声 明 跟踪评级安排 广发证券股份有限公司 2026 年面向专业投资者公开发行 短期公司债券（第一期）信用评级报告 广发证券股份有限公司 2026 年面向专业投资者公开发行 短期公司债券（第一期）信用评级报告 中诚信国际信用评级有限责任公司 | 编号:CCXI-20260108D-01 中诚信国际信用评级有限责任公司 2026 年 1 月 15 日 2 本次评级为委托评级，中诚信国际及其评估人员与评级委托方、评级对象不存在任何其他影响本次评级行为独立、 客观、公正的关联关系。 本次评级依据评级对象提供或已经正式对外公布的信息，以及其他根据监管规定收集的信息，中诚信国际按照相关 性、及时性、可靠性的原则对评级信息进行审慎分析，但中诚信国际对于相关信息的合法性、真实性、完整性、准 确性不作任何保证。 中诚信国际及项目人员履行了尽职调查和诚信义务，有充分理由保证本次评级遵循了真实、客观、公正的原则。 评级报告的评级结论是中诚信国际依据合理的内部信用评级标准和方法、评级程序做出的独立判断，未受评级委托 方、评级对象和其他第三方的干预和影响。 本评级报告对评级对象信用状况的任何表述和判断仅作为相关决策参考之用，并不意味 ...

（住所：广东省广州市黄埔区中新广州知识城腾飞一街 2 号 618 室） （股票简称：广发证券；股票代码：000776.SZ、1776.HK） 2026 年面向专业投资者公开发行短期公司债券（第一期） 募集说明书 | 发行人： | 广发证券股份有限公司 | | --- | --- | | 牵头主承销商、受托管理人： | 华泰联合证券有限责任公司 | | 联席主承销商、簿记管理人： | 平安证券股份有限公司 | | 受托管理人： | 华泰联合证券有限责任公司 | | 发行金额： | 不超过人民币 30 亿元（含） | | 增信措施情况： | 无担保 | | 信用评级结果： | 主体评级：AAA；评级展望：稳定；债项评级：A-1 | | 信用评级机构： | 中诚信国际信用评级有限责任公司 | 牵头主承销商/债券受托管理人 （住所：深圳市前海深港合作区南山街道桂湾五路 128 号前海深港基金小镇 B7 栋 401） 联席主承销商/簿记管理人 （深圳市福田区福田街道益田路 5023 号平安金融中心 B 座第 22-25 层） 签署日期： 年 月 日 广发证券股份有限公司 2026 年面向专业投资者公开发行短期公司债券（ ...

Core Viewpoint - Guangfa Securities has received approval from the China Securities Regulatory Commission (CSRC) to publicly issue perpetual subordinated bonds totaling up to 20 billion yuan to professional investors [1]. Group 1: Stock Performance - As of January 20, 2026, Guangfa Securities closed at 22.77 yuan, an increase of 1.52% from the previous trading day [1]. - The stock opened at 22.38 yuan, reached a high of 22.92 yuan, and a low of 22.35 yuan, with a trading volume of 1.337 billion yuan and a turnover rate of 1.0% [1]. Group 2: Bond Issuance - The company has been approved to issue perpetual subordinated bonds with a total face value not exceeding 20 billion yuan [1]. - The issuance must strictly follow the prospectus submitted to the Shenzhen Stock Exchange and is valid for 24 months from the date of approval [1]. - The company is required to report any significant events during the registration period and fulfill its information disclosure obligations [1].

