收益 | 4 | | --- | --- | | 1.1 | 团队情况：团队内有多位从业经验丰富的投资经理 4 | | 1.2 | 产品布局：涵盖多种类型的宽基指增及其他策略产品 5 | | 1.3 | 投资框架：构建系统化投研体系，打造高效且持续迭代的量化投资 | | 框架 | 6 | | ...

文 ｜ 陆茗 编辑 ｜ 向现 来源｜ 南风窗(ID:SouthReviews) 封面来源 ｜ unplash 曾在小红书上炫富1.67亿年薪的清华学霸吴舰，如今正面临两起诉讼。 南风窗 . 冷静地思考，热情地生活。 清华学霸涉欺诈遭美通缉， 1.67亿年薪致赔2.55亿。 以下文章来源于南风窗 ，作者陆茗 2025年9月， 一起来自美国纽约南区联邦地区法院发起的刑事诉讼，指控吴舰犯有电信欺诈罪、证券欺诈罪和洗钱罪；另一起来自美国证券交易委员会 （下称"SEC"）发起的民事诉讼，主张判决吴舰归还不义之财、支付罚款、终身禁业。 事件要追溯到2023年1月，小红书上出现一条题为《不敢发朋友圈的》的匿名帖称："工作5年赚到了刚毕业时候想不到的数字……想找个没人的地方偷偷 炫耀下。" 帖子配了一张摄屏照，显示发帖人2022年的薪资收入达到2350万美元，约合人民币1.67亿元，是上一年的10倍——发帖人称，原也想过此举可能被同事认 出来，但很快打消了顾虑，直言："感觉我也不需要那么在意。" 社交平台上的帖子显示， 发帖人2022年的薪资收入达到2350万美元，约合人民币1.67亿元，是上一年的10倍 戏剧化的是，SEC ...

Report Overview - The report is a weekly update on public - offering funds, specifically focusing on the tracking of ETF strategy indices as of September 22, 2025 [1] Core Viewpoints - By leveraging ETFs, it is convenient to transform quantitative models or subjective views into practical investment strategies. The report presents several ETF - based strategy indices and tracks their performance and positions on a weekly basis [12] Index - Specific Summaries 1. ETF Strategy Index Tracking - **Overall Performance**: The table shows the performance of various ETF strategy indices from September 12 - 19, 2025, including their weekly returns, benchmark returns, and excess returns [13] 1.1. Huabao Research Large - Small Cap Rotation ETF Strategy Index - **Strategy**: Utilizes multi - dimensional technical indicator factors and a machine - learning model to predict the return difference between the Shenwan Large - Cap Index and the Shenwan Small - Cap Index. It outputs weekly signals to determine positions and obtain excess returns. - **Performance**: As of September 19, 2025, the excess return since 2024 is 18.52%, the recent one - month excess return is - 0.48%, and the recent one - week excess return is - 0.17% [14] 1.2. Huabao Research SmartBeta Enhanced ETF Strategy Index - **Strategy**: Uses price - volume indicators to time self - built Barra factors and maps timing signals to ETFs based on their exposures to 9 major Barra factors to outperform the market. - **Performance**: As of September 19, 2025, the excess return since 2024 is 17.22%, the recent one - month excess return is 1.25%, and the recent one - week excess return is 0.83% [18] 1.3. Huabao Research Quantitative Windmill ETF Strategy Index - **Strategy**: Adopts a multi - factor approach, including long - and medium - term fundamental analysis, short - term market trend tracking, and analysis of market participants' behaviors. It uses valuation and crowding signals to indicate industry risks and identify potential sectors. - **Performance**: As of September 19, 2025, the excess return since 2024 is 24.66%, the recent one - month excess return is 8.37%, and the recent one - week excess return is 1.07% [22] 1.4. Huabao Research Quantitative Balance ETF Strategy Index - **Strategy**: Employs a multi - factor system covering economic fundamentals, liquidity, technical aspects, and investor behavior to build a quantitative timing system for equity market trend analysis. It also predicts the large - small cap style to adjust equity market positions. - **Performance**: As of September 19, 2025, the excess return since 2024 is - 9.37%, the recent one - month excess return is - 2.00%, and the recent one - week excess return is 0.28% [27] 1.5. Huabao Research Hot - Spot Tracking ETF Strategy Index - **Strategy**: Tracks market sentiment, major industry events, investor sentiment, professional opinions, policy changes, and historical trends to identify hot - spot index products and build an ETF portfolio to capture market trends. - **Performance**: As of September 19, 2025, the recent one - month excess return is 0.59%, and the recent one - week excess return is - 1.88% [30] 1.6. Huabao Research Bond ETF Duration Strategy Index - **Strategy**: Uses bond market liquidity and price - volume indicators to select effective timing factors and predicts bond yields through machine learning. It reduces long - duration positions when expected yields are below a certain threshold. - **Performance**: As of September 19, 2025, the recent one - month excess return is 0.34%, and the recent one - week excess return is 0.04% [33]

