AI market split widens

Wall Street is rewarding semiconductor and infrastructure names while software stocks lag, signalling investors prefer hard scarcity over application promises right now. (finance.yahoo.com) Analysts point to continued bullishness on chip makers like Nvidia and AMD even as some banks argue the software selloff may be overdone, creating a clear bifurcation in capital flow across the AI stack. (thestreet.com)

Wall Street is splitting the artificial intelligence trade: money is still flowing to chip and data-center companies, while many software stocks remain under pressure. (finance.yahoo.com) Nvidia reported fiscal 2026 revenue of $215.9 billion on February 25, with data-center revenue of $62.3 billion in the latest quarter and $193.7 billion for the full year. Advanced Micro Devices reported full-year 2025 revenue of $34.6 billion on February 3, including a record $16.6 billion from data center. (nvidianews.nvidia.com) (ir.amd.com) At the same time, Goldman Sachs said in February that the software selloff reflected a fast shift in investor sentiment, not a sudden collapse in company fundamentals. A Yahoo Finance report published April 12 said Goldman built a framework and a basket of software names it believes have been punished too broadly. (finance.yahoo.com) The divide runs through the artificial intelligence stack, the layers of companies that build chips, rent computing power, and sell applications. Investors are paying up for the layer with hard capacity limits, especially advanced processors and data-center buildout, while questioning how quickly software companies can turn artificial intelligence features into revenue. (finance.yahoo.com) (goldmansachs.com) Goldman said on April 10 that global technology stocks have had one of their weakest relative periods in nearly 50 years, after worries about artificial intelligence disrupted software business models and raised questions about returns on heavy spending. The bank described the slump as a potential “technology value opportunity,” not a uniform verdict on the sector. (goldmansachs.com) In software, the core fear is that artificial intelligence agents could do tasks inside existing products without the same seat-based pricing that made software companies rich. Goldman-linked reporting said software forward price-to-earnings multiples fell to about 20 times earnings, the lowest level since 2014 and close to the smallest premium to the Standard & Poor’s 500 since 2010. (finance.yahoo.com) (newsycanuse.com) In chips, the story is still supply, scale, and backlog. Nvidia’s latest results showed gross margin above 70% and year-over-year revenue growth of 65%, numbers that reinforce why investors still treat compute as the scarce part of the market. (nvidianews.nvidia.com) That does not mean software is uniformly broken. Goldman’s recent research and related coverage highlighted buy-rated names it said could adapt, arguing the market may be treating all software companies as if artificial intelligence will erase their pricing power at the same speed. (finance.yahoo.com) (thestreet.com) The next test is earnings. If software companies show that artificial intelligence features lift sales faster than they cut pricing, the gap can narrow; if chipmakers keep posting data-center growth like Nvidia and Advanced Micro Devices just did, the market split is likely to stay in place. (nvidianews.nvidia.com) (ir.amd.com)

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.