Core Viewpoint - The report from GF Securities indicates a significant increase in the total output of market pigs by listed companies in 2025, with a year-on-year growth of 25% to 111.53 million heads, and a 30% increase to 90.39 million heads when excluding Muyuan Foods [1][3]. Group 1: Market Output - In December 2025, the total output of market pigs from listed companies reached 19.05 million heads, reflecting a month-on-month increase of 7.1% and a year-on-year increase of 11.3% [2]. - The output of market pigs from listed companies, excluding Muyuan Foods, was 12.07 million heads in December, with a month-on-month growth of 8.0% and a year-on-year growth of 35.7% [2][3]. - Major companies such as Muyuan Foods, Wens Foodstuff Group, New Hope Liuhe, and Dekang Agriculture showed varying month-on-month growth rates in December, with increases of 5.7%, 4.8%, 15.4%, and 4.1% respectively [3]. Group 2: Piglet Sales and Prices - The overall sales volume of piglets from listed companies saw a significant increase in 2025, with a notable rise in the proportion of piglet sales [2][4]. - The price of 7 kg piglets has rebounded to 307 RMB per head, attributed to the upcoming replenishment season and positive market sentiment regarding pig prices in the second half of 2026 [1][5]. - The average selling price of pigs in December was estimated at 11.53 RMB per kilogram, showing a month-on-month decline of 1.4% [4]. Group 3: Company Performance - In 2025, major companies reported the following cumulative outputs: Muyuan Foods at 77.98 million heads (+19%), Wens Foodstuff Group at 40.48 million heads (+34%), New Hope Liuhe at 17.55 million heads (+6%), and Dekang Agriculture at 10.83 million heads (+23%) [3]. - Smaller companies like Tangrenshen, Tiankang Biological, Shennong Group, and Juxing Agriculture also reported varying outputs, with Tangrenshen showing a year-on-year increase of 23% [3]. Group 4: Industry Outlook - The industry is currently facing cumulative losses, which may lead to continued reduction in pig production capacity [1][5]. - The breeding sow inventory decreased by 0.22% month-on-month in December, indicating potential challenges in production [5].

Market Overview - On January 20, the securities sector rose by 0.42% compared to the previous trading day, with Dongwu Securities leading the gains [1] - The Shanghai Composite Index closed at 4113.65, down 0.01%, while the Shenzhen Component Index closed at 14155.63, down 0.97% [1] Individual Stock Performance - Dongwu Securities (601555) closed at 9.29, up 2.31% with a trading volume of 945,100 shares and a transaction value of 873 million [1] - Huatai Securities (601688) closed at 23.26, up 1.88% with a trading volume of 860,000 shares and a transaction value of 1.9957 billion [1] - GF Securities (000776) closed at 22.77, up 1.52% with a trading volume of 588,800 shares and a transaction value of 1.337 billion [1] - Other notable performers include Xiangcai Co. (600095) up 1.15%, Huazheng Securities (600909) up 0.99%, and Guohai Securities (000750) up 0.94% [1] Fund Flow Analysis - The securities sector saw a net inflow of 480 million from institutional investors, while retail investors experienced a net outflow of 354 million [2] - The overall fund flow indicates a mixed sentiment, with institutional buying contrasting with retail selling [2] Detailed Fund Flow for Selected Stocks - Zhaoyuan Group (601211) had a net inflow of 281 million from institutional investors, while retail investors saw a net outflow of 184 million [3] - CITIC Securities (600030) experienced a net inflow of 197 million from institutional investors, with retail investors having a net outflow of 1.746 million [3] - Huatai Securities (601688) recorded a net inflow of 123 million from institutional investors, while retail investors had a net outflow of 1.39 million [3] - Other stocks like Zhongjin Company (566109) and China Merchants Securities (600999) also showed significant net inflows from institutional investors [3]