招商基金在指数增强领域厚积薄发，取得较好超额回报，场内交易的中证2000增强ETF近一年超23.36% 的超额收益；招商中证1000指数增强基金成立至今8年多时间，累计超额收益高达114.47%，位居指数 增强基金同期表现前列。 当前，伴随市场回暖，指数化投资迎来高速发展期。作为向上涨场环境下的优质配置工具，指数增强基 金成为投资者捕捉赛道机遇、优化资产组合的重要引擎。 这份可观的业绩离不开招商基金量化团队的深厚底蕴与不懈创新，团队在传统的动量因子之上，引入了 更多基本面分析，兼具主动管理和量化的优势，追求相对于传统量化策略更明确的长期投资回报。 多点开花，过往持续超额回报彰显实力 A股市场企稳反弹，宽基指数普遍上涨。指数增强基金亮眼依旧，其在弱市中控回撤的韧性，与上涨市 中创超额的能力，构成了其相较于被动指数的优势。 指数增强基金是指在跟踪标的指数的基础上，通过量化策略积极获取超额收益，为投资者提供一种既追 求享受市场整体增长回报，又能争取超越市场平均水平的投资途径。指数增强基金不仅能瞄准指数的 Beta收益，还争取通过主动管理收获Alpha收益，在市场回暖的当下受到不少投资者追捧。 招商基金在8年前就前瞻 ...

Core Viewpoint - The article discusses the role of quantitative models in enhancing index funds, emphasizing the importance of multi-factor models in evaluating and selecting stocks for better performance [2][4][8]. Group 1: Multi-Factor Models - Factors are described as "labels" or "features" that characterize stocks, similar to how one might describe a person [2]. - The multi-factor model considers various factors to assess stock value and risk, with a library of hundreds of factors used for evaluation [2][3]. - Factors are categorized into fundamental and quantitative types, focusing on deep logic and statistical patterns, respectively [2]. Group 2: Process of Enhancement - The enhancement process involves three main stages: individual stock analysis, portfolio generation, and daily operations [4][5][6]. - In the individual stock analysis stage, stocks are scored based on various factors, leading to a ranking [4]. - The portfolio generation stage involves selecting stocks that meet specific criteria and assigning them different weightings based on their scores [5]. - Daily operations focus on optimizing trades to minimize costs and monitoring portfolio performance [7]. Group 3: Advantages of Quantitative Models - Quantitative models are positioned as ideal partners for index enhancement, providing stability, precision, and efficiency in stock selection [8]. - The use of quantitative strategies helps to maintain consistency in investment management, reducing emotional influences [8]. - Quantitative models can dynamically update to ensure effectiveness, allowing for timely adjustments in the portfolio [8].

