Quantitative Models and Construction Methods Index Enhancement Strategies - **Model Name**: Index Enhancement Strategy for CSI 300, CSI 500, CSI 1000, and CSI 2000 - **Model Construction Idea**: The strategy is based on a multi-factor stock selection model, leveraging an equity factor library to identify effective factors within the constituent stocks of the respective indices[77] - **Model Construction Process**: - **Factor Selection**: Hundreds of factors from the equity factor library are screened for effectiveness within the constituent stocks of CSI 300, CSI 500, CSI 1000, and CSI 2000 indices[77] - **Portfolio Optimization**: - For CSI 300: Strict sector and market capitalization neutrality, individual stock weight capped at 8%, and weight deviation capped at 3%[77] - For CSI 500: Strict sector and market capitalization neutrality, individual stock weight capped at 1%, and weight deviation capped at 1%[77] - For CSI 1000 and CSI 2000: Market capitalization deviation capped at 0.5 standard deviations, sector deviation capped at 2.5%, individual stock weight capped at 1% for CSI 1000 and 0.5% for CSI 2000[77] - **Rebalancing**: Weekly tracking of the performance of the index enhancement strategy within the constituent stocks[77] Model Evaluation - **Evaluation**: The strategy effectively utilizes a multi-factor approach to enhance index performance while maintaining sector and market capitalization neutrality. However, the strategy's performance is subject to transaction costs and historical data limitations[77][83] --- Model Backtesting Results CSI 300 Index Enhancement Strategy - **Weekly Return**: 1.53% (Index Return: 1.64%, Excess Return: -0.12%)[78] - **Monthly Return**: -3.31% (Index Return: -2.46%, Excess Return: -0.85%)[78] - **Year-to-Date Return**: 21.83% (Index Return: 15.04%, Excess Return: 6.8%)[78] - **Maximum Drawdown of Excess Return**: -3.15%[78] CSI 500 Index Enhancement Strategy - **Weekly Return**: 2.97% (Index Return: 3.14%, Excess Return: -0.17%)[78] - **Monthly Return**: -4.54% (Index Return: -4.08%, Excess Return: -0.46%)[78] - **Year-to-Date Return**: 23.41% (Index Return: 22.81%, Excess Return: 0.61%)[78] - **Maximum Drawdown of Excess Return**: -4.77%[78] CSI 1000 Index Enhancement Strategy - **Weekly Return**: 3.77% (Index Return: 3.77%, Excess Return: 0%)[83] - **Monthly Return**: -2.59% (Index Return: -2.3%, Excess Return: -0.29%)[83] - **Year-to-Date Return**: 35.59% (Index Return: 23.1%, Excess Return: 12.49%)[83] - **Maximum Drawdown of Excess Return**: -5.59%[83] CSI 2000 Index Enhancement Strategy - **Weekly Return**: 4.38% (Index Return: 4.99%, Excess Return: -0.61%)[83] - **Monthly Return**: -0.03% (Index Return: -0.4%, Excess Return: 0.37%)[83] - **Year-to-Date Return**: 59.74% (Index Return: 31.65%, Excess Return: 28.08%)[83] - **Maximum Drawdown of Excess Return**: -5.23%[83] --- Quantitative Factors and Construction Methods Single Factors - **Factor Name**: Analyst Forecast ROE-FY3 - **Construction Idea**: Measures the expected return on equity (ROE) for the next three fiscal years as forecasted by analysts[33] - **Construction Process**: Derived from analyst consensus estimates for ROE over the next three fiscal years[33] - **Evaluation**: Demonstrates strong predictive power for stock selection, particularly in CSI 300 and CSI 2000 stock pools[33][36] - **Factor Name**: Standardized Unexpected Quarterly ROE with Drift - **Construction Idea**: Captures the deviation of actual quarterly ROE from expectations, adjusted for drift[35] - **Construction Process**: - Calculate the unexpected component of quarterly ROE - Standardize the values and adjust for drift to account for temporal effects[35] - **Evaluation**: Effective in identifying outperforming stocks, particularly in CSI 1000 and CSI 2000 stock pools[35][36] - **Factor Name**: One-Month Price Change - **Construction Idea**: Reflects short-term momentum by measuring the percentage change in stock price over the past month[36] - **Construction