金融工程 短期：成交量模型所有宽基指数中性。特征龙虎榜机构模型中性。特征成交量 模型看空。智能算法沪深 300 模型中性，智能算法中证 500 模型看空。 中期：涨跌停模型中性。上下行收益差模型国证 2000 指数与全 A 指数信号由 看多转为中性。月历效应模型中性。 长期：长期动量模型看多。 综合：A 股综合兵器 V3 模型看空。A 股综合国证 2000 模型看空。 港股模型： 中期：成交额倒波幅模型看空。恒生指数上下行收益差模型中性。 证 券 研 究 报 告 【金工周报】（20251117-20251121） 指数信号整体中性偏空，短期震荡偏空 ❖ 本周回顾 本周市场普遍下跌，上证指数单周下跌 3.9%，创业板指单周下跌 6.15%。 A 股模型： 本周行业指数全线下跌，跌幅前五的行业为：综合、电力设备及新能源、基础 化工、综合金融、钢铁。从资金流向角度来说，所有行业主力资金净流出，其 中电子、电力设备及新能源、基础化工、医药、机械主力资金净流出居前。 本周股票型基金总仓位为 95.82%，相较于上周减少了 52 个 bps，混合型基金 总仓位 88.71%，相较于上周减少了 218 个 bps。 本周汽 ...

- The report includes multiple quantitative models for market timing, categorized into short-term, medium-term, and long-term models. Short-term models include the "Volume Model" (neutral for all broad-based indices), "Feature Institutional Model" (bullish), "Feature Volume Model" (bearish), and "Smart Algorithm Models" (bearish for CSI 300 and CSI 500 indices) [1][11][62][63]. Medium-term models include the "Limit-Up and Limit-Down Model" (neutral), "Up-Down Return Difference Model" (bullish), and "Calendar Effect Model" (neutral) [12][64]. Long-term models include the "Long-Term Momentum Model" (bullish) [13][65]. Comprehensive models such as "A-Share Comprehensive Weapon V3 Model" and "A-Share Comprehensive CSI 2000 Model" are bearish [14][65]. - The "Volume Model" is constructed based on trading volume data, aiming to capture short-term market sentiment [1][11]. The "Feature Institutional Model" leverages institutional trading patterns observed in the market [1][11]. The "Feature Volume Model" focuses on specific volume characteristics to predict market trends [1][11]. The "Smart Algorithm Models" utilize machine learning algorithms to analyze historical data and predict market movements [1][11][62][63]. The "Limit-Up and Limit-Down Model" analyzes the frequency and impact of limit-up and limit-down events [12][64]. The "Up-Down Return Difference Model" calculates the difference between upward and downward returns to assess market direction [12][64]. The "Calendar Effect Model" incorporates seasonal and calendar-based effects on market performance [12][64]. The "Long-Term Momentum Model" evaluates long-term price trends to predict future movements [13][65]. Comprehensive models combine signals from multiple individual models to provide an overall market outlook [14][65]. - Evaluation of these models indicates that short-term models show mixed signals, with some bullish and others bearish, reflecting market uncertainty [1][11][62][63]. Medium-term models are generally neutral to bullish, suggesting moderate optimism [12][64]. Long-term models are bullish, indicating strong confidence in sustained upward trends [13][65]. Comprehensive models are bearish, signaling caution in the overall market outlook [14][65]. - Backtesting results for the models are not explicitly detailed in the report, but the report mentions the performance of specific indices and their alignment with model predictions. For example, the CSI 300 index showed bearish signals from the "Smart Algorithm Model," aligning with its weekly decline of 1.08% [8][11][63]. Similarly, the "Up-Down Return Difference Model" showed bullish signals, consistent with positive medium-term outlooks [12][64]. - The report also includes quantitative factor-based strategies such as "Double-Bottom Pattern" and "Cup-and-Handle Pattern." The "Double-Bottom Pattern" achieved a weekly return of 4.09%, outperforming the Shanghai Composite Index by 4.61% [40][47]. The "Cup-and-Handle Pattern" achieved a weekly return of 0.6%, outperforming the Shanghai Composite Index by 1.12% [40][41]. These factors are constructed based on technical chart patterns and are evaluated for their relative performance against benchmark indices [40][41][47].

