Core Viewpoint - The rise of AI-driven quantitative strategies is transforming investment approaches, allowing for the identification of complex relationships in data that traditional methods cannot capture [1][4]. Group 1: AI Quantitative Strategies - AI quantitative models can analyze hundreds of high-frequency signals, uncovering non-linear relationships in data, which enhances predictive accuracy compared to traditional models that rely on a limited number of factors [1][7]. - The AI quantitative strategy developed by Swiss Bank Asset Management focuses on around 400 high-frequency signals, contrasting with the typical 20 signals used in traditional quantitative strategies [7]. - The AI model's ability to learn complex relationships allows it to improve the prediction of stock price movements based on analyst ratings and other signals [6][8]. Group 2: Market Expansion and Interest - Global capital interest in the Chinese market is on the rise, with plans to include A-shares in AI quantitative strategies as they expand into emerging markets [4][5]. - The firm is currently exploring the potential of AI-driven strategies for domestic investors in China, contingent on obtaining additional QDLP quotas [5][6]. Group 3: Investment Strategy and Risk Management - The investment horizon for Swiss Bank Asset Management's AI strategies is approximately 20 days, differing from many competitors that focus on ultra-short holding periods [8]. - To mitigate overfitting risks, the firm employs methods such as using economically sound signals, integrating numerous simple models, and utilizing extensive historical data for training [8]. Group 4: Role of Fund Managers - The role of fund managers is evolving with the integration of AI, shifting from model building to training AI models and validating their outputs while still conducting factor research [8].

雷德玮（David Wright），现任瑞士百达资产管理量化投资主管，负责Quest（量化股票和解决方案）以 及指数股票业务。他于2022年加入瑞士百达资产管理。此前，他曾担任量化业务策略Quest联席主管。 在加入瑞士百达之前，雷德玮在贝莱德（2009年之前为巴克莱全球投资）工作了22年，后来担任系统化 主动股票（SAE）业务的欧洲、中东和非洲地区（EMEA）产品策略主管以及SAE ESG产品的首席产品 策略师。 "当一名分析师上调股票评级时，是否真正预示股价即将上涨？对此，AI量化模型可以更精准地判 断。"瑞士百达资管量化投资主管雷德玮近期在接受中国证券报记者专访时表示，AI通过算力提升和开 源工具普及，正在推动量化投资进入2.0时代。传统量化局限于价值、动量等少量因子进行分析，AI量 化可以识别数百个高频信号，挖掘数据中的非线性关系。 瑞士百达资管为瑞士百达集团旗下的机构资产管理和基金管理公司，瑞士百达集团拥有220年历史，截 至2025年6月30日，集团资产管理规模达7110亿瑞士法郎。目前，雷德玮带领的瑞士百达资管量化股票 及解决方案投资团队的管理规模为250亿美元，旗下AI量化策略产品已在发达市场 ...

Core Insights - The article discusses the advancements in AI-driven quantitative investment strategies led by David Wright at Swiss Bank Asset Management, highlighting the transition to a 2.0 era of quantitative investing through enhanced computational power and open-source tools [1][2]. Group 1: AI Quantitative Strategies - AI quantitative models can identify hundreds of high-frequency signals and uncover non-linear relationships in data, surpassing traditional quantitative methods that rely on a limited number of factors [1][5]. - The AI quantitative strategy team at Swiss Bank Asset Management manages $25 billion, with plans to expand into emerging markets, including A-shares in China [2][3]. - The interest of global capital in the Chinese market is on the rise, with potential AI quantitative strategies targeting A-shares expected to launch next year [2][3]. Group 2: Market Adaptation and Performance - AI models have shown that the signal relationships identified can be generalized across countries, indicating that these models can be adapted to the Chinese market [3][4]. - Emerging markets may offer higher potential excess returns compared to developed markets, although trading costs are also higher, leading to similar overall excess returns relative to benchmarks [3][4]. - The integration of fundamental signals alongside emotional and price signals in emerging markets has been found to enhance model performance [3][4]. Group 3: Differentiation and Risk Management - Swiss Bank Asset Management's AI quantitative strategy focuses on a holding period of approximately 20 days, contrasting with many competitors that prefer shorter holding periods [5][6]. - The firm emphasizes the use of traditional data for model training, covering longer historical periods, and maintaining factor neutrality across various investment styles [5][6]. - To mitigate overfitting risks, the company employs economically sound signals, integrates numerous simple models for training, and utilizes a cross-validation method with 15 years of data [6]. Group 4: Evolving Role of Fund Managers - The role of fund managers is evolving with the rise of AI in quantitative investing, shifting from model building to training AI models and validating outputs [6]. - Fund managers will continue to conduct factor research and oversee investment portfolio construction, maintaining the same number of personnel despite changes in responsibilities [6].

