Core Insights - The A-share market is experiencing significant activity, with financing balances reaching a nearly ten-year high of 22,454.72 billion yuan, and the Shenzhen market setting a historical record [1] - Despite the apparent market prosperity, with the Shanghai Composite Index stabilizing above 3,800 points and 80% of stocks rising, less than 50% of stocks have increased by more than 6%, indicating a typical bull market scenario [3] Group 1: Market Dynamics - The current market presents four major misleading phenomena during a bull market: 1. The "waiting for a rise" illusion, where 80% of stocks expected to rise end up falling [4] 2. The "hot and cold" illusion, where true market hotspots show sustainability rather than frequent rotation [5] 3. The "rise and fall" illusion, where stock price fluctuations do not reflect underlying institutional movements [6] 4. The "high and low" illusion, where retail investors focus on price levels rather than behavioral data, leading to misinterpretation of market signals [7] Group 2: Institutional Behavior - The analysis of institutional trading characteristics reveals that active institutional participation is crucial for sustaining stock price increases. Stocks lacking institutional involvement are likely to experience declines [9][11] - Approximately 70% of seemingly strong stocks lack sustained institutional participation, explaining why many investors incur losses in a bull market by chasing superficial trends rather than understanding underlying dynamics [11] Group 3: Leverage and Investment Strategy - The current leverage data indicates that leveraged funds act as both a driving force and an amplifier in the market. Successful investors are those who can see beyond appearances and grasp the essence of market movements [12] - In an era of information overload, investors are encouraged to establish a "data moat" by developing a multi-dimensional evaluation system rather than relying solely on price charts or news [12]

Core Viewpoint - The retirement of Zhu Yongqiang, the general manager of Xinda Australia Fund, marks a significant personnel change in the public fund industry, with Fang Jing temporarily taking over the role [1][2]. Company Overview - Xinda Australia Fund, established in June 2006, is the first fund company in China controlled by a state-owned asset management company [5]. - The company has experienced substantial growth under Zhu Yongqiang's leadership, with total assets increasing over 7 times from 127.60 billion yuan at the end of 2019 to a peak of 1374.51 billion yuan by September 2024 [5][6]. Leadership Transition - Zhu Yongqiang retired on September 5, 2025, after more than five years as general manager, during which he significantly expanded the company's asset management scale [2][5]. - Fang Jing, the current deputy general manager, will temporarily assume the role of general manager [1][3]. Performance Metrics - Under Zhu's management, the fund's non-monetary asset scale grew from 118.46 billion yuan to 681.11 billion yuan, with the company ranking 61 out of 183 in the industry by mid-2025 [5]. - The company achieved a total scale increase of 7.07 times and a non-monetary fund scale increase of 4.75 times during Zhu's tenure [5]. Investment Strategy - The company focused on enhancing its equity research capabilities by investing in experienced researchers and building a diverse talent structure across key sectors such as technology and consumer goods [6]. - In fixed income, the company emphasized risk control and developed a rigorous credit rating system to manage various risks dynamically [6]. - The company has also embraced financial technology trends by integrating AI algorithms into its investment strategies, creating a comprehensive quantitative research ecosystem [6]. Recent Performance - As of August 31, 2025, 41 of the company's fund products achieved over 30% returns in the past year, with 34 products rising over 50% and 13 products doubling their returns [7].

Group 1 - The article highlights the contrast between long-term holding strategies of certain funds and the short-term speculative behavior in the A-share market, illustrating a "race to run" mentality among investors [1][3] - Specific funds are noted for their long-term positions in stocks like Tencent and Kweichow Moutai, with some holding these positions for over 30 reporting periods, indicating a deep commitment to their investment choices [2][5] - The narrative suggests that while these funds appear passive, they are actually leveraging data-driven models to validate the predictability of their investments, particularly in companies like Moutai [5][10] Group 2 - The article emphasizes the importance of data over news in investment decisions, suggesting that successful investors build a quantifiable understanding of market dynamics rather than reacting to headlines [4][18] - An example is provided with Tongyuan Petroleum, which saw a significant price increase attributed to market conditions, but prior data signals indicated institutional activity that foreshadowed this movement [6][17] - The discussion encourages investors to develop their own data observation systems, focus on capital flows rather than news, and understand market expectations, highlighting that true investment insights often lie in overlooked quantitative signals [18]

