| 2025年3月4日 | | --- | | 各股市走势图 . | 股市短评 | 1 | | --- | --- | --- | | 主要经济数据公布 . | 主要新闻总结 . | | | 国际经济要闻 . | 大中华经济要闻 | | | 企业要闻 . | 中港两地上市的Ｈ股股份 | | | 免责声明 . | | | | 指數 | 收市 | 日比(%) | 5日(%) | 17,000 | | --- | --- | --- | --- | --- | | 恒生 | #N/A | 0.28 | (1.44) | 16,000 15,000 | | | Requesting | | | 14,000 | | H股 | #N/A | 0.03 | (2.31) | | | | Requesting | | | | | 紅籌 | #N/A | #N/A | #N/A | | | | Requesting | Requesting | Requesting | | | HSC大型股 | #N/A | #N/A | #N/A | | | | Requesting | Requesting | Requesting | ...

环球市场动态 美 国 通 胀 下 半 年 存 反 弹 隐 忧 股 票 中国股市周二回调，A 股窄幅震 荡，港股情绪偏弱，市场观望特朗 普关税下一步走向；欧洲股市周三 收高，从近期跌势中反弹，市场对 乌克兰可能停火及美国 CPI 数据作 出积极反应；美股科技股推动反 弹，通胀数据缓解经济忧虑。 外 汇 / 商 品 俄乌冲突持续，加上美国公布最新 汽油日均需求为 11 月份以来最 高，周三国际油价上 2%；美国 CPI 数据温和，投资者正在评估美 国经济前景和美联储的利率路径， 伦铜走高，金价上升。 固 定 收 益 美国短期国债领跌，尽管 CPI 数据 温和，但市场对美联储降息的押注 继续减弱。美国 10 年期国债发行中 标收益率低于发行前交易水平。亚 洲投资级债券相对平稳，流动较轻。 产品及投资方案部 注：bp/bps=基点；pt/pts=百分点 中信证券财富管理 (香港) 免责声明请参考封底 2025 年 3 月 13 日 .s 入 a. 竹 ▪ 美国 2 月通胀增速全面低于预期和前值，总体 CPI 环比升 0.2% (前值 0.5%)，核心 CPI 环比升 0.2% (前值 0.4%)，总体 CPI 同比升 ...

Investment Rating - The report maintains a "Buy" rating for the companies discussed, including TuoSiDa and BaoFeng Energy, based on their growth potential and financial performance [8][9][10]. Core Insights - The semiconductor industry is witnessing a significant shift towards self-developed AI chips by major companies, driven by the increasing demand for AI applications and the need for efficient computing solutions [4][6]. - The healthcare sector is advancing with the introduction of brain-computer interface technologies, supported by new pricing guidelines from the National Healthcare Security Administration, which will facilitate clinical applications [7]. - The macroeconomic environment shows mixed signals, with U.S. employment data indicating a slight cooling but not severe enough to trigger recession fears, while fiscal policies under the Trump administration are impacting market sentiment [1][14]. Industry Summaries Semiconductor Industry - The competition between GPGPU and ASIC chips highlights the strengths and weaknesses of each technology, with ASICs excelling in low-precision tasks but lagging in memory bandwidth compared to GPGPUs [4]. - Major companies are investing heavily in R&D for AI chips, with the expectation that the demand for AI inference will continue to grow significantly [6]. Healthcare Sector - The successful implementation of brain-computer interface surgeries marks a breakthrough in medical technology, with new pricing projects established to support these innovations [7]. - The National Healthcare Security Administration's new guidelines will help standardize costs associated with brain-machine interface services, paving the way for broader clinical adoption [7]. Macroeconomic Environment - Recent U.S. economic data presents a mixed picture, with non-farm employment figures slightly below expectations, yet still within acceptable limits, alleviating some recession concerns [1][14]. - The divergence in fiscal narratives between the U.S. and Europe, particularly the shift towards tighter fiscal policies in the U.S., is creating volatility in market sentiments, impacting asset prices [1][14].

Investment Rating - The report maintains an "Overweight" investment rating for the electronic industry [1] Core Viewpoints - The competition between GPGPU and ASIC chips is highlighted, with ASICs focusing on low-precision tasks and showing better power efficiency, but still lagging behind GPGPU in certain performance metrics [5][8] - Major companies are increasingly investing in self-developed AI chips to meet the growing demand for AI applications, with significant capital expenditures expected to cover initial development costs [5][16] - The report recommends investing in companies like Cambricon and Haiguang Information, while also suggesting to pay attention to ZTE, Aojie Technology, and Chipone [5] Summary by Sections 1. GPGPU vs ASIC Performance Comparison - ASICs primarily target low-precision data types, which are sufficient for large model training, while GPGPU excels in high-precision tasks [8] - In terms of power efficiency, ASICs generally have better power control and efficiency ratios compared to GPGPU [8][11] - GPGPU's memory bandwidth and capacity still surpass those of ASICs, although ASICs have higher computational density [11][12] 2. Reasons for Major Companies Developing AI Chips - The cost structure for chip companies includes employee salaries, EDA and IP costs, manufacturing expenses, and sales costs, with salaries making up a significant portion [16][17] - The report estimates that a digital chip Fabless company requires approximately 9.7 billion yuan for salaries alone for a development team [17][18] - The demand for AI inference is expected to grow significantly, with major companies building large-scale clusters to support this demand [18][19] 3. Who Can Manufacture AI Chips for Major Companies? - Broadcom is identified as a leader in AI interconnect technology, with a strong IP ecosystem and significant market share in AI custom chip services [21][24] - Marvell is noted for its rapid growth in the AI chip market, with a significant increase in AI-related revenue and partnerships with major cloud service providers [25][27] - AIchip is recognized for its advanced 3DIC and process technology, addressing efficiency and performance challenges in AI and high-performance computing [28][29]