Group 1 - The core point of the news is that Huicheng Co., Ltd. has experienced a continuous decline in stock price, dropping 1.38% on November 4, with a total market value of 13.53 billion yuan and a cumulative decline of 8.42% over four days [1] - Huicheng Co., Ltd. is located in Hefei, Anhui Province, and was established on December 18, 2015. It was listed on August 18, 2022. The company specializes in the manufacturing of gold bumping, wafer testing, and various packaging services for display driver chips [1] - The main business revenue composition of Huicheng Co., Ltd. is 90.25% from display driver chip testing and packaging, while other services account for 9.75% [1] Group 2 - From the perspective of fund holdings, one fund under Shenwan Hongyuan has Huicheng Co., Ltd. as its second-largest holding, with 21,900 shares, accounting for 1.97% of the fund's net value [2] - The fund, Shenwan Hongyuan Intelligent Life Quantitative Selection Mixed Fund A, has experienced a floating loss of approximately 48,180 yuan today and a total floating loss of 31,800 yuan over the past four days [2] - The fund was established on March 17, 2023, with a current scale of 19.53 million yuan, and has achieved a return of 30.57% this year, ranking 2,983 out of 8,150 in its category [2]

- The report mentions multiple quantitative models for market timing, including short-term, mid-term, and long-term models. Short-term models include the "Volume Model" (neutral for all broad-based indices), "Feature Volume Model" (bearish), "Feature Institutional Model" (bearish), and "Smart Algorithm Model" (bearish for CSI 300, neutral for CSI 500)[1][13][66]. Mid-term models include the "Limit-Up-Limit-Down Model" and "Calendar Effect Model," both neutral[14][67]. The long-term model is the "Long-Term Momentum Model," which is bullish[15][68]. Comprehensive models like "A-Share Comprehensive Weapon V3 Model" and "A-Share Comprehensive CSI 2000 Model" are bearish[16][69]. - The "Volume Model" is constructed based on trading volume trends, while the "Feature Volume Model" and "Feature Institutional Model" focus on specific volume characteristics and institutional trading patterns, respectively. The "Smart Algorithm Model" utilizes machine learning techniques to predict market movements[1][13][66]. The "Limit-Up-Limit-Down Model" analyzes price limits, and the "Calendar Effect Model" incorporates seasonal patterns[14][67]. The "Long-Term Momentum Model" evaluates price trends over extended periods[15][68]. - The "Comprehensive Weapon V3 Model" and "Comprehensive CSI 2000 Model" combine signals from multiple models across different timeframes to provide a holistic market outlook[16][69]. - The report evaluates these models qualitatively, noting that short-term models are generally neutral to bearish, mid-term models are neutral, and long-term models are bullish. Comprehensive models are bearish for A-shares[1][13][66][16][69]. - Testing results for the models are summarized as follows: Short-term models show mixed signals, with bearish predictions for specific indices like CSI 300 and CSI 2000. Mid-term models remain neutral, while the long-term momentum model indicates a bullish outlook. Comprehensive models suggest a bearish trend for A-shares[1][13][66][16][69]. - For Hong Kong stocks, the "Turnover Inverse Volatility Model" is bearish, indicating potential downward movement for the Hang Seng Index[16][70]. - The report also highlights shape-based models like the "Double Bottom Pattern" and "Cup-and-Handle Pattern." The "Double Bottom Pattern" portfolio outperformed the Shanghai Composite Index by 2.57% this week, with cumulative returns of 34.32% since December 31, 2020[43][48]. The "Cup-and-Handle Pattern" portfolio outperformed the Shanghai Composite Index by 1.28% this week, with cumulative returns of 70.89% since December 31, 2020[43][44]. - The report evaluates these shape-based models positively, noting their consistent outperformance compared to the benchmark index over time[43][44][48]. - Testing results for shape-based models: "Double Bottom Pattern" portfolio weekly return of 3.0%, cumulative return of 34.32% since December 31, 2020[43][48]. "Cup-and-Handle Pattern" portfolio weekly return of 1.71%, cumulative return of 70.89% since December 31, 2020[43][44].

