Group 1 - The company Jin Xin Nong (SZ 002548) announced an investor research meeting scheduled for November 11, 2025, where the board secretary and securities representative will address investor inquiries [1] - For the first half of 2025, Jin Xin Nong's revenue composition was as follows: feed processing accounted for 66.97%, livestock breeding for 32.1%, other industries for 0.58%, and raw material trading for 0.35% [1] - As of the report date, Jin Xin Nong's market capitalization was 4.6 billion yuan [1]

Investment Rating - The report maintains a "Positive" investment rating for the industry [8]. Core Insights - On November 6, the company "月之暗面" officially released and open-sourced the Kimi K2 Thinking model, which is characterized by its general Agentic capabilities and reasoning abilities, excelling in deep reasoning and capable of solving complex problems through multiple tool calls [2][5]. - Kimi K2 Thinking is based on the "model as Agent" concept, allowing for up to 300 rounds of tool calls autonomously without human intervention, potentially opening new paradigms for cost reduction in AI applications [2]. - The model has shown significant improvements in reasoning performance, surpassing GPT-5 in various benchmarks, including achieving a 44.9% accuracy in HLE tests compared to GPT-5's 41.7% [10]. - The model's ability to autonomously search and browse has also improved, scoring 60.2% in BrowseComp tests, which is significantly higher than the human average of 29.2% and GPT-5's 54.9% [10]. - Kimi K2 Thinking can perform 200-300 tool calls in an uninterrupted manner, enhancing its Agentic and reasoning performance, which is crucial for complex problem-solving [10]. - The model's pricing is set at 4 RMB/M tokens for input and 16 RMB/M tokens for output, with a Turbo API available at higher rates, indicating a competitive pricing strategy in the market [10]. Summary by Sections Event Description - The Kimi K2 Thinking model was released on November 6, featuring a MoE architecture with 1 trillion parameters and 32 billion active parameters, supporting a context length of 256K, optimized for domestic GPUs [5]. Performance Analysis - The model has demonstrated enhanced reasoning stability due to improved long-range planning and autonomous search capabilities, allowing it to break down complex tasks into manageable steps and integrate results effectively [10]. - The introduction of native INT4 quantization has improved reasoning efficiency and hardware compatibility, potentially reducing costs for large-scale inference servers [10]. Market Outlook - The report expresses optimism about the domestic AI industry chain, recommending stocks with significant positioning advantages and shovel stocks that support the AI ecosystem [2].

Investment Rating - The report does not explicitly provide an investment rating for the industry Core Insights - The report addresses the increasing concerns regarding electricity shortages in the U.S., focusing on the extent of shortages, the integration of data centers into the grid, and the economic viability of self-built power sources for data centers [3][6] Summary by Sections Objective Indicators of Electricity Shortages - In 2024, the average power outage duration for U.S. electricity users increased significantly to 662.6 minutes, approximately 11 hours, marking an 81% year-on-year increase and the highest level in nearly a decade, indicating a substantial worsening of electricity shortages [7][18] - The average retail electricity price in the U.S. reached 13.54 cents per kWh from January to August 2025, a 5.2% increase year-on-year, with the PJM region seeing an average price of 13.14 cents per kWh, up 9% [23][24] Impact of Power Plant Retirements and Load Transfers - From 2025 to 2030, approximately 54 GW of thermal power plants are expected to retire, and around 10.3 GW of cryptocurrency mining facilities may exit the market due to rising electricity prices. If these plants do not retire and the mining facilities exit, the total potential relief for electricity shortages could be about 53.49 GW, which is only 36.2% of the previously predicted 201 GW shortage [8][27] Policy Developments for Data Center Integration - The U.S. Department of Energy has initiated a rule-making process to ensure large electricity loads, such as AI data centers, can connect to interstate grid systems in a timely and non-discriminatory manner. This trend is supported by various state policies encouraging data centers to match local resources [9][31] Project Developments in Data Center Energy Storage - There is a growing number of projects where AI data centers are integrating energy storage solutions. For instance, the GW Ranch off-grid power park in Texas plans to pair 5 GW of gas power with 1.8 GW of storage, while the Masdar&EWEC project in Abu Dhabi will provide 5.2 GW of solar and 19 GWh of storage for a 1 GW data center [10][38] Economic Comparison of Power Sources - The estimated costs per kWh for various power generation technologies are as follows: natural gas combined cycle at $0.0468, distributed internal combustion engines at $0.0608, solid oxide fuel cells (SOFC) at $0.0717, and solar plus storage at $0.0815. Natural gas generation is deemed the most cost-effective option [11][42] Capital Expenditure Implications for Internet Giants - The initial investment increase for data centers due to self-built power sources is relatively modest, with combined cycle units increasing costs by about 6%, distributed engines by 9%, SOFC by 7%, and solar plus storage by 16%. The report suggests that electricity shortages will not hinder AI development but rather drive infrastructure investment [12][46]

