Data is the fuel that keeps Wall Street's engine turning -- and one of the most important data releases of the entire quarter occurred two weeks ago. By no later than Feb. 17, institutional investors with at least $100 million in assets under management were required to file Form 13F with the Securities and Exchange Commission.A 13F offers a way for investors to track which stocks Wall Street's brightest investors, including billionaire Dan Loeb of Third Point, bought and sold in the most recent quarter. Th ...

1. Cisco Systems (CSCO) In a market where valuations remain elevated, free cash flow (FCF) is one of the most reliable indicators of financial strength. Companies that consistently generate strong cash flow can: Why it looks undervalued: Investment thesis: Cisco's (NASDAQ:CSCO) shift toward software and recurring revenue improves margins and predictability. Despite slower headline growth, its steady free cash flow supports dividends and aggressive buybacks. Cisco Daily Chart 2. Pfizer (PFE) Below are 7 unde ...

证券研究报告 1、全球AI产业仍在快速发展。美国科技巨头在2026年仍在进一步加码AI相关基建投资，希望获得未来竞争优势。其共识可能是：过度投资的风险小 于投资不足的风险。而2026年以Anthropic为代表的大模型公司发布了一系列AI Agent产品，显示出AI相关应用的落地速度较快。AI的发展在短期内可 能会产生一定的通缩效应，但长期影响则值得期待。对中国而言，AI发展的空间与潜力也非常可观。 2、AI产业链持续拉动中国出口增长。2025年3-4月美国关税宣布后，AI产品与其他产品出口趋势发生了明显分野：AI产品出口增速继续向上，其他产 品出口增速则开始回落，2025年8月关税落地后其他产品增速中枢进一步回落。美国科技巨头四季报电话会给出的全年资本开支指引大幅上调，若实 际开支按照此指引执行，2026年AI产品出口同比或将再创新高，为2026年出口韧性带来支撑，预计2026年出口或在4.8%-5.6%之间。 提质增效，科技突围 ——2026年春季宏观展望 姓名 何宁（分析师） 证书编号：S0790522110002 邮箱：hening@kysec.cn 姓名 陈策（分析师） 证书编号：S07905240 ...

India is fast becoming one of the world’s biggest AI user bases. The question now is how it can turn that scale into superpower status rather than just training Silicon Valley for free.That will be a tall order for a country largely caught flat-footed by the boom. But let’s start with the basics: The three main building blocks of AI are talent, compute (including high-end chips and infrastructure), and data. India doesn’t lack engineers, but it currently doesn’t have foundational research training at scale ...

Summary of Conference Call Notes Companies Involved - Alphabet (GOOGL) - Meta Platforms (META) - Microsoft (MSFT) - Oracle (ORCL) Key Points and Arguments Depreciation Estimates - Updated depreciation model indicates that cumulative depreciation for MSFT, ORCL, META, and GOOGL could exceed **$520 billion** over the next three years, an increase of **$30 billion** from previous estimates, primarily driven by GOOGL [2][5] - Depreciation expense is expected to become a significant portion of these companies' cost structures due to rising capital expenditure (capex) investment rates, which are projected to surpass the dot-com peak this year [2][6] GOOGL's Capex Guidance - GOOGL has notably increased its 2026 capex guidance to between **$175 billion and $185 billion**, significantly above consensus expectations of approximately **$120 billion** [7] - This increase is expected to result in an additional **$20 billion** of depreciation over the next three years, with the most substantial impact on depreciation estimates occurring in **2028**, where estimates rose by approximately **25%**, equating to **13% of revenue** [7][11] Impact on Cost Structure - The shift towards higher capital investment indicates a transition to a larger fixed cost structure, moving away from legacy asset-light business models [14] - Analysts anticipate that expense forecasts will need to be revised higher to reflect the increased capex, which could lead to lower margin expectations if revenue growth does not keep pace [14] Depreciation Growth Rate - Depreciation among hyperscalers is projected to grow at a **CAGR of approximately 60%** over the next three years [3] - The rising depreciation as a percentage of revenue for MSFT, ORCL, META, and GOOGL is illustrated, indicating a trend that could affect overall profitability [15] Financial Implications - Non-depreciation expenses will need to decline rapidly relative to revenue to meet margin expectations, highlighting the pressure on operational efficiency [17][18] Additional Important Information - The estimates for depreciation in FY1 to FY3 are derived from the depreciation model based on Microsoft’s capex estimates, with specific coverage by analysts for each company [5][16] - The lag between capital investment and asset deployment contributes to higher depreciation in future years, emphasizing the importance of timing in financial projections [6] This summary encapsulates the critical insights from the conference call, focusing on the financial implications of increased depreciation and capex for the involved companies.