Quantitative Models and Construction Methods 1. Model Name: RNN-LIN - **Model Construction Idea**: Simplify the traditional GRU model by using a linear RNN structure, reducing parameter complexity while maintaining competitive performance[2][17][20] - **Model Construction Process**: - The model uses a linear RNN structure with only a forget gate and an output gate. The hidden state is updated without non-linear activation functions - Equations: $ h_{t} = f_{t} \otimes h_{t-1} + (1 - f_{t}) \otimes c_{t} $ $ y_{t} = o_{t} \otimes h_{t} $ $ f_{t} = Sigmoid(x_{t}W_{f}) $ $ o_{t} = Sigmoid(x_{t}W_{o}) $ $ c_{t} = SiLU(x_{t}W_{c}) $ - $f_{t}$: Forget gate - $o_{t}$: Output gate - $c_{t}$: Candidate state[20][21] - The model reduces parameters by approximately 50% compared to GRU[21] - **Evaluation**: The linear RNN model shows slightly weaker performance than GRU but remains competitive. Adding GLU modules improves its performance significantly[22][53] 2. Model Name: DecompGRN - **Model Construction Idea**: Extend the linear RNN by integrating cross-sectional information directly into the RNN gating mechanism, enabling simultaneous modeling of temporal and cross-sectional data[2][50] - **Model Construction Process**: - The first RNN layer outputs individual stock representations at each time step - Cross-sectional information is incorporated by grouping stocks based on market capitalization and calculating group de-meaned values - The second RNN layer combines temporal and cross-sectional information in the forget and output gates - Equations: $ h_{t} = f_{t} \otimes h_{t-1} + (1 - f_{t}) \otimes c_{t} $ $ y_{t} = o_{t} \otimes h_{t} $ $ f_{t} = Sigmoid(x_{t}W_{f}) $ $ o_{t} = Sigmoid(x_{t}W_{o}) $ $ c_{t} = SiLU(x_{t}W_{c}) $ - $f_{t}$: Forget gate - $o_{t}$: Output gate - $c_{t}$: Candidate state[50][55] - **Evaluation**: DecompGRN outperforms the GRU baseline in terms of RankIC and RankICIR while maintaining only 43% of the GRU's parameter count[74][53] --- Model Backtest Results 1. RNN-LIN - **RankIC**: - CSI All Share: 0.13 - CSI 300: 0.10 - CSI 500: 0.09 - CSI 1000: 0.12[36][37] - **RankICIR**: - CSI All Share: 1.08 - CSI 300: 0.62 - CSI 500: 0.71 - CSI 1000: 0.96[36][37] - **IC Win Rate**: - CSI All Share: 0.88 - CSI 300: 0.74 - CSI 500: 0.78 - CSI 1000: 0.86[36][37] - **Annualized Return (Top Group)**: - CSI All Share: 42.59% - CSI 300: 28.59% - CSI 500: 23.68% - CSI 1000: 32.81%[42] 2. DecompGRN - **RankIC**: - CSI All Share: 0.141 - CSI 300: 0.099 - CSI 500: 0.098 - CSI 1000: 0.127[55][58] - **RankICIR**: - CSI All Share: 1.26 - CSI 300: 0.65 - CSI 500: 0.77 - CSI 1000: 1.08[55][58] - **IC Win Rate**: - CSI All Share: 0.89 - CSI 300: 0.74 - CSI 500: 0.78 - CSI 1000: 0.88[55][58] - **Annualized Return (Top Group)**: - CSI All Share: 57.68% - CSI 300: 31.69% - CSI 500: 26.9% - CSI 1000: 40.35%[57][58] --- Index Enhancement Test Results (DecompGRN) - **Annualized Excess Return**: - CSI 300: 10.24% - CSI 500: 10.05% - CSI 1000: 19.58%[75][85] - **Tracking Error**: - CSI 300: 5.07 - CSI 500: 6.1 - CSI 1000: 6.75[75][85] - **Cumulative Excess Return (as of 2025-08-27)**: - CSI 300: 3.93% - CSI 500: 6.72% - CSI 1000: 18.26%[75][85]

