持续上涨后，波动加剧 股指周报 2025/10/18 0755-23375128 jiangwb@wkqh.cn 从业资格号：F3048844 交易咨询号：Z0017196 蒋文斌（宏观金融组） 目录 06 估值 周度评估及策略推荐 周度评估及策略推荐 重要消息：1、北京时间10月18日上午，国务院副总理何立峰与美方牵头人、美国财政部长贝森特和贸易代表格里尔举行视频通话，双方同 意尽快举行新一轮中美经贸磋商；2、工信部：到2027年实现城域算力1毫秒时延圈覆盖率不低于70% 打造高品质毫秒入算底座；3、寒武纪 第三季度营收为17.27亿元，同比增长1332.52%；净利润为5.67亿元。紫金矿业：前三季度归母净利润同比增长55%；4、摩根大通宣布未来 十年将为美国稀土、AI、能源、机器人等关键产业提供高达1.5万亿美元融资与项目支持；这项计划的出台恰逢美国政府到处入股关键产业 公司的节点。 经济与企业盈利：1、国家统计局：8月份规模以上工业增加值同比实际增长5.2%，2025年1—7月份全国固定资产投资增长0.5%，7月份社会 消费品零售总额同比增长3.4%，增速延续回落态势；2、中国9月官方制造业PMI为4 ...

Core Insights - Sequoia Capital emphasizes that AI is not merely a software revolution but a labor revolution, targeting the $10 trillion labor market rather than the $650 billion software market [2][8] - The historical context of software development shows that AI is creating new markets similar to how SaaS transformed the software industry [5][7] AI as a Labor Revolution - AI aims to replace certain labor functions rather than just enhance software capabilities, with a focus on sectors like customer service, administration, sales, financial analysis, and education [8] - The current automation level of AI in the U.S. service industry is less than 0.2%, indicating significant potential for growth [8] Comparison with Historical Innovations - The AI revolution is likened to the Industrial Revolution, where the true impact came from the establishment of factory systems rather than the invention of steam engines [10][11] - The development of AI infrastructure, akin to the assembly line in manufacturing, is crucial for widespread adoption and efficiency [12] Future Trends in AI - Sequoia identifies five key trends for AI: enhancing efficiency while accepting uncertainty, the rise of reinforcement learning, the integration of AI into the physical world, the shift in productivity metrics towards computational power, and the need for companies to adapt to these changes [13][14] - The demand for computational power is expected to increase dramatically, creating new opportunities for infrastructure providers [14] Implications for Businesses and Individuals - Companies that can effectively utilize AI will have a competitive edge, while those that do not adapt may face obsolescence [14] - The future workforce will be smaller and more efficient, with a focus on collaboration with AI rather than traditional labor roles [12][14]

Core Insights - Multiple public fund companies, including Beixin Ruifeng and Tongtai, have disclosed their Q3 reports, primarily focusing on equity funds, with a notable emphasis on technology and military sectors [1][2] Group 1: Fund Performance - The top-performing fund, Quan Guo Xu Yuan, reported a scale of 19.069 billion yuan, benefiting from heavy investments in technology and military sectors, leading to significant growth in both performance and scale [1] - Tongtai Digital Economy A achieved a net value increase of over 70% in Q3, focusing on domestic computing power and reducing exposure to overseas supply chains [2] - Beixin Ruifeng Advantage Industry fund saw a net value increase of over 50%, concentrating on strategic emerging industries represented by artificial intelligence [2] - Quan Guo Xu Yuan reported a net value increase of over 45%, driven by heavy investments in technology, new energy, and military sectors [2] Group 2: Market Trends - The technology sector has shown significant growth, with funds focusing on computing power, artificial intelligence, robotics, and semiconductors achieving good returns [2] - The market style has shifted towards growth, with traditional value sectors remaining weak [4] - Fund managers continue to view technology innovation, particularly artificial intelligence, as a core investment theme for the future [4][5] Group 3: Investment Strategies - Some funds, like Tongtai Huize, have focused on niche markets such as the pet economy, despite underperforming compared to broader indices [2] - The Huafu CSI Artificial Intelligence Industry ETF reported a net value increase of over 70%, highlighting the complexity and diversity of the AI sector as a favorable investment avenue [3] - Fund managers express optimism about the long-term potential of the pet economy, despite short-term setbacks [2] Group 4: Bond Market Outlook - The bond market experienced weak fluctuations in Q3, with a notable increase in the yield of 10-year government bonds by 20 basis points [6] - Fund managers anticipate a favorable economic backdrop for the bond market in Q4, supported by moderate monetary policy easing and improving market sentiment [6]