Quantitative Models and Construction - **Model Name**: Competitive Advantage Portfolio **Construction Idea**: Incorporates competitive environment and strategic factors into stock selection, focusing on industries with distinct competitive characteristics[10][11] **Construction Process**: 1. Classify industries into four types: "Barrier Shield", "Intense Competition", "Steady Progress", and "Seeking Breakthrough"[10] 2. Focus on "Barrier Shield" industries to identify "dominant leaders" and "cooperative win-win" companies[10] 3. Combine "dominant leaders + cooperative win-win" stocks with "efficient operators" from non-barrier industries to form the portfolio[11] **Evaluation**: Provides a unique value quantification perspective beyond traditional factor investing[10] - **Model Name**: Margin of Safety Portfolio **Construction Idea**: Focuses on internal value estimation and competitive advantage to ensure sustainable profitability[15] **Construction Process**: 1. Calculate intrinsic value using profitability metrics like ROIC and NOPAT[15] 2. Select top 50 stocks with the highest margin of safety from a competitive advantage pool[15] 3. Weight stocks by dividend yield to maximize portfolio safety margin[15][17] **Evaluation**: Emphasizes reliable intrinsic value estimation and sustainable competitive advantage[15] - **Model Name**: Dividend Low Volatility Adjusted Portfolio **Construction Idea**: Avoids "high dividend traps" by considering dividend sustainability and excluding extreme cases[21] **Construction Process**: 1. Predict dividend yield and exclude stocks with extreme price performance or abnormal debt ratios[21] 2. Optimize portfolio by focusing on stocks with stable dividend yields[21] **Evaluation**: Addresses the risks of chasing high dividend yields without considering long-term value[21] - **Model Name**: AEG Valuation Potential Portfolio **Construction Idea**: Utilizes abnormal earnings growth (AEG) to capture valuation potential[25] **Construction Process**: 1. Calculate AEG using the formula: $$\begin{array}{c}{{A E G=Y_{t}-N_{t}=(E_{t}+r*D P S_{t-1})-(1+r)*E_{t-1}}}\\ {{\frac{V_{0}}{E_{1}}=\frac{1}{r}+\frac{1}{r}*\frac{\left(\frac{A E G_{2}}{1+r}+\frac{A E G_{3}}{(1+r)^{2}}+\frac{A E G_{4}}{(1+r)^{3}}+\cdots\right)}}}\\ {{\frac{E_{1}}{E_{1}}}}\end{array}$$[25] 2. Select top 100 stocks based on AEG_EP factor, then narrow down to top 50 with high dividend reinvestment/P ratio[29] **Evaluation**: Captures undervalued growth potential in companies overlooked by the market[25][29] - **Model Name**: Cash Cow Portfolio **Construction Idea**: Evaluates companies' cash generation efficiency using CFOR analysis[32] **Construction Process**: 1. Use CFOR metrics to assess free cash flow stability and operational asset returns[32] 2. Combine high-quality stocks from non-financial sectors with ROE above the 40th percentile[33] 3. Select stocks with low volatility and valuation factors for final portfolio construction[33] **Evaluation**: Provides a comprehensive view of operational performance and financial stability[32] - **Model Name**: Distress Reversal Portfolio **Construction Idea**: Captures valuation-driven short-term fluctuations and recovery potential[39] **Construction Process**: 1. Use inventory cycles to identify distress reversal opportunities[39] 2. Combine factors like accelerated recovery and undervaluation to select top 50 stocks[39] **Evaluation**: Complements momentum strategies by focusing on valuation-driven returns during downturns[39] --- Model Backtesting Results - **Competitive Advantage Portfolio**: Annualized return 20.36%, Sharpe ratio 0.95, IR 0.12, max drawdown -19.32%, Calmar ratio 1.05[14] - **Margin of Safety Portfolio**: Annualized return 23.37%, Sharpe ratio 1.17, IR 0.13, max drawdown -16.89%, Calmar ratio 1.38[19] - **Dividend Low Volatility Adjusted Portfolio**: Annualized return 16.81%, Sharpe ratio 0.98, IR 0.16, max drawdown -21.61%, Calmar ratio 0.78[22] - **AEG Valuation Potential Portfolio**: Annualized return 24.88%, Sharpe ratio 1.13, IR 0.17, max drawdown -24.02%, Calmar ratio 1.04[31] - **Cash Cow Portfolio**: Annualized return 14.15%, Sharpe ratio 0.71, IR 0.10, max drawdown -19.80%, Calmar ratio 0.71[37] - **Distress Reversal Portfolio**: Annualized return 25.17%, Sharpe ratio 1.01, IR 0.15, max drawdown -33.73%, Calmar ratio 0.75[41]

