美国企业刚刚交出了近年来最强劲的一份财报季答卷，但美国股市并未随之起舞，强劲的盈利表现与低迷的市场走势形成了显著的脱钩。 据彭博报道，标普500指数成分股公司第四季度的盈利增长了13%，比预期高出近6个百分点。然而，在以摩根大通和沃尔玛公司发布财报为起止 的六周时间里，标普500指数却下跌了1.7%，追平了过去10个季度财报季期间的最差表现。 人工智能引发的"恐慌交易"、全球地缘政治风险以及对私募信贷的担忧，共同压制了市场的乐观情绪。投资者正对那些容易受AI技术应用冲击的 行业进行快速重新定价，导致市场资金从高估值板块流出，转向更为安全的资产。 尽管短期内多重不确定性导致美股呈现横盘震荡态势，但市场参与者依然认为，企业基本面的韧性最终将主导市场走向。一旦投资者理清了AI颠 覆性影响的范围与步伐，美股有望重拾升势。 强劲业绩遭遇市场冷眼 第四季度美国企业的盈利基本面极其稳健。据彭博援引Jefferies Financial Group Inc.的数据显示，不仅盈利增速超出预期，企业对来年的盈利前景 也充满乐观。罗素3000指数中，上调业绩指引的公司数量与下调指引的公司数量之比达到了4比1，这一比例上次出现还是在 ...

深度价值投资者李·罗奇（Lee Roach）指出，当市场预期一家公司会"立即死亡"，而实际上它只是"缓慢 衰退"时，投资回报可能非常惊人。罗奇认为，尽管人工智能（AI）削弱了软件公司的"护城河"，但这 并不意味着其终值为零。许多软件股的市盈率已跌至2008年危机以来罕见的水平。 来源：滚动播报 ...

Core Insights - DISCO has launched an all-inclusive platform for litigation, investigations, and regulatory responses, integrating eDiscovery, Cecilia AI, deposition management, and timelines into a single solution at a transparent price [1] - The new platform aims to address the inadequacies of one-off eDiscovery products by providing a seamless end-to-end solution for modern litigation [1] - A new pricing model has been introduced, featuring a single, transparent per GB price on processed data without ingest fees, facilitating clearer cost comparisons [1] Group 1: Product Features - The DISCO platform includes premier eDiscovery technology for instant access to critical data [1] - Cecilia AI offers advanced AI-powered legal tools to streamline and accelerate litigation processes [1] - Deposition management and timeline capabilities are now integrated into the platform, previously sold separately as Case Builder [1] Group 2: Strategic Direction - The company emphasizes the importance of integrated tools that address the full litigation lifecycle for complex matters [1] - DISCO's CEO highlighted the need for innovation in response to fast-evolving generative AI capabilities [1] - The platform will be showcased at Legalweek, indicating the company's commitment to demonstrating its advancements in legal technology [1]

2026 has gotten off somewhat to a rocky start. At the time of writing, the S&P ( SP500 ) is in the red, driven by the decline in Technology ( XLK ) stocks as AI disruption fears in software continueFormerly known as "The Dividend Collectuh." Top 1% of financial experts on TipRanks. Contributing analyst to the iREIT+Hoya Capital investment group. Dividend Collection Agency is not a registered investment professional nor financial advisor and these articles should not be taken as financial advice. This is for ...

Core Insights - Workday reported strong Q4 FY2026 results with a non-GAAP EPS of $2.47, exceeding the consensus estimate of $2.32, and total revenue of $2.532 billion, reflecting a year-over-year growth of 14.5% [1] - The company's guidance for FY2027 subscription revenue is projected between $9.925 billion and $9.950 billion, indicating a growth rate of 12-13%, which is a deceleration compared to FY2026 [1] - Following the guidance announcement, Workday's shares fell 8% in after-hours trading and nearly 13% in premarket trading, signaling investor concerns about future growth [1] Financial Performance - Workday's Q4 FY2026 non-GAAP EPS increased by 28.6% year-over-year to $2.47 [1] - Total revenue for the quarter reached $2.532 billion, driven by subscription revenue of $2.36 billion, which grew by 15.7% year-over-year [1] - The non-GAAP operating margin improved by 420 basis points to 30.6%, and full-year free cash flow rose by 26.86% to $2.777 billion [1] Guidance and Market Reaction - The FY2027 subscription revenue guidance of $9.925 billion to $9.950 billion represents a slowdown from the previous year's growth rate [1] - The 12-month subscription backlog stands at $8.833 billion, reflecting a 15.8% increase, which suggests sustained demand [1] - The market's reaction to the guidance was negative, with shares dropping significantly, raising questions about whether the guidance reflects conservative estimates or actual demand weakness [1] Strategic Initiatives - Co-founder and CEO Aneel Bhusri emphasized the company's commitment to innovation in HR and finance through AI, highlighting the completion of 1.7 billion AI actions across its platform in FY2026 [1] - Workday has made strategic acquisitions, including Paradox, Sana, and Pipedream, to enhance its AI capabilities [1]