Process**: Calculate the percentage change in stock price over the last 30 days[36] - **Evaluation**: Demonstrates strong performance in CSI 2000 and CSI 1000 stock pools, indicating momentum effects[36] Factor Neutralization - **Neutralization Process**: - Apply absolute median method for outlier removal - Perform Z-score standardization - Conduct cross-sectional regression using log market capitalization and industry dummy variables as independent variables, with the factor as the dependent variable - Use the residuals as the neutralized factor values[32] --- Factor Backtesting Results CSI 300 Stock Pool - **Top Factors (Year-to-Date Excess Return)**: - Single-Quarter Revenue Growth Rate: 25.24%[33] - Single-Quarter ROE: 22.28%[33] - Single-Quarter ROA Change: 22.21%[33] CSI 500 Stock Pool - **Top Factors (Year-to-Date Excess Return)**: - Analyst Forecast Net Profit Growth Rate FY3: 14.53%[34] - Analyst Forecast Revenue Growth Rate FY3: 13.69%[34] - Analyst Forecast Revenue FY3 120-Day Change: 12.81%[34] CSI 1000 Stock Pool - **Top Factors (Year-to-Date Excess Return)**: - Standardized Unexpected Quarterly ROE with Drift: 19.18%[35] - Analyst Forecast ROE-FY3 120-Day Change: 18.4%[35] - Standardized Unexpected Quarterly Net Profit with Drift: 18.34%[35] CSI 2000 Stock Pool - **Top Factors (Year-to-Date Excess Return)**: - 90-Day Report Upward Revision Ratio: 25.01%[36] - Standardized Unexpected Quarterly Net Profit with Drift: 24.46%[36] - 5-Minute Volume Skewness: 23.74%[36] CSI All-Share Stock Pool - **Top Factors (Year-to-Date Excess Return)**: - Analyst Forecast ROE-FY3 120-Day Change: 23.52%[37] - Single-Quarter Revenue Growth Rate: 20.47%[37] - Analyst Forecast Revenue Growth Rate FY3: 19.35%[37]

Investment Rating - The report assigns an "Accumulate" rating to the company with a target price of 14.10 CNY [4][10]. Core Insights - The company's main automotive parts business is experiencing steady growth while also expanding into humanoid robotics and low-altitude economy sectors. The report anticipates that the company will maintain stable growth in its core business and potentially open new growth avenues [2][10]. - Revenue projections for 2025, 2026, and 2027 are estimated at 20.872 billion CNY, 22.062 billion CNY, and 23.165 billion CNY, respectively, reflecting year-on-year growth rates of 6.3%, 5.7%, and 5.0% [3][20]. - The net profit attributable to the parent company is forecasted to be 630 million CNY, 699 million CNY, and 775 million CNY for 2025, 2026, and 2027, indicating growth rates of 23.8%, 10.9%, and 10.9% [3][20]. Financial Summary - Total revenue for 2023 is reported at 20,766 million CNY, with a projected decrease to 19,636 million CNY in 2024, followed by a recovery in subsequent years [3]. - The net profit for 2023 is 521 million CNY, with a slight decline expected in 2024 to 509 million CNY, before increasing in the following years [3]. - The earnings per share (EPS) for 2023 is 0.70 CNY, projected to rise to 1.04 CNY by 2027 [3]. Business Segment Forecast - The automotive interior segment is expected to generate revenues of 115.89 billion CNY, 121.69 billion CNY, and 127.77 billion CNY for 2025, 2026, and 2027, respectively, with a consistent growth rate of 5% [14]. - The automotive bumper segment is projected to achieve revenues of 46.71 billion CNY, 49.04 billion CNY, and 51.50 billion CNY over the same period, also reflecting a 5% growth rate [15]. - The lighting segment is forecasted to see revenues of 14.81 billion CNY, 16.29 billion CNY, and 17.10 billion CNY, with growth rates of 15%, 10%, and 5% respectively [16]. Valuation - The report suggests a valuation based on a price-to-earnings (PE) ratio of 15 times for 2026, leading to a target price of 14.10 CNY, which is below the average PE of comparable companies at 20.95 times [20][21]. - The company is positioned favorably due to its strong customer relationships and proactive expansion into new technology sectors, which supports its growth outlook [10][20].