Core Viewpoint - The article emphasizes the rapid development of AI and quantitative technology in the investment sector, highlighting the importance of continuous strategy evolution for the long-term success of quantitative private equity firms like Tianyan Capital, which was founded in 2014 and has a strong focus on innovation and adaptation [2]. Company Overview - Tianyan Capital was co-founded by Xie Xiaoyang and Zhang Sen, both of whom have over ten years of industry experience. The company’s name reflects its commitment to change and deep insights into the essence of investment [2]. - The firm has received multiple industry awards, including the "Golden Changjiang Award" and "Yinghua Award," and ranks among the top ten quantitative private equity firms in terms of performance [3][4]. Performance Metrics - As of September 2025, Tianyan Capital's products have achieved impressive returns, with an average return of ***% over the past three years, placing it first in the industry [3][4]. - The firm manages approximately 2.1 billion yuan across 11 products that meet ranking criteria, showcasing its strong long-term performance [3]. Investment Strategy - The core strategy of Tianyan Capital is centered around a multi-factor model for stock selection, which allows for higher alpha returns at a lower cost [8]. - The flagship product, "Tianyan Saineng," has been operational since May 2016 and has demonstrated significant returns, with a focus on maintaining model autonomy and stability in risk control [10][11]. Team and Culture - The investment research team at Tianyan Capital consists of over half PhD holders from prestigious institutions, fostering a culture of free exploration and innovation [12]. - The company emphasizes long-termism in its operations, avoiding arbitrary changes to risk parameters and maintaining a stable risk control model [10][11]. Market Position and Future Outlook - Tianyan Capital has strategically positioned itself to balance scale and performance, understanding that growth in assets under management should align with long-term performance and research capabilities [14]. - The firm has also obtained a Hong Kong license to enhance its global asset allocation capabilities, focusing on capturing unique alpha opportunities in the Chinese market while catering to international investors [16].

- The short-term trading volume model is neutral for A-shares[2][12] - The low volatility model is neutral for A-shares[2][12] - The characteristic institutional model is bearish for A-shares[2][12] - The characteristic trading volume model is bearish for A-shares[2][12] - The intelligent algorithm model for the CSI 300 is bearish for A-shares[2][12] - The intelligent algorithm model for the CSI 500 is bearish for A-shares[2][12] - The mid-term limit-up and limit-down model is neutral for A-shares[2][13] - The mid-term calendar effect model is neutral for A-shares[2][13] - The long-term momentum model is bullish for A-shares[2][14] - The comprehensive A-share model V3 is bearish[2][15] - The comprehensive A-share model for the CSI 2000 is bearish[2][15] - The mid-term trading volume to volatility model is bearish for Hong Kong stocks[2][16]

Fund Overview - The fund is named "Suxin Ruijian Quantitative Stock Selection Equity Investment Fund" and has been approved for fundraising by the China Securities Regulatory Commission [1] - The fund will be publicly offered from October 15, 2025, to October 31, 2025, with a maximum fundraising period of three months [2][30] - The fund is categorized as an open-ended equity fund with an indefinite duration and aims to achieve stable asset appreciation while controlling risks [21] Fund Structure - The fund offers two classes of shares: Class A and Class C, with different fee structures for subscription and redemption [1][36] - Class A shares have a code of 025404 and charge subscription fees, while Class C shares (code 025405) do not charge subscription fees [1][21] Investment Strategy - The fund's investment range includes liquid financial instruments such as stocks, depositary receipts, bonds, asset-backed securities, and stock index futures [22][24] - The fund will invest 80%-95% of its assets in stocks and depositary receipts, with a minimum of 5% in cash or short-term government bonds [24] Subscription Details - The minimum initial subscription amount is RMB 10,000, and subsequent subscriptions can be as low as RMB 1,000 [3][40] - Investors can make multiple subscriptions during the fundraising period, but once a subscription application is accepted, it cannot be revoked [4][41] Fund Management - The fund is managed by Suxin Fund Management Co., Ltd., which is responsible for the fund's operations and investment decisions [64] - The fund's custodian is Shanghai Pudong Development Bank Co., Ltd., ensuring the safekeeping of the fund's assets [65] Regulatory Compliance - The fund must meet specific conditions to complete its fundraising, including raising at least 200 million shares and having a minimum of 200 investors [6][31] - If the fundraising conditions are not met by the end of the period, the fundraising will be deemed unsuccessful, and the management company will return the funds to investors [33][63]