Core Viewpoint - Oracle has experienced significant volatility, dropping 54.9% from its all-time high, yet it holds potential for substantial returns if its ambitious cloud revenue projections materialize [2][4]. Group 1: Company Overview - Oracle has transitioned from a legacy software company to a key player in the AI and cloud infrastructure space, aiming to compete with major players like Amazon, Microsoft, and Alphabet [1][4]. - The company’s database and data management software segment is high-margin and generates substantial free cash flow, but it is insufficient to support Oracle's cloud infrastructure ambitions [3]. Group 2: Financial Projections - Oracle aims to grow its Oracle Cloud Infrastructure (OCI) revenue from $18 billion in fiscal 2026 to $144 billion by fiscal 2030, indicating a strong growth trajectory [4]. - The projected OCI revenue would surpass Amazon Web Services' current revenue of $128.7 billion, highlighting Oracle's aggressive growth strategy [5]. Group 3: Investment Thesis - If Oracle's forecasts are accurate and OCI achieves margins similar to AWS's 35.6% in 2025, the stock could yield significant returns for investors over the next five years [6]. - The investment thesis is contingent on Oracle's ability to manage its debt and leverage its remaining performance obligations (RPO), which stood at $523 billion, with $300 billion linked to OpenAI [7].

原来AI也在水字数。 大模型的思维链越长，推理能力就越强？谷歌Say No—— token数量和推理质量，真没啥正相关，因为token和token还不一样，有些纯凑数，深度思考token才真有用。 新研究抛弃字数论，甩出衡量模型推理质量的全新标准DTR，专门揪模型是在真思考还是水字数。 基于DTR，还提出了Think@n策略，让GPT-OSS、DeepSeek-R1等推理模型实现准确率不降、算力成本直接砍半的效果。 长逻辑不等于好推理 长期以来，一个比较常见的观点是思维链越长越牛。 这种思路的逻辑也比较直接，推理步骤多=思考更充分=答案更准。 于是不少研发者也为了追求长推理轨迹开始堆算力。 谷歌的研究团队在AIME2024/2025、HMMT 2025、GPQA-Diamond四个数据集上，测了GPT-OSS、DeepSeek-R1、Qwen3等8个模型变体； 结果发现，token长度和准确率的平均相关系数是-0.54……负相关。 也就是说，在某些情况下，思维链越长，推理越容易跑偏，甚至还会陷入逻辑死循环或者过度推理。 那么问题来了：如果长度靠不住，那该怎么判断模型是不是在真思考？ 谷歌这次的视角比较有意思， ...

PresentationGreat. Thanks, everyone, for joining us for the next session. We're delighted to have SAP back at the conference and particularly Sean Kask, who's the Chief AI Strategy Officer. Sean, I think this is probably third year in a row coming back, but we live in interesting times in the -- on the kind of AI front, there seems to be almost an exponential curve in terms of the development. So I wanted to sort of kick off and get your perspective on -- in the past 12 months, what have been the really big ...

Deep value investor Lee Roach is looking at beaten-down software stocks for the first time ever. ...

Anthropic这两天又成了风口浪尖。 前脚刚因为指责国内公司「蒸馏」数据，被AI圈广泛群嘲。 就连马斯克都忍不住开怼： 你们自己训练模型时不也未经许可窃取版权书籍，还赔了十几亿美元和解金？贼喊捉贼啊！ 后脚就又搞出大事。 要知道，过去半年Claude已经两次把全球SaaS公司吓得魂飞魄散。 第一次，Cowork上线法律工具，软件板块市值蒸发万亿； 第二次，Claude Code Security剑指代码安全，网安板块集体崩盘。 | Ticker | | Company | Price | Market Cap | P/S | P/E | % YTD Chart 1Y | % 1Y | A 52w High | RS Rank 1M 20SMA 50SMA 200SMA | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | | FIG | Figma | | $27.19 | $13.58 | 13.91 | n/a | -28.62% | 16 85 | -80.98% | V | A | ...