广发证券首席金工分析师 安宁宁 SAC: S0260512020003 anningning@gf.com.cn 广发证券资深金工分析师 张钰东 SAC: S0260522070006 zhangyudong@gf.com.cn 广发金工安宁宁陈原文团队 摘要 最近5个交易日，科创50指数跌0.59%，创业板指跌0.82%，大盘价值涨0.01%，大盘成长跌0.39%，上证50涨0.20%，国证2000代表的小盘涨1.09%，国防军 工、石油石化表现靠前，通信、综合表现靠后。 风险溢价，中证全指静态PE的倒数EP减去十年期国债收益率，权益与债券资产隐含收益率对比，截至2025/12/31指标2.69%，两倍标准差边界为4.70%。 估值水平，截至2025/12/31，中证全指PETTM分位数82%，上证50与沪深300分别为75%、75%，创业板指接近58%，中证500与中证1000分别为62%、 64%，创业板指风格估值相对历史总体处于中位数水平。 使用卷积神经网络对图表化的价量数据与未来价格进行建模，并将学习的特征映射到行业主题板块中。最新配置主题为化工、非银和卫星等，具体包括中 证细分化工产业主题、国证商 ...

Market Performance - The Sci-Tech 50 Index increased by 2.85% over the last five trading days, while the ChiNext Index rose by 3.90%. The large-cap value index fell by 0.02%, and the large-cap growth index increased by 2.70%. The Shanghai 50 Index gained 1.37%, and the small-cap index represented by the CSI 2000 rose by 3.55%. The sectors of non-ferrous metals and national defense performed well, while beauty care and social services lagged behind [1]. Valuation Levels - As of December 26, 2025, the static PE ratio of the CSI All Share Index is at the 82nd percentile. The Shanghai 50 and CSI 300 both stand at 74%, while the ChiNext Index is close to 59%. The CSI 500 and CSI 1000 are at 62% and 64%, respectively. The valuation of the ChiNext Index is relatively at the historical median level [1]. Risk Premium - The risk premium, calculated as the inverse of the static PE of the CSI All Share Index minus the yield of the 10-year government bond, is at 2.69% as of December 26, 2025. The two standard deviation boundary is at 4.70% [1]. Fund Flows - In the last five trading days, ETF inflows amounted to 41.6 billion yuan, and the margin trading balance increased by approximately 45.7 billion yuan. The average daily trading volume across the two markets was 1.9454 trillion yuan [2]. Thematic Investment - The latest thematic investment configuration includes sectors such as chemicals, non-bank financials, communications, and satellite industries. Specific indices mentioned are the CSI Sub-Industry Chemical Index, CSI 300 Non-Bank Financial Index, CSI All Share Communication Equipment Index, and the National Satellite Communication Industry Index [2][3]. AI and Machine Learning Application - The application of convolutional neural networks (CNN) for modeling price and volume data has been explored, focusing on standardizing chart data to predict future prices and mapping learned features to industry themes [10].

广发证券首席金工分析师 安宁宁 SAC: S0260512020003 anningning@gf.com.cn 广发证券资深金工分析师 张钰东 SAC: S0260522070006 zhangyudong@gf.com.cn 广发金工安宁宁陈原文团队 摘要 最近5个交易日，科创50指数跌2.99%，创业板指跌2.26%，大盘价值涨1.52%，大盘成长跌1.39%，上证50涨0.32%，国证2000代表的小盘跌0.37%，商贸零 售、非银金融表现靠前，电子、电力设备表现靠后。 风险溢价，中证全指静态PE的倒数EP减去十年期国债收益率，权益与债券资产隐含收益率对比，截至2025/12/19指标2.79%，两倍标准差边界为4.71%。 估值水平，截至2025/12/19，中证全指PETTM分位数80%，上证50与沪深300分别为74%、73%，创业板指接近55%，中证500与中证1000分别为59%、 60%，创业板指风格估值相对历史总体处于中位数水平。 资金交易层面，最近5个交易日，ETF资金流入721亿元，融资盘5个交易日减少约76亿元，两市日均成交17380亿元。 | 日期 | 指数代码 | 指数名称 | ...

Market Performance - The Sci-Tech 50 Index increased by 1.72% and the ChiNext Index rose by 2.74% over the last five trading days, while the large-cap value index fell by 1.21% and the large-cap growth index increased by 0.86% [1] - The static PE of the CSI All Share Index minus the yield of 10-year government bonds indicates a risk premium of 2.82% as of December 12, 2025, with a two-standard deviation boundary at 4.71% [1] Valuation Levels - As of December 12, 2025, the CSI All Share Index's PETTM percentile is at 79%, with the Shanghai Stock Exchange 50 and CSI 300 at 73% and 71% respectively, while the ChiNext Index is close to 55% [1] ETF and Fund Flows - In the last five trading days, ETF inflows amounted to 18.6 billion yuan, and the margin trading balance increased by approximately 23.5 billion yuan, with an average daily trading volume of 193.37 billion yuan across the two markets [2] Thematic Indexes - The latest thematic allocations include communication, artificial intelligence, and momentum growth in the ChiNext, specifically focusing on the CSI Communication Equipment Theme Index, ChiNext Artificial Intelligence Index, CSI 5G Communication Theme Index, and ChiNext Momentum Growth Index [2][3] Neural Network Analysis - A convolutional neural network (CNN) is utilized to model price and volume data, mapping learned features to industry thematic sectors, indicating a focus on AI and communication sectors [11]