Group 1 - The core viewpoint is that the industry has a vast long-term market space due to the active layout of space computing by China and the US, combined with the cost and performance advantages of space computing itself [3][4] - Space computing is transitioning from a "ground-based calculation" model to a "space-based calculation" model, allowing for direct data processing in space [1][3] Group 2 - Space computing has operational cost advantages, with a significant focus on marginal energy costs, which are the core factor in overall operational expenses [2] - For example, a single space-based 40MW computing cluster can operate for 10 years at a total cost of $8.2 million, saving approximately $159 million compared to traditional computing clusters, with over $130 million saved in marginal energy costs [2] Group 3 - The demand for solar wings is expected to increase due to the expansion of power and area requirements driven by space computing, leading to the adoption of flexible technology routes [4] - Flexible solar wings can achieve a weight reduction of 20%-40%, a storage volume reduction of over 60%, and improved performance, making them a key component in the power system [4] Group 4 - Investment recommendations include focusing on companies related to space photovoltaics, such as: - Maiwei Co., Ltd. (300751.SZ), which is expected to become a core equipment supplier for space computing photovoltaic segments [5] - Gaomei Co., Ltd. (688556.SH), which aligns with the cost reduction route for space photovoltaics [5] - Jiejia Weichuang (300724.SZ), which is positioned to benefit from the expansion of flexible solar wings in the space computing sector [5]

智通财经APP获悉，广发证券发布研报称，3D打印在火箭制造领域应用优势明显，目前，3D打印技术 多用于火箭发动机环节。近年来我，国商业航天发展加速推进，3D打印作为火箭制造核心工艺，将充 分受益于商业航天β带来的运力需求提升。同时，3D打印技术在火箭制造环节中应用比例提升已是可预 见趋势，若仅考虑火箭发动机制造环节，3D打印技术市场规模已超4000亿元市场空间。 广发证券主要观点如下： 3D打印已从火箭制造领域的"可选技术"升级为"必选工艺" 3D打印作为制造异形复杂、多尺度、整体化结构件的理想技术路径，凭借"增材制造"底层逻辑，显著 突破了传统制造及铸造等工艺在负载结构加工中的瓶颈。目前，3D打印已在众多火箭型号中获得应 用，成为火箭制造中必选工艺。 3D打印技术具备效率、成本和性能三重技术优势 3D打印技术在火箭制造尤其是发动机环节中的应用可以在效率、成本和性能三维度方面实现大幅提 升，具备显著技术优势。综合多个案例的单项数值最优效果来看，3D打印技术可以使火箭发动机重量 大幅度减少，制造周期可缩短70%以上，发动机成本下降能达90%，其他综合性能得到显著提升。 3D打印技术应用渗透率提升已成确定趋势 受 ...

Core Insights - The development of AI data centers in the U.S. is driving an increase in electricity demand, leading to a surge in gas turbine demand due to power shortages [1][2] - The delivery time for newly ordered H-class gas turbines has significantly lengthened, with expected delivery now between 2028 and 2030 [1][2] - The commercial operation cycle for modified aircraft engines is much shorter than that of large gas turbines, making them an ideal transitional solution for data center construction [4] Group 1: AI Data Centers and Electricity Demand - The global electricity consumption of data centers is projected to grow from 49 GW in 2023 to 96 GW by 2026, with 90% of this growth driven by AI [2] - Aging power grid infrastructure in developed economies, with over 50% of equipment exceeding 20 years of use, is prompting a need for upgrades [2] Group 2: Gas Turbine Demand and Supply - The demand for gas turbines is increasing due to electricity shortages, resulting in a higher order-to-delivery ratio for turbine manufacturers [1][2] - The supply-demand mismatch is evident, with a significant backlog in orders for modified aircraft engines, as seen in recent contracts and deliveries [4] Group 3: Investment Opportunities - The current supply-demand mismatch in modified aircraft engines presents opportunities for companies with supporting technologies and capacities to secure long-term contracts [5] - Companies such as航亚科技, 振华股份, and others are highlighted as potential beneficiaries in the modified aircraft engine market [6]