江证券股份有限公 (住所:湖北省武汉市江汉区淮海路 88 号) 2025 年面向专业投资者公开发行公司债券 (第四期) 发行公告 联席主承销商、簿记管理人 (III) 中信建投证券 (北京市朝阳区安立路66号4号楼) 联席主承销商、债券受托管理人 联席主承销商 牵头主承销商 信证券 (广东省深圳市福田区中心 三路8号卓越时代广场(二 期)北座) 长江证券承销保存有限公司 (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) (中国(上海)自由贸易 试验区世纪大道1198号28 层) 签署日期:二〇二五年十一月六日 发行人及董事会全体成员保证公告内容的真实、准确和完整，对公告的虚 假记载、误导性陈述或者重大遗漏负连带责任。 重要提示 1、长江证券股份有限公司（以下简称"发行人"或"公司"）面向专业机构投资 者公开发行（不超过）人民币 200 亿元（含 200 亿元）的公司债券（以下简称"本 次债券"）已于 2025 年 9 月 24 日获得中国证券监督管理委员会（以下简称"证监 会"）注册批复《关于同意长江证券股份有限公司向专业投资者公开发行公司债 券注册的批复》（证监许可〔2025 ...

长江证券股份有限公司 2025 年面向专业投资者公开发行公司债券更名公告 2025 年 9 月 24 日，根据中国证监会"证监许可〔2025〕2145 号"文件同意 了长江证券股份有限公司 2025 年面向专业投资者公开发行公司债券（以下简称 "本次债券"）的发行申请。 本次债券申报时命名为"长江证券股份有限公司 2025 年面向专业投资者公 开发行公司债券"，鉴于本次债券采用分期发行的方式，按照公司债券命名惯例， 征得主管部门同意，本次发行人拟发行的规模不超过人民币 30 亿元的公司债券 名称确定为"长江证券股份有限公司 2025 年面向专业投资者公开发行公司债券 （第四期）"（以下简称"本期债券"）。 本期债券名称变更不改变原签订的与本次公司债券发行相关的法律文件效 力，原签署的相关法律文件对更名后的公司债券继续具有法律效力。前述法律文 件包括但不限于：《关于长江证券股份有限公司 2025 年面向专业投资者公开发行 公司债券之承销协议》及其补充协议、《国浩律师（上海）事务所关于长江证券 股份有限公司 2025 年面向专业投资者公开发行公司债券之法律意见书》等文件。 特此说明。 （以下无正文） 1 2025 ...

| 发行人: | 长江证券股份有限公司 | | --- | --- | | 牵头主承销商: | 中信证券股份有限公司 | | 联席主承销商: | 中信建投证券股份有限公司、华泰联合证券有限责 任公司、长江证券承销保荐有限公司 | | 受托管理人: | 华泰联合证券有限责任公司 | | 注册金额: | 200 亿元 | | 发行金额 | 不超过 30 亿元(含) | | 增信措施情况: | 元 | | 信用评级结果: | 主体评级 AAA | | 信用评级机构: | 联合资信评估股份有限公司 | 长江证券股份有限公司2025年面向专业投资者公开发行公司债券（第四期）募集说明书 声 明 发行人将及时、公平地履行信息披露义务。发行人及其全体董事、监事、高 级管理人员或履行同等职责的人员保证募集说明书信息披露的真实、准确、完整， 不存在虚假记载、误导性陈述或重大遗漏。主承销商已对募集说明书进行了核查， 确认不存在虚假记载、误导性陈述和重大遗漏，并对其真实性、准确性和完整性 承担相应的法律责任。 发行人承诺在本期债券发行环节，不直接或者间接认购自己发行的债券。债 券发行的利率或者价格应当以询价、协议定价等方式确定，发行 ...