在公募量化产品线中，除了指数增强型品种外，量化选股型品种也占据重要地位，截止 2025 年二季末，全市场 共计 277 只量化选股型基金，合计管理规模 903.20 亿元。这类基金由于不属于标准指数增强范畴，因而，不受限于 指数成分股的投资比例和跟踪误差的硬性约束，拥有更广泛的投资范围和更高的风格暴露自由度，其业绩往往也展现 出更高的弹性。同时，得益于相对宽松的策略环境，不同基金经理能够根据自身的偏好及特长，构建更具差异化的投 资策略。 然而，对于投资者而言，由于量化选股型基金的工具化属性不如指数增强型基金明确，在产品选择时，往往面临 对其定位与策略认知不够清晰的挑战。为此，本篇专题将对部分量化选股型基金的投资策略框架进行梳理，主要聚焦 信达澳亚基金冯玺祥、国泰基金高崇南、信达澳亚基金林景艺、鹏华基金时赟超、西部利得基金翟梓舰等 5 名在量化 选股型产品上投资框架体系各有特色的基金经理（按照姓名拼音字母顺序，下同），涵盖风格配置、中小市值、偏股 混增强等差异化产品定位，以供投资者参考。 冯玺祥（信达澳亚基金）：采用统一框架管理，十分看重因子在整个投资域的有效性以及阿尔法模型的普适性， 通过多样化的因子和模型提 ...

1. Report Industry Investment Rating There is no information provided regarding the industry investment rating in the given content. 2. Core Viewpoints of the Report - The report focuses on the analysis of fund manager Yang Lin from Minsheng Jiachen Fund. Yang Lin has over 13 years of experience in the securities industry and about 1.6 years as an investment manager. His investment framework is a multi - dimensional AI quantitative investment framework that continuously iterates and evolves. He makes full use of various information, constructs factors through both manual and machine - learning methods, and applies AI technology in multiple dimensions [3][7][9]. - The representative product, Minsheng Jiachen Smart Selection Growth, has shown excellent performance in the active equity category. Since its establishment until June 27, 2025, it has achieved a return of 27.94%, ranking in the top 20% of active equity products. It also has a relatively low volatility and a high return - risk ratio [3][29]. - The product has distinct investment characteristics. Its overall position is similar to the combination of CSI 500 and CSI 1000, with low individual - stock concentration, small industry over - or under - weighting, and a focus on small - and medium - cap stocks without micro - cap stocks. Trading turnover may be the main source of its income [3][43][53]. 3. Summary Based on the Table of Contents 3.1 Fund Manager Overview - Yang Lin is a CFA, with a bachelor's degree in finance from Xi'an Jiaotong University and a master's degree in financial mathematics from the University of Warwick. He has held various positions in different financial institutions and joined Minsheng Jiachen Fund in October 2019. Currently, he manages one product. His fund manager index lagged slightly behind the CSI 300 before September 2024 due to the small - cap style of the products he managed but has significantly outperformed the CSI 300 since then [3][7]. 3.2 Investment Framework: A Continuously Iterating and Evolving Multi - dimensional AI Quantitative Investment Framework - Yang Lin uses multiple dimensions of information to create effective investment factors. He constructs factors through both manual and machine - learning methods. Manual factors include fundamental (about 20%) and price - volume factors (about 40%), mainly for longer - term factors, while machine - learning processes relatively short - term price - volume factors (about 40%). The ratio of long/medium/short/ultra - short - term factors is about 2:2:3:3 [9][11]. - AI technology is applied in multiple dimensions of his investment framework, such as in ultra - short - term factor construction, manual stock - selection, and comprehensive data processing [11]. - Yang Lin believes that continuous iteration is crucial for maintaining competitiveness. He iterates his models in three dimensions: data, model, and training. He has iterated his quantitative strategies multiple times in the past few years, with an optimization frequency of 1 - 2 times a year, and only implements new strategies in real - time after verifying their superiority [14][15]. 3.3 Representative Product: Minsheng Jiachen Smart Selection Growth - The product was established in June 2024, managed by Cai Xiao and Yang Lin. Its goal is to use a quantitative investment model to achieve long - term stable appreciation of fund assets while controlling risks. It has a management fee rate of 1.20% and a custody fee rate of 0.20% [18]. - In actual operation, it uses a stock strategy of "50% weight of CSI 500 - like all - market enhancement + 50% weight of CSI 1000 - like all - market enhancement". Its holdings are consistent with this positioning, with about half of the holdings being CSI 500 component stocks and about 40% being CSI 1000 component stocks [20]. - Due to the adoption of the high - order PB - ROE framework, the product shows a significant low - PB characteristic in its holdings. Its weighted average PB in 24H2 full - position and 25Q1 heavy - position stocks is in the lower 9.1% and 1.8% levels respectively among active equity products, while ROE is around the 20% quantile [21]. 3.4 Performance Analysis of Minsheng Jiachen Smart Selection Growth - **Performance in the Same Category**: The product's return since its establishment (as of June 27, 2025) is 27.94%, ranking in the top 20% of active equity products. Its annualized volatility is 24.80%, in the lower - middle level among active equity products. Its annualized Sharpe ratio is 1.18, and the Calmar ratio is 2.40, both ranking around the top 10% [29]. - **Relative Performance**: Compared with the composite index of "45% CSI 500 + 45% CSI 1000+10% ChinaBond New Composite Wealth (Total Value)", the product has significantly outperformed the composite index since November 2024. From October 2024 to June 2025, its monthly winning rate is 77.8%, and the average monthly excess return is 0.34% [33][37]. 3.5 Investment Characteristic Analysis of Minsheng Jiachen Smart Selection Growth - **Positioning Characteristics**: The product has a highly diversified individual - stock position. The proportion of the top ten stocks in 24H2 full - position is only about 6%, and the top thirty is about 17%. It focuses on small - and medium - cap stocks without micro - cap stocks, strictly controlling non - linear market - value exposure [43][48]. - **Industry Distribution**: Compared with the industry distribution of CSI 500 and CSI 1000, the product slightly over - weights industries such as basic chemicals, machinery, and environmental protection, and slightly under - weights industries such as electronics, power equipment, and media, with a small overall over - or under - weighting range [49]. - **Income Source**: Using the Brinson model to split the fund's income, trading has made a significant contribution to the product's excess return since its management, while other income sources contribute less. The product's relatively high turnover rate suggests that trading turnover may be the main reason for its leading performance. The product's absolute income comes from various sectors, with financial real - estate and technological innovation contributing more, and it also has a strong ability to obtain relative income in these sectors [51][53].