Core Insights - The report from GF Securities highlights that deep integration is expected to address the most challenging issues of "decision trust" and "payment breakpoints" in the deployment of AI Agents [1][2] - The full-stack advantage is anticipated to create significant opportunities for AI Agents, with a focus on the domestic computing power industry chain and infrastructure investments [1][3] Group 1: AI Agent Development - The launch of Alibaba's Qianwen Agent, which integrates with various Alibaba ecosystem applications, is seen as a major advantage [1] - The deep integration of Qianwen with Alibaba's services aims to resolve critical challenges in AI Agent deployment [2] - The "task assistant" feature of Qianwen is being tested, showcasing capabilities in multi-step planning and complex task handling [2] Group 2: Computing Power Investment Opportunities - Alibaba's target of 380 billion yuan in AI capital expenditures over the next three years may be conservative and subject to upward revision [3] - ByteDance reported a significant increase in token consumption, indicating a growing demand for computing power [3] - The sale of shares by GDS Holdings to fund domestic AI data center investments reflects optimism in the infrastructure investment landscape [3] Group 3: Domestic Super Node Acceleration - Alibaba introduced the Panjiu AI Infra 2.0 AL128 super node server, enhancing inference performance by 50% under the same AI computing power [4] - Tencent is developing the ETH-X architecture to optimize GPU and memory communication, with plans for an ultra version [4] - Huawei's upcoming Ascend series is expected to contribute to the super node market, with the 9508192 card anticipated for release in Q4 2026 [4]

