New York, February 16, 2026, 16:03 (EST) — Market closed.
- U.S. markets were shut for the Presidents Day holiday; trading resumes Tuesday.
- Investors are questioning whether heavy AI infrastructure spending can deliver near-term profits.
- Nvidia’s Feb. 25 results are the next big checkpoint for the AI trade.
AI-linked megacaps have bled market value this year as investors question whether a costly push into artificial intelligence will pay off fast enough. U.S. stock markets were closed on Monday for Washington’s Birthday, leaving the sector in limbo until Tuesday’s reopen. (Reuters)
Why it matters now: the “AI trade” has started to look like a bill coming due. Capital spending, or capex — money poured into data centers, chips and other equipment — is rising just as investors are pressing for clearer profit and cash-flow payback.
That timing is awkward. Many of the same firms driving demand for AI chips are also fighting for share in AI models and cloud services, where pricing power can be fickle and competition shows up quickly.
In the last U.S. session before the holiday, AI bellwethers finished mixed. Nvidia fell 2.2% to $182.81, Alphabet slipped 1.1% to $305.72 and Meta dropped 1.5% to $639.77, while Broadcom lost 1.8% to $325.17. AMD rose 0.7% to $207.32; Microsoft and Amazon ended slightly lower.
“The market’s viewpoint is that the AI build-out trade … got too pricey,” said Andrew Wells, chief investment officer at SanJac Alpha. Nvidia CEO Jensen Huang has described demand for AI gear as “sky-high,” calling the spending push sustainable. (Reuters)
The competitive backdrop is not standing still. Alibaba on Monday unveiled a new AI model, Qwen 3.5, and said it was “built for the agentic AI era,” while touting cost and performance gains — the kind of claim that can deepen investor worries about AI becoming a price war. (Reuters)
Policy and politics are in the mix, too. India this week hosts a global AI summit as it tries to attract investment, with Google, Microsoft and Amazon among firms that have already committed a combined $68 billion in AI and cloud infrastructure investment in the country through 2030, Reuters reported. (Reuters)
The risk for AI stocks is simple: spending keeps rising while returns arrive slowly. That can squeeze free cash flow, drag on valuations and leave shares exposed if growth cools or a cheaper model resets the pricing bar.
The counter-case is that demand stays strong enough to justify the buildout, especially for the hardware suppliers. A single earnings print that shows tight supply, firm pricing and resilient cloud growth can change the mood quickly.
The next hard catalyst on most traders’ calendars is Nvidia’s fourth-quarter fiscal 2026 report on February 25, due at 2 p.m. PT, the company’s investor relations site shows. (NVIDIA Investor Relations)
Until then, Tuesday’s reopen is the first test: whether investors keep trimming AI exposure on capex anxiety, or start buying back into the group ahead of Nvidia and the next round of spending and cloud demand updates.