Core Insights - The 2025 Autumn Strategy Conference by CITIC Futures focused on the theme "Tides Surge, Breakthroughs and Innovations," analyzing investment opportunities across various sectors for Q4 and 2026 [1] Macro and Precious Metals Forum - The macroeconomic outlook for Q4 is characterized by a "steady progress" approach, with policies aimed at stabilizing growth through 500 billion yuan in financial tools and potential interest rate cuts [2] - Gold is expected to show a strong oscillation in Q4, with long-term strategic allocation opportunities due to the anticipated decline in real interest rates and ongoing geopolitical tensions [2] Financial Forum - Equity assets are projected to perform positively in Q4, driven by new capital inflows and policy expectations, with a focus on IM long positions and strategies to capitalize on market movements [3] - The bond market may shift from a weak stance, with a potential recovery in bullish sentiment, although the 10-year government bond yield is expected to fluctuate between 1.65% and 1.95% [3] Energy and Chemical Forum - The energy and chemical sectors are facing slightly weak supply and demand dynamics in Q4, with oil prices influenced by geopolitical factors and supply disruptions [4] - The chemical industry is under pressure from increasing production capacities, particularly in PVC and styrene, which may hinder demand growth without supportive consumption policies [4] Non-Ferrous Metals Forum - The non-ferrous metals sector is expected to see a positive shift in Q4, with copper, aluminum, and tin being highlighted as potential bullish opportunities due to supply disruptions and macroeconomic support from interest rate cuts [5][6] - Industrial silicon and lithium carbonate may face downward pressure, while polysilicon is expected to benefit from supply-side contraction policies [6] Agricultural Forum - Agricultural products are in a transitional phase between old and new crops, with inventory dynamics and international trade relations significantly impacting market conditions [7] - The soybean market is expected to remain stable, while palm oil may see bullish opportunities due to seasonal production declines [7] Black Metals Forum - The black metals market is anticipated to experience a mixed trend, with short-term price support from a favorable macro environment, but potential long-term weakness due to inventory pressures [8] - Iron ore prices are expected to fluctuate widely, while coal and coke prices may initially rise before facing downward pressure [8] Innovation Forum - The energy sector is under pressure from oversupply, with fossil fuels facing challenges, while the demand for new energy sources is expected to grow steadily [9] - The shipping market is projected to perform strongly due to production increases and sanctions, with coal supply tightening expected to support prices [9]

Group 1 - The report highlights the significant role of quantitative stock selection funds in the public fund market, with a total of 277 funds managing a combined scale of 90.32 billion yuan as of the end of Q2 2025, offering broader investment scope and higher style exposure flexibility compared to standard index-enhanced funds [3][12][13] - The report focuses on five fund managers with distinctive investment frameworks in quantitative stock selection, including Feng Xixiang from Xinda Australia Fund, Gao Chongnan from Guotai Fund, Lin Jingyi from Xinda Australia Fund, Shi Yunchao from Penghua Fund, and Zhai Zijian from Western Li De Fund, providing insights into their strategies and product positioning [3][12][13] Group 2 - Feng Xixiang employs a unified framework emphasizing the effectiveness of factors and the universality of alpha models, integrating static multi-factor linear models with machine learning dynamic weighting models since 2023, achieving balanced allocation in his representative products [4][16][23] - Gao Chongnan focuses on the Calmar ratio, selecting high dividend, quality, and growth styles to enhance the stability of risk-return profiles, with a product positioning aimed at low volatility value style [4][35][36] - Lin Jingyi implements a "HI+AI" approach using an integrated research platform, employing a three-step method to replicate successful peer consensus and enhance index tracking through multiple alpha models [5][22] - Shi Yunchao's strategy combines multi-factor linear models with a higher proportion of non-linear models, focusing on short prediction cycles and higher turnover rates, while maintaining a diversified portfolio to mitigate risks [6][24] - Zhai Zijian adopts an AI quantitative investment strategy with a "core + satellite" multi-strategy balanced configuration, utilizing machine learning for long-term predictions and high-frequency data analysis [6][24] Group 3 - The report indicates that as of the end of Q2 2025, Feng Xixiang manages a total of 4.54 billion yuan across seven quantitative stock selection products, with representative products achieving cumulative returns of 40.66% and 74.91% since inception, significantly outperforming their benchmark indices [17][21] - Gao Chongnan's strategy iteration has led to improved performance, with the National Strategy Yield Fund achieving an annualized return of 28.72% in 2024, reflecting a notable enhancement in risk-adjusted returns [36][37] - The quantitative team at Xinda Australia Fund consists of experienced professionals, with a comprehensive product line that includes 11 quantitative stock selection products and 2 quantitative fixed income + strategy products, aiming to reduce volatility while seeking absolute returns [32][33]