Core Viewpoint - The article highlights the growing popularity and impressive performance of quantitative private equity funds, particularly focusing on Ming Stone Fund, which has established itself as a leading player in the industry since its inception in 2010 [1][3]. Group 1: Company Overview - Ming Stone Fund was founded in December 2010 and currently employs over 100 staff globally, with more than 80% of the research team holding advanced degrees from prestigious universities [1]. - The founder and general manager, Dr. Yuan Yu, has a strong academic background, having obtained a PhD in Finance from the Wharton School and previously worked at the Federal Reserve Bank [3][4]. Group 2: Investment Strategy - The fund has developed a unique three-factor model tailored to the Chinese market, which includes market, size, and value factors, effectively explaining most cross-sectional return anomalies in the A-share market [13]. - Ming Stone Fund employs a proprietary "Five-Ring Multi-Core" quantitative research and investment system, which includes five key research stages: factor, AI, optimization, risk control, and trading [9][14]. Group 3: AI Integration - Since 2021, the fund has significantly increased the role of AI in its investment research, establishing the G-Lab AI laboratory to enhance efficiency across all research stages [14]. - The fund's factor library consists of 30,000 factors, primarily derived from manual research, supplemented by machine learning, ensuring a balance between interpretability and differentiation [15]. Group 4: Performance Metrics - Ming Stone Fund's quantitative stock selection product, "Ming Stone Spring 28," ranked third among top private equity quantitative stock selection products in terms of excess returns over the past three years [16]. - The fund attributes its strong performance to its efficient research system, adaptive strategies, and the favorable market environment, which has enhanced its ability to capture liquidity premiums [18]. Group 5: Market Insights - The article discusses the potential of small-cap stocks represented by the CSI 1000 index, which is expected to perform well due to its low institutional coverage and pricing inefficiencies [20]. - The fund emphasizes the importance of risk control, utilizing a self-developed multi-factor risk control model to manage volatility and exposure effectively [21].

（原标题：当A股遇上AI，股神的诞生？） 21世纪经济报道记者 章驰 沪指创十年新高，但比大盘更疯的是，朋友圈的AI股神！ AI模型和技术虽然可以用于金融数据分析，但它具体有没有构成违规荐股及投顾服务，甚至一旦AI大 模型涉及投资领域的合规问题时，责任权属该如何界定？这些问题目前法律界还没有统一的定论，责任 人类炒股太容易受贪婪与恐惧影响，而AI完全基于数据逻辑操作。可以说在这个赛道里，AI真是赢麻 了！ 实际上，不少大型投资机构早已开始将AI纳入投资决策流程。今年初爆火的DeepSeek，本身就是国内 头部量化资管幻方量化旗下公司的AI大模型。国内多家百亿级量化私募多年以前就已经纷纷开始布局 AI领域，量化投资的技术迭代基本上与AI的技术迭代同步，抢占AI高地已成为共识。 目前，多数券商平台也纷纷接入DeepSeek部署，加码AI功能。股民们用的软件基本都有一个AI入口， 所以AI其实已经渗透到了广大股民的投资过程中。 不过专业人士们都知道，股市是一个复杂且不确定的系统，并不具备规律性，单纯靠AI来战胜市场是 很难的。尽管有了AI，还是必须"以人为主，机器为辅"，把AI作为辅助工具来提高信息获取、分析和 决 ...

Core Viewpoint - The current A-share market is experiencing a rapid rotation of styles, leading investors to focus on index-enhanced products to achieve stable excess returns [1] Group 1: Market Structure and Investment Strategy - The market is transitioning from a "barbell" structure to a more balanced recovery, with funds beginning to flow back to mid-cap stocks, which have solid fundamentals and reasonable valuations [1] - The China Securities A500 Index is highlighted for its dual advantages in market capitalization style and alpha space, making it a favorable path for capturing excess returns [1] - Approximately 70% of the A500 index constituents overlap with the CSI 300 index, providing stability, while the remaining 30% includes more growth-oriented sectors like TMT, offering higher growth potential [1][2] Group 2: Factor Selection and Risk Management - The investment strategy employs a combination of 60% fundamental factors and 40% price-volume factors, focusing on fundamental momentum and turnaround situations [2] - The risk management approach has evolved from static to dynamic, allowing for adjustments based on individual stock liquidity, estimated impact costs, and fundamental strength [2] - The strategy emphasizes a disciplined and systematic approach to quantitative investment, prioritizing long-term, reusable frameworks over short-term gains [4]