本文首发于公众号"私募排排网"。 （点击↑↑ 上图订阅专栏 ） 编者按 通常，投资者在了解私募时，会关注公司团队水平、策略运作、中长期业绩、风险控制等内容，为此，私募排排网推出# 深度揭秘100家私募 #栏目 ，对这些内容进行详解。 本期揭秘的是 百亿量化私募鸣石基金 。据私募排排网截至7月底数据显示， 鸣石基金在量化多头近一年收益TOP10榜单中以***%收益夺得第 一（ 点此看收益 ） ，公司旗下的量化选股产品"鸣石春天28号"则在近三年超额收益位居头部私募量化选股10强的第3名，该产品近三年超额 收益为***%。 （ 点此查看基金收益 ） 下面，就让我们一起来揭秘鸣石基金取得优秀业绩的核心原因。 | 排名 | 产品简称 | 公司简称 | 公司规模 | 基金经理 | 产品规模 | 今年来 今年来近3年 | 沂3年 | | --- | --- | --- | --- | --- | --- | --- | --- | | | | | | | (万元) | 收益 超额 收益 | 超额 | | 7 | 阿巴马乘风破浪A类 份额 | | 阿巴马投资 100亿以上 | 震海道 | | | | | 2 | 明泫股票精 ...

□本报记者 王雪青 2025年，量化持续成为市场"关键词"。伴随着A股市场震荡上行，单日两万亿元成交额已成常态，量化 产品备案数量同比翻倍，私募江湖风起云涌。平方和投资成立已有十周年，这十年历程，既是行业风雨 的缩影，也是其自身穿越周期的注脚。 回望2015年成立之初，外界对量化投资是否适合中国市场仍存疑问；十年来，平方和投资不仅在策略上 实现了"策略十年、十年长青"，更在业绩方面交出了扎实的成绩单，率先验证了量化方法论在中国市场 的长期有效性。 近日，中国证券报记者专访了平方和投资创始人、总经理吕杰勇。在他位于中关村的办公室，这位亲历 中国量化从萌芽到壮大的资深投资者，分享了他一路走来的成长与感悟。 初心如磐：深耕中国市场 一切的起点源于1999年。这一年，吕杰勇考入北京大学数学科学学院。在这里，他不仅被数学的纯粹与 理性深深吸引，更从知名量化投资者詹姆斯·西蒙斯的实践中获得启发：数学不仅能在公式推导中展现 优势，也可以成为优化金融市场效率的工具。一颗量化投资的种子，就此在他心中扎根。 "我始终觉得，量化投资是科学的价值发现方法论，能让市场更有效率，而且衍生品工具在中国市场肯 定是未来的大方向。"站在公司十 ...

民生加银基金周帅： 运用量化工具 打造专精特新投资利器 ◎记者 陈玥 过去两年，当"专精特新"成为市场一大热门主题的同时，如何从一个边界并不清晰的概念中挖掘超额收 益，对基金经理的投资能力提出了考验。经过一系列整合和摸索，周帅打造出了一套"攻略"。 "当前沪深两市的国家级'专精特新企业'数量超1000家，大部分企业市值在100亿元以下；在板块分布 上，近80%企业属于创业板、科创板等科技成长板块；行业主题上分布在新一代信息技术、高端装备制 造、新材料、节能环保、生物产业、新能源等战略性新兴行业。企业数量多、行业主题跨度大、市值相 对较小，较为适合我们通过量化的方式去挖掘主题内的阿尔法收益。在投资策略上，我们将量化策略与 基本面研究相结合，去发掘市场中的投资机会。"周帅说。 周帅表示，通过对专精特新企业风格特征的分析，表明其具备较为清晰的小盘、成长、科技主题特征， 因此从这两个方向去寻找超额收益，最终搭建了小盘价值、科技成长优选两套子策略。 对于科技成长优选策略，周帅表示，主要从行业逻辑出发选择科技成长板块长期选股效果更佳的基本面 因子，同时纳入高频量价及机器学习类因子，单独搭建科技成长板块选股策略。小盘价值策 ...