Group 1 - The article emphasizes the growing scale of passive index funds, which reached a total of 4.9 trillion yuan by the end of September, with 1,548 passive index funds (ETFs and off-market funds) [7][11] - The article discusses the periodic adjustments of major indices like the SSE 50, CSI 300, and CSI 500, which occur every June and December, potentially creating investment opportunities due to significant changes in constituent stocks [4][5] Group 2 - Historical adjustment effects show that stocks added to indices tend to outperform the index in the two weeks prior to their inclusion, while stocks removed from indices generally underperform [12][14] - The average excess return for stocks added to the SSE 50 index in the two weeks before inclusion is 4.89%, with a success rate of 66.67% [15] - For the CSI 300 index, the average excess return for added stocks is 4.04%, with a success rate of 59.39% [18] Group 3 - The article presents the latest adjustment impact calculations, indicating that 572 stocks are involved in adjustments, with 20 stocks seeing net buy amounts exceeding 1 billion yuan and 14 stocks with net sell amounts over 1 billion yuan [25] - The impact coefficients for stocks show that 18 stocks have coefficients exceeding 2, indicating significant buying pressure, while 46 stocks have coefficients below -2, indicating selling pressure [25] Group 4 - The article outlines the index compilation schemes, noting that most indices have a weight limit of 10%, with some allowing up to 15% for individual stocks [24] - The methodology for calculating the impact of adjustments includes assessing the expected buy and sell amounts based on the total scale of tracking funds and the weights of constituent stocks [23]

Market Performance - The Sci-Tech 50 Index decreased by 0.08% over the last five trading days, while the ChiNext Index increased by 1.86%. The large-cap value index rose by 0.74%, and the large-cap growth index increased by 1.61%. The Shanghai 50 Index gained 1.09%, and the small-cap index represented by the CSI 2000 rose by 0.19%. The metals and communications sectors performed well, while media and real estate lagged behind [1]. Risk Premium and Valuation Levels - As of December 5, 2025, the risk premium, calculated as the inverse of the static PE of the CSI All Share Index minus the yield of ten-year government bonds, stands at 2.81%. The two-standard deviation boundary is 4.72% [1]. - The valuation level indicates that the CSI All Share Index's PETTM is at the 80th percentile, with the Shanghai 50 and CSI 300 at 75% and 72%, respectively. The ChiNext Index is close to 49%, while the CSI 500 and CSI 1000 are at 61% and 57%, respectively. The ChiNext Index's valuation is relatively at the historical median level [1]. ETF Fund Flow - Over the last five trading days, ETF funds experienced an outflow of 1.4 billion yuan, while margin trading increased by approximately 11.5 billion yuan. The average daily trading volume across both markets was 168.24 billion yuan [2]. Thematic Indexes - The latest thematic allocations include the CSI Communication Equipment Index, the CSI Chengdu-Chongqing Economic Circle Index, the CSI Low Volatility Dividend 100 Index, the ChiNext Momentum Growth Index, and the National Food Index [2][3][11]. Market Sentiment and Risk Appetite - The report includes observations on market sentiment based on the proportion of stocks above the 200-day moving average and tracks the risk appetite between equity and bond assets [12][13]. Financing Balance - The financing balance statistics indicate trends in margin trading and overall market leverage [15]. Individual Stock Performance - The report provides a distribution of individual stock performance based on year-to-date return ranges, highlighting the performance of various stocks in the current market environment [17]. Oversold Indices - An analysis of indices that are currently considered oversold is included, providing insights into potential investment opportunities [19].

Core Insights - The article emphasizes the importance of data collection and analysis for quantitative investors to uncover hidden market patterns and gain an edge in stock market trading [1][4][5]. Group 1: Data Types and Importance - Level 1 data includes basic market information such as highest price, lowest price, opening price, closing price, trading volume, and trading amount, updated every three seconds [6][7]. - Level 2 data provides more detailed information, including tick data that captures every order during trading sessions, allowing for deeper analysis of market trends and trading signals [6][9]. Group 2: Key Period Factors - The article introduces a set of Level 2 factors based on key trading periods, categorized into four main types: price changes, price levels, trading amounts, and volume-price coordination, totaling 123 factors [12]. - Specific factors such as KeyPeriod_ret_zero and KeyPeriod_ret_low5pct show historical RankIC averages of -5.36% and 5.47% respectively, with win rates of 85.1% and 84.1% [2]. Group 3: Factor Performance - The performance of various factors is highlighted, with low price period factors like KeyPeriod_price_low5pct achieving a 20-day RankIC average of 5.59% and a win rate of 85.3% [2]. - Trading amount factors such as KeyPeriod_amount_top30pct show a 20-day RankIC average of 11.23% with a win rate of 84.8%, indicating strong predictive power [2]. Group 4: Research and Development - The article outlines ongoing research efforts to refine and develop new factors from Level 2 data, with a focus on enhancing the predictive capabilities of trading strategies [10][12]. - Previous reports have successfully identified effective factors, with some achieving historical RankIC averages above 9.2% and win rates around 76% [10].