Quantitative Models and Construction Methods 1. **Model Name**: Amihud Illiquidity Indicator - **Model Construction Idea**: Measures the price impact of trading volume to assess the liquidity level of an asset. A higher value indicates lower liquidity. [11][12][13] - **Model Construction Process**: The formula is: $$ Amihud = \frac{1}{D} \sum_{d=1}^{D} \frac{\left| R_{i,d} \right|}{Vol_{i,d}} $$ Where: - \( D \): Number of trading days in the window - \( R_{i,d} \): Absolute return of security \( i \) on day \( d \) - \( Vol_{i,d} \): Trading volume of security \( i \) on day \( d \) This indicator reflects the sensitivity of price to trading volume. A higher value indicates that smaller trading volumes cause larger price changes, implying lower liquidity. [12][13] 2. **Model Name**: Pastor-Stambaugh Liquidity Indicator - **Model Construction Idea**: Based on the reversal of asset returns to measure liquidity. Assets with lower liquidity tend to exhibit higher return reversals. [14] - **Model Construction Process**: The formula is: $$ r_{i,d+1}^{e} = \alpha + \beta_{i} r_{i,d} + \gamma_{i} sign(r_{i,d}^{e}) \cdot v_{i,d} + \epsilon_{i,d+1} $$ Where: - \( r_{i,d+1}^{e} \): Excess return of security \( i \) on day \( d+1 \) - \( r_{i,d} \): Return of security \( i \) on day \( d \) - \( v_{i,d} \): Trading volume of security \( i \) on day \( d \) - \( \gamma_{i} \): Liquidity indicator, with a significantly negative value indicating poor liquidity. [14] 3. **Model Name**: Turnover Rate Indicator - **Model Construction Idea**: Reflects the trading activity of an asset by measuring the frequency of its turnover. Higher values indicate higher market liquidity. [15] - **Model Construction Process**: The turnover rate is calculated as: $$ Turnover\ Rate = \frac{Trading\ Volume}{Market\ Capitalization} $$ Where: - \( Trading\ Volume \): Total trading volume of the asset - \( Market\ Capitalization \): Total market value of the asset. [15] 4. **Model Name**: Component Stock Diffusion Indicator - **Model Construction Idea**: Measures the consistency of trends among individual stocks within an industry to assess crowding. Higher values indicate a more crowded market. [16] - **Model Construction Process**: The indicator is calculated as the proportion of stocks in an industry that exhibit a bullish trend, defined as the closing price being above the short-term, medium-term, and long-term moving averages. [16] 5. **Model Name**: Component Stock Pairwise Correlation Indicator - **Model Construction Idea**: Quantifies the homogeneity of stock movements within an industry to evaluate crowding. Higher values indicate stronger synchronization and higher crowding. [17] - **Model Construction Process**: The indicator is the average of pairwise correlation coefficients of returns among all component stocks in an industry over a given window. [17] 6. **Model Name**: Component Stock Return Kurtosis Indicator - **Model Construction Idea**: Captures the extremity of trading by analyzing the tail characteristics of return distributions. Higher kurtosis indicates more extreme returns, suggesting heightened market crowding. [18] - **Model Construction Process**: The indicator is the average kurtosis of daily cross-sectional returns within a window. Kurtosis measures the "peakedness" or "flatness" of a distribution, with higher values indicating fatter tails. [18] 7. **Model Name**: Heat Indicator - **Model Construction Idea**: Uses principal component analysis (PCA) to measure the contribution of a single industry to systemic market risk, reflecting its trading heat. [21][22] - **Model Construction Process**: The formula is: $$ AR_{m} = \frac{\sigma_{m}^{2}}{\sum_{j=1}^{N} \sigma_{j}^{2}} $$ $$ C_{i} = \frac{\sum_{j=1}^{n} AR_{j} \cdot \frac{\left| EV_{i}^{j} \right|}{\sum_{k=1}^{N} \left| EV_{k}^{i} \right|}}{\sum_{j=1}^{n} AR_{j}} $$ Where: - \( N \): Total number of industries - \( n \): Number of principal components - \( \sigma_{m}^{2} \): Variance of the \( m \)-th principal component - \( \sigma_{j}^{2} \): Variance of the \( j \)-th industry return - \( EV_{i}^{j} \): Exposure of the \( j \)-th principal component to the \( i \)-th industry. A higher value indicates that the industry contributes more to systemic market risk, suggesting higher trading heat. [21][22] 8. **Model Name**: Herding Effect Indicator - **Model Construction Idea**: Captures the consistency of market participants' behavior. A significant negative value indicates strong herding behavior, often signaling extreme market sentiment and crowded trading. [23][24] - **Model Construction Process**: The formula is: $$ CSAD_{t} = \gamma_{0} + \gamma_{1} \left| R_{m,t} \right| + \gamma_{2} R_{m,t}^{2} + \mathcal{E}_{t} $$ Where: - \( CSAD_{t} \): Cross-sectional absolute deviation of returns on day \( t \) - \( R_{m,t} \): Market return on day \( t \) - \( \gamma_{2} \): Herding effect indicator. [23][24] 9. **Model Name**: Closing Price-Trading Volume Correlation Indicator - **Model Construction Idea**: Analyzes the stability of the relationship between price and trading volume to predict potential trend reversals. Persistent negative correlation often signals overtrading and potential reversals. [25] - **Model Construction Process**: The indicator is the correlation coefficient between the series of closing prices and trading volumes of an index. [25] 10. **Model Name**: Trading Volume Share Indicator - **Model Construction Idea**: Reflects the concentration of trading in a specific sector or industry. Higher values indicate higher trading concentration and potential overheating. [26] - **Model Construction Process**: The indicator is calculated as the daily trading volume of a sector or industry divided by the total market trading volume. [26] Model Backtesting Results 1. **Amihud Illiquidity Indicator**: No specific backtesting results provided 2. **Pastor-Stambaugh Liquidity Indicator**: No specific backtesting results provided 3. **Turnover Rate Indicator**: No specific backtesting results provided 4. **Component Stock Diffusion Indicator**: No specific backtesting results provided 5. **Component Stock Pairwise Correlation Indicator**: No specific backtesting results provided 6. **Component Stock Return Kurtosis Indicator**: No specific backtesting results provided 7. **Heat Indicator**: No specific backtesting results provided 8. **Herding Effect Indicator**: No specific backtesting results provided 9. **Closing Price-Trading Volume Correlation Indicator**: No specific backtesting results provided 10. **Trading Volume Share Indicator**: No specific backtesting results provided Historical Similarity Analysis Results - Using the Wind Satellite Index (866125.WI) as an example, historical similar segments were identified based on metrics such as component stock count, trading volume share, and market capitalization. - For the next 60 trading days: - **Average maximum return**: 12.79% - **Average time to peak**: 33 days - **Average peak trading volume share**: 4.48% [42][46][49]