[Table_Title] 研究报告 Research Report 1 Mar 2026 中国电子 China (Overseas) Technology 谷歌 TPU 切入 Meta 核心训练池 Google TPUs Enter Meta's Core Training Stack 姚书桥 Barney Yao 吴叡霖 Louis Ng barney.sq.yao@htisec.com louis.yl.ng@htisec.com [Table_yemei1] 热点速评 Flash Analysis [Table_summary] (Please see APPENDIX 1 for English summary) 事件 Meta 已与 Google 签署多年期协议，将以租用方式使用 Google 自研 AI 芯片 TPU，用于训练与开发新一代 AI 模型，合 同规模达数十亿美元。报道同时提到，Meta 也在评估更进一步合作，包含最早明年在其自建数据中心采购并部署 TPU 的可能性。 对先进封装、HBM、多源拉动。无论 GPU 还是 TPU，核心瓶颈仍是 HBM 供给、系统互连、机房电力与散热。多极化 ...

When Jeff Bezos said that one breakthrough technology would shape Amazon’s destiny, even Wall Street’s biggest analysts were caught off guard. Fast forward a year and Amazon’s new CEO Andy Jassy described generative AI as a “once-in-a-lifetime” technology that is already being used across Amazon to reinvent customer experiences. At the 8th Future Investment Initiative conference, Elon Musk predicted that by 2040 there would be at least 10 billion humanoid robots, with each priced between $20,000 and $25,000 ...

When Jeff Bezos said that one breakthrough technology would shape Amazon’s destiny, even Wall Street’s biggest analysts were caught off guard. Fast forward a year and Amazon’s new CEO Andy Jassy described generative AI as a “once-in-a-lifetime” technology that is already being used across Amazon to reinvent customer experiences. At the 8th Future Investment Initiative conference, Elon Musk predicted that by 2040 there would be at least 10 billion humanoid robots, with each priced between $20,000 and $25,000 ...

2026 年 03 月 01 日 行业研究 评级：推荐(维持) | 1 研 2 1 2 C | | --- | | œ | | 研究所： | | | --- | --- | | 证券分析师： | 董伯骏 S0350521080009 | | | dongbj@ghzq.com.cn | | 证券分析师： | 李永磊 S0350521080004 | | | liyl03@ghzq.com.cn | | 证券分析师： | 杨仁文 S0350521120001 | | | yangrw@ghzq.com.cn | | 证券分析师： | 王鹏 S0350525090001 | | | wangp05@ghzq.com.cn | [Table_Title] 中国 2026 年计划实施 2 次载人飞行、1 次货运 飞船补给任务 ——新材料产业周报 最近一年走势 | 行业相对表现 | | | 2026/02/27 | | --- | --- | --- | --- | | 表现 | 1M | 3M | 12M | | 基础化工 | 6.0% | 26.1% | 52.2% | | 沪深 300 | 0.1% | 4.3% ...

1）英伟达发布 4Q25 业绩，单季度每股盈利 1.62 美元，较分析师预期的 1.53 美元高出 5.81%，同比+82.02%；单季度 营收 681.27 亿美元，较分析师预期的 660.03 亿美元高出 3.22%，同比+73.21%。同时公司给出 1Q26 780 亿美元业绩 指引，超分析师预期的 727.8 亿美元，整体业绩&指引双 beat。但从股价看市场反应温和，我们认为主要系担心竞争 格局、预期交易充分的原因。展望 3 月，NVIDIA GTC 大会即将开幕。将首次公开下一代核心产品 Feynman 芯片，且将 搭载全球首款 1.6nm 制程工艺，同时我们预计有望展示 CPO 交换机、rubin ultra 等方案，或将给再催化光通信板块。 2）旭创发布业绩快报，2025 年全年收入 382.40 亿元，我们测算对应 4Q25 单季度收入 132.35 亿元，同比+102%/环比 +30%，增长稳健。2025 年全年营业利润 135.97 亿元，我们测算对应 4Q25 单季度营业利润 47.59 亿元，同比+190%/环 比+20%。整体处于此前业绩预告中枢。利润增速小于收入增速，我们认为主要 ...