Core Viewpoint - The current A-share market is experiencing rapid style rotation, leading investors to focus on index-enhanced products to achieve stable excess returns [1] Group 1: Market Structure and Investment Strategy - The market has shown a "barbell structure" over the past two years, with funds concentrated in low-valuation large-cap blue chips and small-cap stocks, while mid-cap stocks have been under pressure [2] - As the policy environment improves and risk appetite rises, funds are beginning to flow back into mid-cap stocks, creating a potential allocation window for companies with solid fundamentals and reasonable valuations [2] - The introduction of the CSI A500 index enhancement product aims to balance win rates and odds, targeting more attractive investment opportunities to enhance cost-effectiveness [2] Group 2: Characteristics of CSI A500 Index - Approximately 70% of the CSI A500 index constituents overlap with the CSI 300 index, providing stability in fundamentals, while about 30% are closer to the CSI 500 index, featuring higher growth potential in emerging industries [2] - This balanced structure offers a good equilibrium between safety margins and upside potential [2] Group 3: Factor Selection and Dynamic Risk Control - The quantitative team employs a systematic model to achieve sustainable excess returns, focusing on factor selection and dynamic risk control [3] - The strategy combines 60% fundamental factors and 40% price-volume factors, with an emphasis on fundamental momentum and turnaround factors [3] - Dynamic risk control measures are tailored based on individual stock liquidity, estimated impact costs, sentiment, and fundamental strength, allowing for more precise risk management [4] Group 4: Market Performance and Strategy Adaptation - Recent market performance has seen extreme institutional clustering in certain technology sectors, increasing the demands on quantitative strategies to generate excess returns [5] - The CSI A500 index enhancement strategy will maintain a balanced exposure without extreme bets on style, focusing on stability in win rates and drawdown management [6] - The quantitative investment philosophy emphasizes discipline and systematic approaches, aiming to enhance strategy effectiveness through detailed optimization [6]

在当下A股市场风格快速轮动的背景下，如何通过指数增强产品争取稳定超额收益，成为投资者关注的 焦点。 中信保诚基金量化基金经理姜鹏认为，当前，市场正处于"哑铃行情"向均衡修复的阶段，中证A500指 数在市值风格与阿尔法空间上具备双重优势，叠加精细化的量化模型与动态风控，有望成为捕捉超额收 益的更优路径。 "变化正在发生。"姜鹏敏锐观察，"随着政策环境改善、风险偏好回升，我们可以看到资金开始从哑铃 的两端向中段回流。基本面扎实、估值合理且尚未被充分挖掘的上市公司，或迎来配置窗口"。 他强调，选择在此时推出中证A500指增产品，核心在于胜率与赔率的平衡，"我们希望去投资更具吸引 力的标的，提升投资的性价比。当前环境下，个人认为中证A500指数契合这一特征。" 具体到中证A500指数的特性，姜鹏表示："从截至7月末的指数成份股来看，中证A500指数成份股中， 约70%与沪深300指数重合，这部分提供了基本面的稳定性；另外约30%则更接近中证500指数的特色， 涵盖较多TMT等新兴产业，具备更高成长弹性。这种稳中有进的结构，使其在安全边际与向上空间之 间形成良好的均衡。" 精细化因子挖掘 与动态风控 在指数底座之上，中 ...

- The report does not contain any quantitative models or factors for analysis[3][4][5]

Group 1 - The report highlights the performance of various quantitative investment strategies, with the "Excellent Fund Performance Enhancement Portfolio" achieving a weekly absolute return of 4.17% and a year-to-date return of 22.11% [1][21] - The "Super Expected Selection Portfolio" reported a weekly return of 3.71% and a year-to-date return of 39.73%, outperforming the mixed equity fund index by 15.01% this year [1][29] - The "Brokerage Golden Stock Performance Enhancement Portfolio" had a weekly return of 3.94% and a year-to-date return of 27.90%, exceeding the mixed equity fund index by 3.18% [1][36] - The "Growth and Stability Portfolio" achieved a weekly return of 3.38% and a year-to-date return of 45.63%, outperforming the mixed equity fund index by 20.91% [2][41] Group 2 - The "Excellent Fund Performance Enhancement Portfolio" is constructed by benchmarking against active equity funds rather than broad market indices, utilizing quantitative methods to enhance performance [3][17] - The "Super Expected Selection Portfolio" is built by screening stocks based on expected performance and analyst profit revisions, focusing on both fundamental and technical aspects [4][22] - The "Brokerage Golden Stock Performance Enhancement Portfolio" leverages a stock pool from brokerage recommendations, optimizing the portfolio to align with the performance of the brokerage stock pool [5][30] - The "Growth and Stability Portfolio" employs a two-dimensional evaluation system for growth stocks, prioritizing those closer to their earnings report dates to capture potential excess returns [6][62] Group 3 - The report provides a comprehensive performance overview of the portfolios, indicating that the "Excellent Fund Performance Enhancement Portfolio" ranks in the 49.21 percentile among active equity funds this year [1][21] - The "Super Expected Selection Portfolio" ranks in the 14.36 percentile among active equity funds, demonstrating strong performance relative to peers [1][29] - The "Brokerage Golden Stock Performance Enhancement Portfolio" ranks in the 33.90 percentile among active equity funds, indicating competitive performance [1][36] - The "Growth and Stability Portfolio" ranks in the 8.91 percentile among active equity funds, showcasing its effectiveness in capturing growth opportunities [2][41]