Core Viewpoint - The stock market experienced a broad rally in August, but there was significant divergence among individual stocks, leading to a general decline in excess returns for quantitative strategies. Only 20.83% of the 696 quantitative long products achieved positive excess returns in August [2][3]. Summary by Sections Market Performance - In August, the quantitative long strategies saw a decline in excess returns, particularly after the second week, when technology stocks like Yizhongtian, Hanwujing, and Industrial Fulian became the main drivers of market gains, while other sectors lagged [2][3]. - The average excess return for quantitative long products was -2.28% for the month, with only 145 out of 696 products achieving positive excess returns [3]. Weekly Performance Breakdown - **Week 1**: Average excess return of 0.81%, with 84.20% of products achieving positive excess [3]. - **Week 2**: Average excess return of -1.06%, with only 21.98% of products achieving positive excess [3]. - **Week 3**: Average excess return of -0.61%, with 23.85% of products achieving positive excess [3]. - **Week 4**: Average excess return of -1.99%, with only 12.21% of products achieving positive excess [3]. Notable Products - Despite the overall decline, 20 products managed to achieve positive excess returns for all four weeks in August, including "Mingyue Songjian Zhaozhao 1" managed by Huang Chen and "Nianjue Lianzhen Quantitative Selected Excellent 1" managed by Wang Xiao [3][4]. - The top-performing product in August was "Nianjue Lianzhen Quantitative Selected Excellent 1," which achieved a significant excess return, outperforming other quantitative long products [9]. Performance by Fund Size - **Over 100 Billion**: 51 products achieved positive excess returns, with a total excess return of -1.69% [5][10]. - **50-100 Billion**: 20 products achieved positive excess returns, with a total excess return of -1.35% [10][11]. - **20-50 Billion**: 9 products achieved positive excess returns, with a total excess return of -2.53% [13]. - **10-20 Billion**: 14 products achieved positive excess returns, with a total excess return of -2.57% [16]. - **5-10 Billion**: 19 products achieved positive excess returns, with a total excess return of -2.55% [18]. - **0-5 Billion**: 32 products achieved positive excess returns, with a total excess return of -3.02% [20]. Key Players - Mingyue Songjian Zhaozhao 1 and Nianjue Lianzhen Quantitative Selected Excellent 1 were highlighted for their strong performance in August [4][9]. - Mingyue Investment and Mengxi Investment were noted for their significant number of products achieving positive excess returns [8][9].

Core Viewpoint - The private equity market has seen a significant increase in the number of registered products, with a year-on-year growth of over 82% in 2025 compared to 2024, indicating a strong recovery and investor interest in private equity securities products [2]. Group 1: Growth in Private Equity Products - A total of 7,907 private equity securities products have been registered in 2025, compared to 4,340 in the same period of 2024, marking an increase of 82.19% [2]. - Monthly registration data shows a notable recovery since March 2025, with monthly registrations exceeding 1,100 products for three consecutive months from June to August [2]. - Factors contributing to this growth include a steady recovery in the A-share market, the rise of strategic emerging industries like artificial intelligence, improved regulatory frameworks, and the strong performance of private equity products [2]. Group 2: Dominance of Stock Strategies - Stock strategy products have become the mainstay of the registered private equity market, with 5,173 products accounting for 65.42% of the total, reflecting a 91.31% increase from 2,704 products in 2024 [3]. Group 3: Diverse Strategy Growth - The demand for diversified asset allocation has led to stable growth in multi-asset strategies and futures and derivatives strategies, with 1,116 and 841 products registered respectively, representing increases of 76.58% and 66.87% compared to 2024 [6]. Group 4: Rise of Quantitative Products - Quantitative products have seen a significant rise, with 3,584 products registered, making up 45.33% of the total, and a year-on-year growth of 100.34% from 1,789 products in 2024 [8]. - Within quantitative strategies, stock strategies dominate, with 2,601 products registered, accounting for 72.57% of quantitative products [8]. Group 5: Leading Private Equity Managers - A total of 2,154 private equity managers have registered products this year, with 1,436 managers (66.7%) registering 2,690 products, while 76 billion-level managers accounted for 1,936 products [10]. - The top three billion-level quantitative private equity managers by registered products are KuanDe with 118 products, HeiYi with 112 products, and MingHeng with 101 products [12][14].

Core Viewpoint - The private equity market in A-shares is experiencing a resurgence, with a significant increase in the issuance of new products, particularly in quantitative strategies, driven by a recovering market and improved regulatory environment [1][2][3]. Group 1: Market Recovery and Product Issuance - As of August 31, 2025, the number of newly registered private equity securities products has exceeded 7900, representing a year-on-year increase of 82.19% from 4340 in the same period of 2024 [2]. - The monthly registration volume has remained above 1100 for three consecutive months, indicating sustained enthusiasm for private equity securities products [2]. - In July 2025, the new registration scale reached 1074.27 billion, marking a nearly four-year high, compared to only 156 billion in the same month last year [2]. Group 2: Factors Driving Growth - The growth in registered private equity securities products is attributed to several factors, including the steady recovery of the A-share market and the strong performance of strategic emerging industries like artificial intelligence, which have increased market risk appetite [3]. - The improvement in the regulatory framework and transparency of the private equity fund industry has bolstered investor confidence, creating a stable environment for healthy industry development [3]. - The strong profitability of private equity securities products has further stimulated investor participation [3]. Group 3: Quantitative Investment Strategies - Quantitative investment strategies have seen rapid development and demonstrated significant advantages in the current market environment, leading to increased investor interest [4]. - The number of registered quantitative products has doubled year-on-year, reaching 3584, accounting for 45.33% of the total [7]. - Among the quantitative products, stock strategies are the most popular, making up 72.57% of the total, followed by futures and derivatives (13.81%) and multi-asset strategies (10.66%) [7]. Group 4: Performance of Specific Strategies - Stock strategy products have been particularly noteworthy, with 5173 registered, representing 65.42% of the total, and a year-on-year increase of over 90% [5]. - Multi-asset and futures/derivatives strategies have also seen steady growth, with registrations increasing by 76.58% and 66.87%, respectively [5]. - The performance of quantitative products has been strong, with some leading institutions reporting excess returns of 20% to 30% [6]. Group 5: Dominance of Leading Private Equity Firms - A total of 76 private equity firms with over 10 billion in assets have registered 1936 products this year, averaging at least 25 products per firm, which accounts for nearly a quarter of the total [8]. - Three leading quantitative private equity firms have registered over 100 products each, with KuanDe Private Equity leading at 118 products, followed by HeiYi Asset at 112, and MingFa Investment at 101 [8][10]. - The concentration of product registrations among top firms highlights the significant headwind effect in the industry [9].