Investment Rating - The report maintains an "Overweight" investment rating for the electronic industry [1]. Core Viewpoints - The competition between GPGPU and ASIC chips is highlighted, with ASICs focusing on low-precision tasks and showing better power efficiency, but still lagging behind GPGPU in certain performance metrics [5][8]. - Major companies are increasingly investing in self-developed AI chips to meet the growing demand for AI applications, with significant capital expenditures expected to cover initial development costs [5][16]. - The report recommends investing in companies like Cambricon and Haiguang Information, while also suggesting to pay attention to ZTE, Aojie Technology, and Chipone [5]. Summary by Sections 1. GPGPU vs. ASIC Performance Comparison - ASICs primarily target low-precision data types, which are sufficient for large model training, while GPGPU excels in high-precision tasks [8]. - ASICs have better power control and efficiency for specific tasks, but GPGPU still outperforms in certain metrics, such as NVIDIA's GB200 [8][11]. - The report notes that ASICs have high computational density but face challenges in memory bandwidth compared to GPGPU [5][11]. 2. Reasons for Major Companies to Develop AI Chips - The report outlines the cost structure of chip companies, emphasizing that employee salaries constitute a significant portion of expenses [16][17]. - It estimates that a digital chip Fabless company requires approximately 9.7 billion yuan for employee salaries over a two-year product development cycle [18]. - The demand for AI inference is expected to grow significantly, with NVIDIA reporting that 40% of its data center revenue comes from inference business [18] . 3. Who Can Manufacture AI Chips for Major Companies? - Broadcom is identified as a leader in AI interconnect technology, with a strong IP ecosystem and significant market share in AI custom chip services [21][24]. - Marvell is noted for its rapid growth in the AI chip market, with a significant increase in AI-related revenue and partnerships with major cloud service providers [25][27]. - AIchip is recognized for its advanced 3DIC and process technology, which addresses efficiency and performance challenges in AI and high-performance computing [28][29].

Investment Rating - The report does not explicitly provide an investment rating for the industry Core Insights - On March 12, 2025, the market experienced fluctuations, with the computing power concept gaining strength. The Shanghai Composite Index fell by 0.23%, the Shenzhen Component Index decreased by 0.17%, and the ChiNext Index dropped by 0.58%. The total market turnover was 1,683.51 billion yuan, an increase of 201.88 billion yuan compared to the previous trading day [2][14]. Market Performance - The market saw 2,666 stocks rise and 2,587 stocks fall. The performance of various indices ranked as follows: Growth > Stability > Cyclical > 0 > Financial > Consumer; Small-cap Value > Small-cap Growth > Mid-cap Value > Large-cap Growth > Large-cap Value > Mid-cap Growth. The CSI All Share Index outperformed the fund-heavy positions [2][19]. Industry Performance - Among the 30 CITIC first-level industries, the top performers were Media (1.84%), Comprehensive Finance (1.33%), and Communication (1.11%). The laggards included Home Appliances (-0.84%), Coal (-0.80%), and Food & Beverage (-0.78%). Most concept sectors declined, with significant increases in Broadcasting, Digital Watermarking, and Blind Box Economy, while PLC, Aquaculture, and Engineering Machinery saw declines [2][19][23]. Fund Flow Analysis - On March 12, 2025, the main funds experienced a net outflow of 11.935 billion yuan. Specifically, large orders saw a net outflow of 12.615 billion yuan, while small orders continued to see a net inflow of 18.238 billion yuan [3][24]. ETF Fund Flow - On March 12, 2025, the trading volume of major ETFs such as the Huaxia SSE 50 ETF and the Huatai-PB CSI 300 ETF decreased compared to the previous trading day. The trading volumes were 1.493 billion yuan, 2.457 billion yuan, and others, with changes of -184 million yuan and -1.571 billion yuan respectively [3][30]. Global Market Performance - On March 12, 2025, major Asia-Pacific indices closed mixed, with the Hang Seng Index down 0.77% and the Nikkei 225 up 0.07%. European indices also saw declines, with the DAX down 1.29% and the FTSE 100 down 1.21%. The US markets also experienced declines, with the Dow Jones Industrial Average down 1.14% [4][5][34].