第二批科创债 ETF 集中成立，A500 增强工具持续扩容 —— 量化市场追踪周报（2025W38） [Table_ReportTime] 2025 年 9 月 21 日 请阅读最后一页免责声明及信息披露 http://www.cindasc.com 1 [Table_ReportType] 金工点评报告 [Table_Author] 于明明 金融工程与金融产品 首席分析师 执业编号：S1500521070001 联系电话：+86 18616021459 邮 箱：yumingming@cindasc.com 吴彦锦 金融工程与金融产品 分析师 执业编号：S1500523090002 联系电话：+86 18616819227 邮 箱：wuyanjin@cindasc.com 周君睿 金融工程与金融产品 分析师 执业编号：S1500523110005 联系电话：+86 19821223545 邮 箱：zhoujunrui@cindasc.com [Table_Title] 量化市场追踪周报(2025W38)：第二批科创债 ETF 集中成立，A500 增强工具持续扩容 证券研究报告 金工研究 [Table_Repo ...

Quantitative Models and Construction Methods Model Name: Three-Dimensional Timing Model - **Model Construction Idea**: The model uses three dimensions: liquidity, divergence, and prosperity to judge market trends[8] - **Model Construction Process**: - The model evaluates the current liquidity trend, market divergence, and prosperity level - It uses technical indicators to assess the market status, such as the overbought condition of the CSI 300 index[8] - The model's historical performance is visualized to validate its effectiveness[17] - **Model Evaluation**: The model indicates a downward trend in a high turnover market, suggesting a low probability of short-term upward movement[8] Model Name: ETF Hot Trend Strategy - **Model Construction Idea**: The strategy selects ETFs based on their price trends and market attention[28] - **Model Construction Process**: - Identify ETFs with both highest and lowest price trends using K-line highest and lowest price shapes - Construct support and resistance factors based on the relative steepness of the regression coefficients of the highest and lowest prices over the past 20 days - Select the top 10 ETFs with the highest turnover rate in the past 5 and 20 days to form a risk parity portfolio[28] - **Model Evaluation**: The strategy includes ETFs from semiconductor, non-ferrous metals, 5G communication, battery industries, and growth styles[29] Model Name: Capital Flow Resonance Strategy - **Model Construction Idea**: The strategy monitors the resonance of margin trading and large order funds to select favored industries[32] - **Model Construction Process**: - Define the margin trading capital factor as the net buying of financing minus the net selling of securities lending, neutralized by the Barra market value factor - Define the active large order capital factor as the net inflow of the industry, neutralized by the time series of trading volume over the past year - Combine the two factors to construct the strategy, excluding extreme industries and large financial sectors to improve stability[35] - **Model Evaluation**: The strategy has shown stable positive excess returns since 2018, with an annualized excess return of 13.5% and an IR of 1.7[35] Model Backtesting Results - **Three-Dimensional Timing Model**: Historical performance shows a consistent downward trend in high turnover markets[17] - **ETF Hot Trend Strategy**: The strategy has achieved cumulative excess returns over the CSI 300 index this year[30] - **Capital Flow Resonance Strategy**: The strategy recorded a negative excess return last week, with an absolute return of -2.4% and an excess return of -2.0% relative to the industry equal weight[35] Quantitative Factors and Construction Methods Factor Name: Beta Factor - **Factor Construction Idea**: Measures the sensitivity of a stock's returns to market returns[40] - **Factor Construction Process**: - Calculate the beta coefficient of each stock based on its historical returns relative to the market index - Form portfolios of high and low beta stocks to compare their performance[40] - **Factor Evaluation**: High beta stocks significantly outperformed low beta stocks, recording a positive return of 2.19% last week[40] Factor Name: Growth Factor - **Factor Construction Idea**: Measures the growth potential of stocks based on their earnings and revenue growth[40] - **Factor Construction Process**: - Calculate the growth rate of earnings and revenue for each stock - Form portfolios of high and low growth stocks to compare their performance[40] - **Factor Evaluation**: Growth stocks continued to outperform value stocks, with the growth factor achieving a return of 1.51% last week[40] Factor Backtesting Results - **Beta Factor**: - Year-to-date: 26.61% - Last month: 2.39% - Last week: 2.19%[41] - **Growth Factor**: - Year-to-date: -0.44% - Last month: 4.74% - Last week: 1.51%[41]