Investment Rating - The report assigns an "Accumulate" rating for the industry, indicating a potential increase of over 15% relative to the CSI 300 index [4][9]. Core Insights - The release of DeepSeek-V3.2 and its enhanced version V3.2-Speciale marks a significant breakthrough in reasoning capabilities, tool invocation, and the open-source ecosystem, promoting the prosperity of large model open-source and developer ecosystems [2][4]. - DeepSeek-V3.2 achieves top-tier performance comparable to closed-source models, particularly in reasoning capabilities, and integrates innovative thinking modes into tool invocation, providing more efficient and cost-effective solutions for AI application development [4]. - The Speciale version has excelled in international competitions, achieving second place in the ICPC and demonstrating the potential of open-source models to reach human-level intelligence in complex reasoning tasks [4]. Summary by Sections Industry Overview - The report highlights the computer industry, focusing on advancements in AI and large models, particularly the DeepSeek series [4]. Investment Recommendations - The investment suggestion emphasizes that the release of DeepSeek-V3.2 signifies a new phase in the performance and practicality of open-source large models, with a balanced focus on performance and efficiency [4]. Technological Advancements - DeepSeek-V3.2 incorporates a systematic approach to chain thinking within tool invocation processes, significantly enhancing the model's generalization and execution capabilities in complex scenarios [4]. - The model has undergone reinforcement learning across over 85,000 complex instructions in more than 1,800 environments, achieving the highest level among open-source models in tool invocation assessments [4]. Ecosystem Development - The comprehensive upgrade of DeepSeek-V3.2's open-source and API services is expected to accelerate technological penetration and drive transformative changes in industry applications [4]. - The open strategy is anticipated to attract numerous developers to build vertical applications based on DeepSeek, further solidifying its leading position in the open-source domain [4].

Group 1 - The report highlights that Japan's bond market experienced its most severe sell-off since 1999, driven by a combination of fiscal expansion, central bank policy shifts, and supply-demand imbalances [6][7][8] - The Japanese government's economic stimulus plan of 21.3 trillion yen (approximately 3.5% of GDP) raised concerns about debt sustainability, leading to increased selling pressure in the bond market [6][7] - The Bank of Japan's reduction in long-term bond purchases exacerbated supply pressures, with the 30-year bond yield reaching a historic high of 3.26% [7][8] Group 2 - The report identifies common characteristics of global bond market volatility, noting that developed markets have also experienced significant adjustments in response to central bank policy signals [11][12] - In the UK, a crisis of fiscal credibility led to a surge in 30-year gilt yields to the highest levels since 1998, reflecting concerns over government debt sustainability [12] - Australia's bond market saw a sharp increase in yields following unexpected inflation data, indicating a shift in market expectations regarding interest rate movements [13][15] Group 3 - The report discusses the vulnerabilities of emerging markets, highlighting that their bond markets are particularly sensitive to changes in central bank policies, leading to amplified volatility [20][21] - Argentina's recent crisis exemplifies this vulnerability, with a significant rise in sovereign debt risk premiums amid concerns over fiscal sustainability [21][22] - The report notes that emerging markets face challenges due to shallow liquidity and reliance on foreign capital, which can lead to rapid capital outflows in response to policy shifts [20][23] Group 4 - The report emphasizes the importance of balancing fiscal expansion, central bank operations, and market absorption capacity in the context of Japan's bond market [28][29] - It suggests that while Japan's experience offers lessons, significant differences exist in capital account management and monetary policy tools between Japan and other countries [28][29] - The report warns that ongoing fiscal stimulus in China could lead to reassessments of long-term interest rate levels, particularly if nominal growth does not meet expectations [28][30] Group 5 - The report outlines potential scenarios for Japan's bond market, particularly in light of the upcoming Bank of Japan policy meeting, where tensions between fiscal stimulus and monetary tightening may influence market reactions [33][34] - It notes that the yield curve could steepen if interest rate hikes materialize, but economic data surprises could limit long-term yield increases [34][35] - The report highlights the differentiated risk profiles of various bond maturities, with longer-duration bonds facing greater price volatility in a low liquidity environment [35][36]