Group 1 - In 2025, AI quantitative investment and the Federal Reserve's interest rate hikes are key variables reshaping investment logic in the financial markets [1] - AI technology has rapidly advanced in quantitative investment, with firms like Luminus Fund utilizing deep neural networks to extract market features from vast datasets [3] - Luminus Fund's quantitative simulation shows that over 70% of excess returns come from stock selection, highlighting AI's potential in enhancing returns through individual stock analysis [3] Group 2 - The persistent inflation in 2025, with core PCE inflation nearing 3% and CPI inflation expected to rise to 5.4%, increases pressure on the Federal Reserve to consider interest rate hikes [4] - Wall Street's betting on the likelihood of rate hikes has surged from under 10% to 34.6%, with predictions of a potential increase of 75 basis points from major financial institutions [4] - The evolving investment logic indicates that while traditional AI models may struggle with market volatility due to reliance on historical data, models that can adapt quickly may seize more opportunities [5] Group 3 - Different asset classes are affected differently by AI quantitative investment and Federal Reserve rate hikes, with notable divergence in the tech stock market [6] - Stocks like Intel surged by 16% due to market sentiment and AI-driven funds, while growth stocks like Meta and Netflix face challenges from anticipated rate hikes [6] - In the bond market, rising rates lead to falling bond prices, but AI models can optimize bond allocations across various maturities and credit ratings [6] Group 4 - The gold market is also impacted, with short-term dollar strength from rate hikes suppressing gold price increases, while AI quantitative investment can analyze multidimensional data to capture short-term price fluctuations [6] - Investors in 2025 must reassess their strategies, recognizing both the advantages and limitations of AI quantitative investment while closely monitoring Federal Reserve rate hike developments [7] - Adjusting asset allocations, such as increasing cash reserves and focusing on stable, cash-rich companies less affected by rate hikes, is essential for navigating the complex market environment [7]

Core Insights - The rise of AI in the quantitative private equity sector is highlighted by the recent achievements of domestic firms like NianKong Technology, which has made significant strides in AI applications for finance [1][2] - As of May 23, 2023, there are 39 domestic quantitative private equity firms managing over 10 billion yuan, with 15 of them actively engaging in AI-related investments [1] Group 1: Company Developments - NianKong Technology has collaborated with Shanghai Jiao Tong University to submit a research paper on large model training methodologies to the international NIPS conference [1] - The establishment of Shanghai QuanPin Siwei Artificial Intelligence Technology Co., Ltd. by NianKong Technology focuses on cutting-edge AI research [1] - NianKong Technology, founded in 2015, operates two quantitative private equity firms, NianJue Assets and NianKong Data Technology, and is recognized as an early adopter of AI in the financial sector [1] Group 2: Performance Metrics - NianKong Technology's quantitative products have shown impressive performance, with an average return of 21.50% over the past year across four products [1] - Specific products managed by NianKong's founder, Wang Xiao, achieved returns of 33.96% and 23.24% over the past year [1] - Among the 15 billion-yuan AI quantitative private equity firms, 13 have reported products with more than three performance records, yielding an average return of 29.91% over the past year [2] Group 3: Industry Trends - The integration of AI in quantitative investment has become increasingly significant, with a noticeable divergence in returns between subjective long and quantitative long strategies since 2020 [2] - The technological advancements in AI are paving the way for quantitative investment, making "AI + Quantitative" a crucial consideration for investors in the future [2] - DeepSeek, backed by Liang Wenfeng's firms, has set a benchmark in the industry, with its affiliated firm, Huanfang Quantitative, leading in average returns over the past six months and year [2]