Core Viewpoint - Yitian Intelligent has established a private equity fund, Shanghai Deshu Cloud Fund, to invest in AI computing power services, acquiring 77.4% of Company X, which holds orders worth approximately 10 to 11 billion yuan in computing power services over five years [1] Group 1: Investment and Fund Details - The Shanghai Deshu Cloud Fund has completed its registration and filing with the China Securities Investment Fund Industry Association, with a filing date of October 15, 2025 [1] - The total subscribed capital for the private equity fund is 2 billion yuan, focusing on sectors such as artificial intelligence, big data, computing power, IDC, information industry, and advanced manufacturing [2] Group 2: Business Strategy and Market Position - Yitian Intelligent aims to enhance its operational efficiency and returns by leveraging the investment capabilities and project resources of professional institutions while maintaining its core business [1] - The company positions itself as a "one-stop pioneer service provider" in the computing power sector, offering customized services that integrate computing, storage, and networking [3] - Yitian Intelligent is actively expanding its computing power business, including a strategic cooperation agreement to build a domestic computing power cluster and AI demonstration application projects [3]

Core Viewpoint - Hongbo Co., Ltd. has made significant progress in the construction of an intelligent computing center through its wholly-owned subsidiary, Yingbo Digital Technology, by signing a series of contracts with Beijing Jingneng International Holdings Limited, with a total contract value of approximately 999.68 million yuan [1] Group 1: Contract Details - Yingbo Digital Technology signed a contract for the construction of an intelligent computing center with a total computing power of 1024 PFLOPS, with a total transaction amount of 999,682,000 yuan [1] - The company is set to receive payments from Beijing Jingneng on two occasions, totaling approximately 499.84 million yuan and 299.90 million yuan [1] - A supplementary agreement was signed on December 3, 2024, which revised the equipment procurement agreement, reducing the contract amount to approximately 645.63 million yuan while maintaining the computing power scale [1] Group 2: Debt Settlement - On December 3, 2024, Yingbo Digital Technology signed a debt offset agreement with Jingneng Haibei and a third party, allowing for the offset of receivables amounting to approximately 258.19 million yuan against outstanding debts and quality deposits owed to Jingneng Haibei [2] - Following the completion of the debt offset, Yingbo Digital Technology has remaining receivables of approximately 30.72 million yuan from the third party and a quality deposit of approximately 32.28 million yuan from Jingneng Haibei [2] - The completion of the debt offset agreement is expected to enhance the company's operational efficiency, with specific accounting treatments and impacts on business performance to be detailed in periodic reports [2]

Market Overview - On October 17, A-shares experienced a collective decline, with the Shanghai Composite Index falling by 1.95%, the Shenzhen Component Index by 3.04%, and the ChiNext Index by 3.36%. The North Star 50 Index dropped by 3.75% [1] - The total trading volume in the A-share market reached 1.9547 trillion yuan, with over 4,700 stocks declining [1] - Sectors such as electric grid equipment, photovoltaic energy storage, CPO, controllable nuclear fusion, cultivated diamonds, and semiconductors saw significant declines, while gas, precious metals, and airport shipping sectors performed relatively well [1] ETF Performance - Major broad-based ETFs experienced fluctuations, with the A500 ETF (512050) declining by 2.41% [1] - AI-related ETFs collectively fell, with the Chip ETF (159995) down by 3.68% and the AI ETF (515070) down by 2.82% [1] - The 5G Communication ETF (515050) dropped by 4.23%, and the ChiNext AI ETF (159381) fell by 2.33% [1] AI Sector Insights - Despite increased market volatility, the AI sector continues to show strong growth potential, supported by favorable policies and strategic collaborations [2] - The Ministry of Industry and Information Technology is promoting a "millisecond computing" initiative to enhance computing resource accessibility [2] - AMD's strategic partnership with OpenAI and OpenAI's commitment to invest $1 trillion over the next five years for computing capabilities indicate a robust demand for AI computing resources [2] Long-term Outlook - According to Guotai Junan Securities, the prosperity of AI is extending beyond the stock market into credit and bond markets [2] - Future market focus will shift to the sustainability of AI narratives, including the performance of tech companies in Q3 reports and capital expenditure guidance, as well as the Federal Reserve's interest rate path and U.S. fiscal sustainability [2] - The A-share market's attention will return to incremental policies, external liquidity, and regulatory attitudes, balancing growth and stability in investment strategies [2]

Core Insights - The article discusses the significant advancements in the AI field over the past three years, primarily driven by the development of foundational models, which require substantial data, computational power, and human resources [2][4]. Resource Allocation and Research Impact - The relationship between hardware resources and the publication of top-tier AI/ML conference papers has been analyzed, focusing on GPU availability and TFLOPs [4][5]. - A total of 5,889 foundational model-related papers were identified, revealing that stronger GPU acquisition capabilities correlate with higher acceptance rates and citation counts in eight leading conferences [5][9]. Research Methodology - The study collected structured information from 34,828 accepted papers between 2022 and 2024, identifying 5,889 related to foundational models through keyword searches [8][11]. - A survey of 229 authors from 312 papers indicated a lack of transparency in GPU usage reporting, highlighting the need for standardized resource disclosure [9][11]. Growth of Foundational Model Research - From 2022 to 2024, foundational model research has seen explosive growth, with the proportion of related papers in top AI conferences rising significantly [18][19]. - In NLP conferences, foundational model papers have outpaced those in general machine learning conferences [22]. Research Contributions by Academia and Industry - Academic institutions contributed more papers overall, while top industrial labs excelled in single-institution output, with Google and Microsoft leading in paper production [29][32]. - The research efficiency between academia and industry is comparable, with industry researchers publishing an average of 8.72 papers and academia 7.93 papers [31]. Open Source Models and GPU Usage - Open-source models, particularly the LLaMA series, have become the predominant choice in research, favored for their flexibility and accessibility [35][37]. - NVIDIA A100 is the most widely used GPU in foundational model research, with a notable concentration of GPU resources among a few institutions [38][39]. Funding Sources and Research Focus - Government funding is the primary source for foundational model research, with 85.5% of papers receiving government support [41][42]. - The focus of research has shifted towards algorithm development and inference processes, with a significant portion of papers dedicated to these areas [42]. Computational Resources and Research Output - The total computational power measured in TFLOPs is more strongly correlated with research output and citation impact than the sheer number of GPUs used [44][45]. - While more resources can improve acceptance rates, the quality of research and its novelty remain critical factors in the review process [47].