Investment Rating - The report recommends an "Accumulate" rating for leading energy storage companies, specifically Haibo Sichuang and Sunshine Power, along with related companies such as Canadian Solar and Xidian New Energy [5]. Core Insights - The development of AIDC (Artificial Intelligence Data Center) may exacerbate electricity shortages in the U.S., with data center energy storage serving as a short-term solution for peak shaving and frequency regulation, while potentially becoming a self-sufficient power source in the long term [2][3]. - Fluence is currently negotiating over 30 GWh of AIDC energy storage projects, with 80% of these projects initiated after the end of Q4 2025, indicating a significant emerging market opportunity [3][4]. - The energy consumption of data centers in the U.S. is projected to grow significantly, with estimates suggesting an increase from 176 TWh in 2023 to between 325-580 TWh by 2028, which will raise their share of total U.S. electricity consumption from 4.4% to between 6.7% and 12% [3][4]. Summary by Sections AIDC Development and Energy Demand - AIDC's high energy consumption could lead to increased electricity shortages in the U.S. According to the Department of Energy (DOE), data center electricity demand is expected to grow annually by 13%-27% from 2023 to 2028 [4]. - If 50 GW of new data center capacity is added by 2030, the projected electricity gap could reach 23 GW, potentially larger when considering the retirement of existing power plants [4]. Energy Storage Solutions - Short-term energy storage solutions are beneficial for data centers to manage power fluctuations and facilitate grid connection, with the current grid connection process taking several years [4]. - Long-term, solar and storage solutions may evolve into self-sufficient power sources for data centers, with the economic viability of solar storage already being demonstrated [4]. Company Recommendations - The report highlights the potential of long-duration energy storage (6-8 hours) as an emerging opportunity, particularly in markets with high renewable energy penetration like Europe and California [3][4]. - The report emphasizes the advantages of solar storage over gas turbines, particularly in terms of connection timelines and economic feasibility [4].

Investment Rating - The report maintains a rating of "Buy" for the company [5] Core Views - The new tax policy is expected to benefit compliant leading brands in the gold market, with the company poised to increase its market share under this policy [2][11] - The company is projected to achieve significant revenue growth driven by rising gold prices and increased investment demand [16][20] Financial Summary - Total revenue is forecasted to grow from 16,552 million yuan in 2023 to 31,804 million yuan in 2027, with a compound annual growth rate (CAGR) of approximately 9.9% [4][21] - Net profit attributable to the parent company is expected to rise from 707 million yuan in 2023 to 1,023 million yuan in 2027, reflecting a CAGR of about 8.5% [4][21] - Earnings per share (EPS) is projected to increase from 0.91 yuan in 2023 to 1.32 yuan in 2027 [4][21] Revenue Forecast - The company’s revenue from precious metal investment products is expected to grow significantly, with projections of 18,068.91 million yuan in 2025, representing a 40% increase [16][17] - Revenue from gold jewelry is anticipated to grow at a slower pace, with estimates of 6,020.28 million yuan in 2025, reflecting a 10% increase [16][17] - The company is expected to maintain a high dividend payout ratio, exceeding 75% [11][20] Market Position and Strategy - The company operates as a member of the Shanghai Gold Exchange, allowing it to directly procure gold and sell it through a fully owned retail model, which is less affected by the new tax policy [11][42] - The company is expanding its retail presence, with a total of 103 stores by mid-2025, covering key cities including Beijing, Tianjin, and others [11][20] - The new tax policy is expected to enhance the company's competitive pricing advantage in the investment gold market, potentially attracting customers from other brands [11][42] Valuation - The target price for the company is set at 19.26 yuan, based on a price-to-earnings (P/E) ratio of 18x for 2025, which is slightly above the industry average [22]