Group 1 - The non-farm payroll report is crucial for global markets, with expectations of 75,000 new jobs in August, but analysts question the accuracy of the data [1] - The unemployment rate appears stable at 4.1%-4.2%, yet the employment-population ratio is declining, indicating potential discrepancies in the data [3] Group 2 - Institutional investors are the key players in stock price movements, controlling the narrative behind market fluctuations through their trading behaviors [3] - Cost control is essential for identifying "good stocks," as even renowned investors like Warren Buffett emphasize the importance of holding costs [4][5] Group 3 - Quantitative tools reveal institutional trading behaviors, as seen in the case of "Sifang Technology," which surged over 40% in the market due to active institutional inventory data prior to the price increase [6][8] - Concept stocks, such as "Chuan Da Zhi Sheng," also show signs of institutional interest through quantitative data before significant price movements [9] Group 4 - Ordinary investors should focus on real trading behaviors rather than debating data accuracy, as institutions utilize quantitative tools to understand market fundamentals [13]

Market Overview - The market is experiencing significant differentiation, with solid-state battery concept stocks like Tianhong Lithium and Zhengye Technology seeing substantial gains, and Yiwai Lithium Energy rising by 13% in early trading due to the launch of the "Longquan No. 2" all-solid-state battery [1] - The entire battery industry is expanding production, with leading companies like QianDao Intelligent holding orders worth 30 billion [1] Financing Market Insights - The margin financing balance has reached 2.27 trillion, increasing by 1.548 billion in a single day, with the electric power equipment sector being particularly favored, netting 3.338 billion in purchases [4] - Zhongji Xuchuang led with a net purchase of 1.748 billion, indicating strong institutional interest [4] Institutional Behavior Analysis - Despite perceptions of a bull market, there can be significant downturns, often driven by institutional investors who may panic and sell, only to buy back at lower prices [5] - The market's true dynamics are revealed through quantitative data, which can show that institutions are often accumulating shares during perceived downturns [5] Stock Selection and Performance - The importance of selecting the right stocks is emphasized, as not all stocks benefit equally from market uptrends, with less than 50% of stocks having opportunities during rebounds [7] - Quantitative data can highlight the difference in performance between stocks with institutional support and those without, underscoring the risks of poor stock selection [9] Market Psychology and Quantitative Tools - The market is influenced by human psychology, including herd behavior and overconfidence, which can lead to poor investment decisions [14] - Quantitative tools are essential for investors to navigate these psychological traps and understand the true movements of capital in the market [14]

本文首发于公众号"私募排排网"。 （点击↑↑ 上图查看详情 ） 8月，A股市场整体单边震荡上行，沪指站上3800点创下近10年新高，单日成交额最高突破3万亿，市场交投持续火爆，部分基金（公募、私募） 净值也随之创出历史新高。 然而，回顾上一轮结构性牛市，沪指由2019年初低点上涨至2021年2月3731.69的高点，却在随后近1年里于高点附近盘整，并自2022年初起震荡 下跌，迎来了将近3年的熊市行情。 其中， 有不少权益类公募基金和股票多头私募基金正是在 2021年高位时成立的，如今沪指早已"越过"2021年高位的3700多点，那些2021年成 立的基金产品至今收益如何？ 私募排排网数据显示， 截至 2025年8月29日，2021年成立的股票多头私募基金共433只，成立至今实现正收益的 404只，占比93.3% ；根据公募排排网数据，截至2025年8月27日， 2021年成立的权益类公募基金且有业绩展示的（剔除ETF）达1644只，成立 以来正收益产品526只，正收益占比仅32%，相较之下私募基金业绩更为强势 。 那么，有哪些股票多头私募基金收益表现更胜一筹呢？为了给予更多有价值的信息，笔者将筛选出 截至 ...