Quantitative Models and Construction PB-ROE-50 Model - **Model Name**: PB-ROE-50 - **Model Construction Idea**: The model combines Price-to-Book (PB) and Return on Equity (ROE) metrics to identify stocks with strong profitability and reasonable valuation[24] - **Model Construction Process**: The PB-ROE-50 portfolio is constructed by selecting 50 stocks based on a combination of PB and ROE metrics. The portfolio is rebalanced periodically to maintain the desired exposure to these factors[24] - **Model Evaluation**: The model demonstrates consistent positive excess returns across different stock pools, indicating its effectiveness in capturing value and profitability signals[24] --- Quantitative Factors and Construction Standardized Unexpected Earnings (SUE) - **Factor Name**: Standardized Unexpected Earnings (SUE) - **Factor Construction Idea**: Measures the deviation of actual earnings from expected earnings, standardized by historical earnings volatility[12] - **Factor Construction Process**: $ SUE = \frac{E_{actual} - E_{expected}}{\sigma_{earnings}} $ Where: $E_{actual}$ = Actual earnings $E_{expected}$ = Expected earnings $\sigma_{earnings}$ = Standard deviation of historical earnings[12] - **Factor Evaluation**: Demonstrates strong positive returns in the CSI 300 stock pool, indicating its ability to capture earnings surprises effectively[12] Single-Quarter ROE YoY Growth - **Factor Name**: Single-Quarter ROE YoY Growth - **Factor Construction Idea**: Measures the year-over-year growth in Return on Equity (ROE) for a single quarter[14] - **Factor Construction Process**: $ ROE_{YoY} = \frac{ROE_{current\_quarter} - ROE_{same\_quarter\_last\_year}}{ROE_{same\_quarter\_last\_year}} $ Where: $ROE_{current\_quarter}$ = ROE for the current quarter $ROE_{same\_quarter\_last\_year}$ = ROE for the same quarter in the previous year[14] - **Factor Evaluation**: Shows strong positive returns in the CSI 500 stock pool, highlighting its effectiveness in identifying growth trends[14] Total Asset Growth Rate - **Factor Name**: Total Asset Growth Rate - **Factor Construction Idea**: Measures the growth rate of total assets over a specific period[16] - **Factor Construction Process**: $ Growth_{assets} = \frac{Assets_{current} - Assets_{previous}}{Assets_{previous}} $ Where: $Assets_{current}$ = Total assets in the current period $Assets_{previous}$ = Total assets in the previous period[16] - **Factor Evaluation**: Demonstrates strong positive returns across multiple stock pools, indicating its robustness in capturing growth signals[16] --- Backtesting Results of Models PB-ROE-50 Model - **Excess Return (Weekly)**: - CSI 500: 0.47% - CSI 800: 0.25% - All Market: 1.02%[25] - **Excess Return (YTD)**: - CSI 500: 3.22% - CSI 800: 11.76% - All Market: 14.28%[25] --- Backtesting Results of Factors Standardized Unexpected Earnings (SUE) - **Excess Return (Weekly)**: - CSI 300: 4.12% - CSI 500: 0.34% - Liquidity 1500: 1.16%[12][15][17] Single-Quarter ROE YoY Growth - **Excess Return (Weekly)**: - CSI 300: 0.84% - CSI 500: 2.28% - Liquidity 1500: 1.27%[12][15][17] Total Asset Growth Rate - **Excess Return (Weekly)**: - CSI 300: 2.39% - CSI 500: 0.50% - Liquidity 1500: 2.12%[12][15][17]

Group 1 - The core viewpoint of the news is that the comments made by Fed Chair candidate Brad regarding a potential 100 basis point rate cut have significantly influenced global markets, particularly causing a surge in A-shares, which reflects heightened expectations for interest rate cuts [1][3]. - The phenomenon of "buy the rumor, sell the news" is highlighted, indicating that true market opportunities often slip away by the time news is widely disseminated, especially in the A-share market where participants tend to act preemptively [3][15]. - The article emphasizes that institutional investors do not wait for news releases to enter the market, as evidenced by the early movements in oil stocks during the 2025 Israel-Iran conflict, suggesting that significant price movements often precede major news events [3][5]. Group 2 - Behavioral finance principles suggest that irrational behaviors among market participants create specific patterns, with institutional investors often accumulating positions through small trades to avoid drawing attention [7][15]. - The article points out that certain stocks exhibit a common characteristic of having institutional activity prior to significant news, indicating a "preparatory" state that is more valuable than the news itself [15]. - The focus is on identifying which assets have shown unusual fund movements before the Fed takes action, highlighting the importance of data analysis in understanding current market dynamics rather than merely predicting future events [15].

2025 年 08 月 22 日 国联量化团队：传统量化与 AI 投资 并行，力争实现可持续的超额收益 ——基金经理研究系列报告之七十八 本研究报告仅通过邮件提供给 中庚基金 使用。1 股 票 基 证 券 研 究 报 告 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 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 金 相关研究 - ⚫ 国联基金多策略投资部（国联量化团队）是公司量化投资业务的核心团队，覆盖指数及 ETF、 指数增强、主动量化（对冲与 Smart Beta）、FOF、固收+等多个条线。国联量化团队的核 心目标是在市场有效性增强与传统公开因子边际收益递减的背景下实现可持续的超额收 益。为此，团队打造了依托"高频指增+深度学习"的非线性框架"，以期持续获取稳定的 Alpha，并在定 ...