Investment Rating - The report assigns a rating of "Buy" for the company, with a target price of HKD 3.91, corresponding to a 20x P/E for 2025 [7][11]. Core Insights - The company is expected to maintain high growth in performance due to an increase in the proportion of used car financing, a successful transition to a light asset model driving SaaS business growth, and the application of AI in the automotive industry [3][11]. - The company’s used car financing transactions reached 235,000 in Q3 2025, a year-on-year increase of 22.6%, with financing amounts totaling CNY 21.2 billion, and used car credit growth of 51.3% to CNY 12.1 billion, increasing its share to 56.9% [11]. - The SaaS business facilitated financing of CNY 11.4 billion in Q3 2025, a 102% year-on-year increase, contributing 53.7% to total financing [11]. - The company is expected to see continued improvement in performance in the second half of 2025, with AI products like "X Call" enhancing efficiency in credit applications and customer management [11]. Financial Summary - Total revenue projections for 2025-2027 are CNY 115.99 billion, CNY 129.40 billion, and CNY 144.39 billion, representing year-on-year growth of 17%, 12%, and 12% respectively [11]. - Net profit estimates for the same period are CNY 11.86 billion, CNY 14.59 billion, and CNY 18.16 billion, with growth rates of 46%, 23%, and 24% respectively [11]. - The company’s P/E ratios are projected to be 6.55 for 2024, increasing to 15.36 in 2025, and then decreasing to 10.03 by 2027 [11].

Investment Rating - The report maintains an "Overweight" rating for the retail chain pharmacy industry [1][3]. Core Insights - The report highlights that leading pharmacies are beginning to recover from a challenging phase, with a focus on the growth potential in 2026 driven by both organic and external factors [2][3]. - The retail pharmacy market in China is projected to reach a retail scale of 611.9 billion yuan in 2024, reflecting a year-on-year decline of 1.8% due to factors such as population decrease and changes in healthcare insurance [3]. - The report identifies key companies to watch, including Yifeng Pharmacy, Dazhenglin, and Laobaixing, while also suggesting attention to Yixintang, Jianzhijia, and Shuyupingmin [3]. Summary by Sections Market Overview - The retail pharmacy market is experiencing a contraction in supply due to increased competition, with the number of physical pharmacies exceeding 700,000 in 2024, a 60% increase since 2014 [3]. - The report notes a recovery trend in the market, with a sequential growth of 6.7% in September 2025, despite a year-on-year decline of 1.9% in the first three quarters of 2025 [3]. Demand and Supply Dynamics - The demand side is showing signs of recovery, particularly in essential medicine categories, with a 6.9% sequential growth in drug retail sales in September 2025 [3]. - The report indicates that the concentration of leading pharmacy chains is expected to increase as smaller chains face operational challenges, leading to closures [3]. Future Outlook - The report anticipates that leading pharmacy chains will achieve recovery through both internal growth driven by rising flu cases and external growth via acquisitions, with a current low penetration of direct stores at about 1.5% [3][4]. - The long-term outlook remains positive for industry concentration, with expectations for continued growth in the leading pharmacy chains [3].

| | | [Table_Report] 相关报告 服务消费景气提升，科技硬件延续涨价 [Table_Authors] 方奕(分析师) 中观景气 12 月第 1 期 本报告导读： 中观景气延续分化的增长格局，新兴科技景气仍强，高性能存储价格延续快速上涨， 游戏供给偏宽松；服务消费景气明显提升，地产周期和耐用品需求仍承压。 投资要点： | | 021-38031658 | | --- | --- | | | fangyi2@gtht.com | | 登记编号 | S0880520120005 | | | 陶前陈(研究助理) | | | 0755-23976164 | | | taoqianchen@gtht.com | | 登记编号 | S0880125070014 | | | 张逸飞(分析师) | 资产概览：银价铜价创历史新高 2025.12.01 外资与融资资金重回流入 2025.12.01 消费景气线索增多，科技制造延续增长 2025.11.26 ETF 流入明显，融资资金与外资有所流出 2025.11.24 资产概览：美元兑日元逼近 160 关口 2025.11.24 策 略 研 究 告 请务必阅读正文 ...