Group 1 - China COSCO Shipping Group's subsidiary is a leading integrated port and shipping enterprise in Hainan [2] - The company has a new duty-free pickup point, which is part of the duty-free business supply chain [2] Group 2 - Spot gold has risen to $4,330 per ounce, reaching a historical high [3] - Baoding Technology's wholly-owned subsidiary Hexi Gold Mine is engaged in gold mining and sales, with a metal reserve of 4.536 tons [3] - The company has a production capacity of 500 tons of silver annually [3] Group 3 - Tesla will hold a shareholder meeting to showcase new developments in robotics [5] - Pioneering Electronics, through its subsidiary, holds an indirect stake in Benmo Technology, which is rapidly developing in the robotics sector [5] - The company anticipates significant revenue growth in 2024 compared to 2023 [5] Group 4 - Guangdong Mingzhu expects a net profit increase of 858.45%-1071.44% in the first three quarters due to improved operational performance [6] - The actual controller of Cangzhou Mingzhu is set to change to the Guangzhou State-owned Assets Supervision and Administration Commission [6] Group 5 - The company is the first listed firm to produce electronic-grade phosphoric acid [7] - The company is a core supplier for military and armed police procurement, specializing in nuclear pollution prevention technology [7]

Core Viewpoint - Haiguang Information's net profit attributable to shareholders for the first three quarters of 2025 reached 1.961 billion yuan, representing a year-on-year increase of 28.56% [1] - The company's revenue for Q3 2025 was 4.026 billion yuan, showing a year-on-year growth of 69.60%, with a net profit of 760 million yuan, up 13.04% year-on-year [1] Financial Performance - The net profit attributable to shareholders for the first three quarters of 2025 was 1.961 billion yuan, a 28.56% increase compared to the previous year [1] - Q3 2025 revenue was 4.026 billion yuan, reflecting a significant year-on-year growth of 69.60% [1] - Q3 2025 net profit was 760 million yuan, which is a 13.04% increase year-on-year [1] Strategic Initiatives - The company is intensifying its ecological construction efforts, indicating a focus on long-term growth [1] - A stock incentive plan has been announced, demonstrating the company's confidence in its long-term prospects [1] Competitive Position - The company is positioned to become a rare player in the domestic market for chips, interconnects, servers, and liquid cooling, establishing itself as a comprehensive computing power leader [1] - Haiguang Information holds a unique advantage in the domestic GPU competitive landscape [1] - The company maintains a strong leadership position in computing power and is continuously enhancing its R&D efforts to accelerate technological iterations [1] Future Outlook - The stock incentive plan is expected to anchor future growth potential [1] - The company is rated as a "Buy" based on its solid market position and growth prospects [1]

Group 1 - The core viewpoint of the article emphasizes the transformative potential of AI and the strategic investment by Nvidia in OpenAI, betting on AI becoming a new utility akin to electricity [3][5][13] - Nvidia plans to invest $100 billion in OpenAI to build a 10GW computing power cluster, which could generate $400 billion in revenue for Nvidia if powered by its chips [5][13] - The article argues that AI chip market has not reached its peak, with a significant shift from CPU to GPU for applications like search and recommendation systems, indicating a major industry upgrade [14][16] Group 2 - Nvidia's CEO highlights that the global GDP is approximately $100 trillion, and investing $10 trillion in AI could significantly enhance productivity, with a substantial portion directed towards AI infrastructure [16][19] - The article discusses the misconception among Wall Street analysts regarding Nvidia's growth potential, suggesting that they fail to recognize the exponential growth pattern of AI technology [20][21] - The competition from ASIC chips is acknowledged, but Nvidia's focus on efficiency over price is emphasized, as energy consumption is the real cost driver in computing power [24][25] Group 3 - The article addresses concerns about AI leading to job losses, arguing instead that AI will create new job opportunities and enhance human capabilities rather than replace them [28][29] - It stresses the importance of understanding the transformative power of AI and the need for investors to remain open-minded and informed about the evolving landscape [29][30] - The article concludes that the current era represents a pivotal moment for investment in AI and hard technology, with the potential for significant returns [30]