Core Viewpoint - TSMC is considering acquiring a stake in a joint venture that would operate Intel's foundry business, with the aim of reviving Intel's struggling operations and supporting U.S. advanced manufacturing efforts [1][5][6] Group 1: TSMC's Proposal and Negotiations - TSMC has approached companies like Nvidia, AMD, and Broadcom regarding a potential investment in a joint venture for Intel's foundry operations, with TSMC not holding more than 50% of the stake [1][2] - The negotiations are in early stages and require approval from the Trump administration, which is concerned about foreign control over Intel [2][5] - TSMC's proposal is part of a broader strategy to enhance its presence in the U.S. semiconductor market, following a commitment to invest $100 billion in new chip factories [5] Group 2: Intel's Current Situation - Intel's stock has dropped over 50% in the past year, and the company reported a net loss of $18.8 billion in 2024, marking its first loss since 1986, primarily due to significant asset impairments [4] - The company’s foundry division has a book value of $108 billion as of December 31 [4] - Intel's board supports negotiations with TSMC, although there is internal opposition from some executives [6] Group 3: Technical and Operational Challenges - Any collaboration between TSMC and Intel will face significant challenges due to differences in manufacturing processes, chemicals, and equipment used by both companies [6][7] - TSMC aims for potential investors to not only be shareholders in the joint venture but also customers of Intel's advanced manufacturing business [7] - The 18A process technology has become a contentious point in negotiations, with Intel claiming it is more advanced than TSMC's 2nm process [7]

宏 观 研 究 2025.03.10 每 周 经 济 动 态 近期美股市场下跌原因探析 方 正 证 券 研 究 所 证 券 研 究 报 告 《10 月美国 CPI 同比料在低基数下反弹》 2024.11.11 《基建实物工作量同比降幅收窄，地产销售延 续改善——量化经济指数周报-20241103》 2024.11.05 《美国大选投票日观察指南》2024.11.03 金晗 登记编号：S1220525030005 燕翔 登记编号：S1220525030001 相 关 研 究 《特朗普交易或有止盈风险，美联储料暂停降 息进程》2024.11.17 《建材价格延续回升指向建筑需求仍在改善— —量化经济指数周报-20241117》2024.11.17 2025 年 2 月中旬以来美股三大指数出现快速调整，截至 3 月 9 日收盘，标 普 500 指数相较于 2 月中旬的高点已下跌 6.1%，纳斯达克指数下跌 9.3%， 道琼斯指数下跌 4.1%。分行业看，美股可选消费、通讯服务、信息技术行 业领跌。我们认为，近期美股市场下跌的主要原因有以下三点：美国 AI 垄 断格局被打破、美国经济增长动力减弱、特朗普关税政策引发担 ...

Investment Rating - The report maintains a "Buy" rating for Marvell Technology (MRVL US) with a target price of $80.90 [6][7]. Core Insights - Marvell's Q4 FY25 revenue reached $1.817 billion, with a Non-GAAP net profit of $530 million, aligning with Bloomberg consensus estimates [2]. - The company expects Q1 FY26 revenue to be approximately $1.875 billion, with a Non-GAAP EPS of $0.61, also in line with market expectations [2]. - Data center revenue is projected to grow in the mid-high single digits quarter-over-quarter, with AI-related revenue expected to increase by over 10% [2][3]. - ASIC projects contribute about 25% of data center revenue, with ongoing collaborations with major cloud providers [3]. - The demand for optical products remains strong, particularly for 800G PAM and 400ZR DCI products, which are expected to dominate shipments in 2025 [4]. Summary by Sections Financial Performance - Q4 FY25 revenue was $1.817 billion, with a year-over-year increase of 24% and a quarter-over-quarter increase of 78% in data center revenue [2]. - The company projects FY26 revenue of $8.384 billion, reflecting a 45.38% year-over-year growth [5][11]. Business Segments - The ASIC business is expected to grow significantly, with contributions from collaborations with major cloud providers [3]. - The optical product segment is anticipated to maintain stable growth, driven by data center network upgrades [4]. Valuation Metrics - The adjusted net profit for FY26 is forecasted at $2.532 billion, with an adjusted EPS of $2.93 [5][11]. - The report assigns a 27.6x adjusted PE for FY26, slightly above the comparable company average of 27.1x [5].

Core Viewpoint - TSMC is in discussions with major semiconductor companies like NVIDIA, AMD, and Broadcom to explore the possibility of jointly investing in a venture to operate Intel's foundry business, which has led to a significant increase in Intel's stock price by over 10% in after-hours trading [2]. Group 1 - TSMC will manage Intel's foundry operations but will not hold more than 50% ownership to comply with U.S. government regulations on foreign ownership [2]. - The negotiations focus on Intel's 18A technology, which Intel claims is superior to TSMC's 2nm process, with NVIDIA and Broadcom already testing the 18A process and AMD evaluating its applicability [2]. - The discussions involve complex technical and strategic challenges, including intellectual property protection and process differences, with TSMC requiring potential investors to also be customers of Intel's advanced processes to create synergies [2]. Group 2 - Qualcomm has withdrawn from earlier discussions to acquire a stake in Intel, while Intel's board members support the deal, although some executives are firmly opposed [2]. - Intel has rejected discussions about separating its chip design department from its foundry department for sale [2].