www.lhratings.com 1 长江证券股份有限公司 2025 年跟踪评级报告 联合〔2025〕4232 号 联合资信评估股份有限公司通过对长江证券股份有限公司及其 相关债券的信用状况进行跟踪分析和评估，确定维持长江证券股份 有限公司主体长期信用等级为 AAA，维持"21 长江 03""22 长江 02""22 长江 03""22 长江 04""23 长江 02""23 长江 03""23 长江 04""23 长江 05""23 长江 06""24 长江 01""24 长江 02""24 长江 03""25 长江 01"信用等级为 AAA，上调"23 长江 C1""23 长江 Y1""24 长江 Y1""25 长江 Y1"信用等级为 AAA，评级展望为稳 定。 特此公告 联合资信评估股份有限公司 评级总监： 二〇二五年六月二十七日 跟踪评级报告 | 2 声 明 一、本报告是联合资信基于评级方法和评级程序得出的截至发表之日 的独立意见陈述，未受任何机构或个人影响。评级结论及相关分析为联合 资信基于相关信息和资料对评级对象所发表的前瞻性观点，而非对评级对 象的事实陈述或鉴证意见。联合资信有充分理由保证所出 ...

Group 1 - The CSI All Share Securities Company Index fell by 0.35% as of November 11, with notable gains from Northeast Securities, Huazhong Securities, and others [1] - The Broker ETF (159842) experienced a decline of 0.25% with a trading volume exceeding 34 million yuan and a premium rate of 0.03% [1] - The Broker ETF has seen net inflows for two consecutive days, accumulating over 330 million yuan in the last 10 trading days, with 8 days of net inflows [1] Group 2 - China Aviation Securities is optimistic about the growth potential of the brokerage industry, highlighting that regulatory encouragement for industry consolidation is a positive trend [2] - Mergers and acquisitions are viewed as effective means for brokerages to achieve external growth, enhancing overall industry competitiveness and resource allocation [2] - Industry consolidation is expected to improve concentration and create scale effects within the brokerage sector [2]

Core Viewpoint - The A-share market showed signs of recovery on November 10, with the Shanghai Composite Index closing at 4018.6 points, approaching the intraday high of 4025.7 points recorded on October 30 [1] Group 1: Market Performance - The brokerage sector experienced slight movements, with Northeast Securities hitting the daily limit, while other firms like GF Securities, Dongxing Securities, and Changjiang Securities also saw gains [1] - The broker ETF (159842) rose by 1.02%, with a trading volume of 392 million yuan, indicating a significant increase compared to the previous Friday [1] - According to Wind data, the ETF's shares increased by 6 million, bringing its total size to 8.165 billion yuan, marking a historical high [1] Group 2: Future Outlook - Institutions predict that the capital market will maintain a steady upward trend for the year, with sustainable earnings growth expected to enhance the sector's valuation recovery potential [1] - In the context of a continuously improving capital market, brokerage firms' proprietary business income is anticipated to continue its growth trajectory, providing stable support for overall performance [1]

Market Overview - The securities sector increased by 1.31% on November 10, with Northeast Securities leading the gains [1] - The Shanghai Composite Index closed at 4018.6, up 0.53%, while the Shenzhen Component Index closed at 13427.61, up 0.18% [1] Individual Stock Performance - Northeast Securities (000686) closed at 10.20, with a rise of 10.03% and a trading volume of 2.0144 million shares, totaling a transaction value of 2.001 billion [1] - Other notable performers include: - GF Securities (000776) at 23.60, up 3.74% with a transaction value of 2.286 billion [1] - Dongxing Securities (601198) at 12.97, up 3.10% with a transaction value of 996 million [1] - Changjiang Securities (000783) at 9.10, up 2.82% with a transaction value of 1.197 billion [1] Fund Flow Analysis - The securities sector saw a net inflow of 1.097 billion from institutional investors, while retail investors experienced a net outflow of 1.037 billion [1] - Key net inflows from major stocks include: - Northeast Securities with a net inflow of 353 million, accounting for 17.65% of total inflow [2] - Dongfang Caifu (300059) with a net inflow of 199 million, representing 3.67% [2] - Industrial Securities (601377) with a net inflow of 183 million, making up 12.23% [2] Summary of Stock Flows - The table indicates significant net outflows from retail investors in several stocks, including: - Northeast Securities with a retail net outflow of 240 million [2] - Dongfang Caifu with a retail net outflow of 130 million [2] - Industrial Securities with a retail net outflow of 1.20 billion [2]