Group 1 - The company is recruiting interns for positions in Shenzhen, Shanghai, and Beijing, requiring in-person internships with a minimum commitment of three days per week for at least three months [1] - The application deadline for submitting resumes is December 31, 2025 [1] - Interns with outstanding performance may have the opportunity for full-time employment after the internship [1] Group 2 - Responsibilities include data processing, analysis, and assisting researchers with quantitative investment projects [2] - Interns will also assist in the development and tracking of financial engineering strategy models [2] - Additional tasks may be assigned by the team [2] Group 3 - Basic requirements include being a master's or doctoral student in STEM fields or financial engineering, with a strong preference for exceptional fourth-year students [3] - Proficiency in programming languages such as Python and familiarity with SQL databases are essential [3] - Candidates should possess strong self-motivation, analytical skills, and effective communication abilities [3] Group 4 - Preferred qualifications include a solid foundation in financial markets, familiarity with key concepts in stocks, bonds, futures, indices, and funds [4] - A strong mathematical background, research project experience, and published academic papers in SCI or EI are advantageous [4] - Familiarity with financial terminals like Wind, Bloomberg, and Tianruan, as well as knowledge of machine learning and deep learning, is a plus [4] Group 5 - Interested candidates should submit their resumes in PDF format to the specified email address, following a specific naming convention for the email subject [5] - Resumes not adhering to the naming format will be treated as spam [5] - Qualified candidates will be contacted for written tests and interviews after the resume collection deadline [5]

Core Viewpoint - The overall macro analysis indicates a bearish outlook for equity assets, while technical analysis shows an upward trend with moderate valuation and capital outflow [1][2][8] - For bonds, the macro perspective is bullish, and the technical trend is also upward [1][2][8] - Industrial products are viewed negatively from a macro standpoint, with a downward price trend technically [1][2][8] - Gold assets are favored in the macro analysis, with an upward price trend technically [1][2][8] Macro Analysis - The macro analysis categorizes assets based on their performance under different macro indicators, indicating that equity assets are currently under pressure, while bonds and gold are favored [4][8] - The analysis employs T-tests to assess the impact of macro indicators on asset returns, revealing significant differences in average returns based on the trend of macro indicators [4][5] Technical Analysis - The technical analysis utilizes closing prices and various indicators to assess asset trends, with equity, bonds, and gold showing upward trends, while industrial products are on a downward trend [10][13] - The latest trend indicators for equity and bond assets are positive, while industrial products show a negative trend [14][13] Valuation Indicators - The equity risk premium (ERP) for the CSI 800 index is at 55.71%, indicating a moderate valuation level [17][18] - The analysis of capital flow indicates a net outflow of 102.9 billion yuan for equity assets, suggesting a negative sentiment in the market [21][22] Asset Allocation Performance Tracking - Historical performance data shows that a fixed ratio combined with macro and technical indicators yielded a return of 10.50% for 2025, with an annualized return of 12.00% since April 2006 [3][26] - Different asset allocation strategies, including volatility control and risk parity, have also been analyzed, showing varying returns and risk profiles [30][33] Summary of Views - The combined scores from macro and technical indicators suggest a bearish outlook for equity assets, a bullish stance for bonds and gold, and a negative view for industrial products [23][25]

Core Viewpoint - The quantitative convertible bond portfolio experienced a slight decline in November, with a year-to-date return of 20.14% and an excess return of 3.96% [1] Group 1: Portfolio and Performance - The quantitative convertible bond portfolio is generated based on three factor systems: fundamental factors, low-frequency price-volume factors, and high-frequency price-volume factors [5] - The portfolio's performance in November showed a return of -0.72% and an excess return of -0.03% [1] Group 2: Convertible Bond Factors - A total of 32 fundamental factors, 80 low-frequency price-volume factors, and 32 high-frequency price-volume factors are tracked for convertible bonds [2] - The report illustrates the latest data using the pricing deviation factor as an example [2] Group 3: Risk Warnings for Convertible Bonds - The report provides risk warnings for convertible bonds based on forced delisting rules and credit scoring methods, highlighting various risks including trading and financial delisting risks [3][13] Group 4: Timing for Convertible Bond Index - The report employs price-volume models, pricing deviations, and convertible bond elasticity for timing and position management of the CSI Convertible Bond Index, indicating a bullish signal for the end of November with a position recommendation of 1/3 [4][14] Group 5: Timing Signals - The timing signals for the CSI Convertible Bond Index from early November show a mix of bullish and neutral signals, with a position recommendation fluctuating between 0% and 67% throughout the month [15]