Report Industry Investment Rating No relevant content provided. Core Viewpoints of the Report - The timing model constructed in a "scenario-based combat" approach in 2025 performed averagely due to factors such as unstable effectiveness, multicollinearity, and high volatility. A new volume-price factor timing model based on a grouping algorithm was reconstructed, focusing on optimizing three major issues: unstable effectiveness, large signal fluctuations, and insufficient reliability of single signals [4][7]. - By using double standards of grouped IC and thresholds in factor screening, factors that can stably play a predictive role in both high and low value intervals were selected, ensuring the effectiveness of model information from the source [4][10]. - Through the strategies of "random grouping + independent training" and "rolling smoothing + multi - group voting", noise was filtered, effective information was aggregated, and accurate and robust timing signals were generated [4]. - The back - test results showed that the model significantly outperformed the benchmark, especially the 1 - day signal, which demonstrated strong stable timing ability both within and outside the sample [4]. Summary According to the Directory 1. Factor Screening: Double Standards Determine Stability and Effectiveness - **Factor Reserve**: 87 factors covering intraday patterns, price fluctuations, trading volume statistics, trends, momentum, overbought and oversold conditions were constructed around the volume - price characteristics of Treasury bond futures, comprehensively covering core dimensions for timing [11]. - **Screening Standard**: An annual rolling back - test framework was adopted, with a 3 - year data window each year. Factors were sorted and divided into 5 groups. Two conditions were set: at least 4 groups of IC should maintain the same direction, and the average absolute value of the same - direction IC should be no less than 0.05. This mechanism could identify factors with stable cross - sectional prediction ability and adapt to market changes [13][14]. 2. Model Building and Signal Generation: Group Voting Based on Random Grouping and Cross - Validation - **Model Building**: A deep - learning architecture of bidirectional multi - layer GRU + attention mechanism was used, focusing on short - term timing requirements for T + 1 and T + 5 day price movements. Techniques such as Dropout and layer normalization were introduced to avoid overfitting, and the "rolling window" training mode was adopted [16]. - **Random Grouping and Signal Generation** - **Random Grouping of Factors and Parallel Training**: After selecting effective factors based on grouped IC values each year, they were randomly divided into multiple groups. Each group of factors was used to train an independent GRU sub - model, generating diverse prediction results [20]. - **Signal Generation**: The final timing signal was generated through "rolling smoothing + multi - group voting". Rolling smoothing was used to filter noise according to different prediction periods, and multi - group voting was used to confirm the signal direction, reducing the influence of single - sub - model prediction errors. A comprehensive signal was also generated by integrating 1 - day and 5 - day prediction results [21][22]. 3. Strategy Back - test Design and Parameter Selection: Balancing Robustness and Adaptability - **Strategy Back - test Design**: A long - short full - position trading mode was set. When the final signal was 1, a full - position long was taken; when it was - 1, a full - position short was taken; when it was 0, the position remained unchanged. Transaction costs and slippage were ignored, and the closing price of the 10 - year Treasury bond futures main contract was used as the benchmark [25]. - **Parameter Selection**: Three major dimensions of parameters were focused on: training window, number of groups, and number of factors in each group. Different parameter combinations were tested from 2019 to September 2025, and the optimal parameter combinations were selected based on the back - test results, emphasizing overall return - capturing ability and stable advantages over the benchmark in most years [26]. - **Back - test Performance**: The 1 - day signal performed outstandingly both within and outside the sample. Within the sample, the 1 - day signal had an average annualized return of 3.61% and a Sharpe ratio of 1.12; outside the sample, from the National Day holiday in 2025 to the present, the 1 - day signal had a cumulative return of 0.99%, a Sharpe ratio of 5.98, and a maximum drawdown of only 0.13%. The 5 - day signal's adaptability decreased outside the sample [30][34].