Wolfe Research warned in a client note Tuesday that market participants may be assuming an unsustainably strong trajectory for artificial intelligence investment even as markets continue to show elevated volatility.
The firm pointed to recent trading patterns as evidence of how AI-related disruption risk has changed investor behavior. In its note, Wolfe said that investors have adopted a "sell first, ask questions later" stance, allowing even the most benign headlines to push certain industries lower by more than 5%.
Analyst Chris Senyek highlighted near-term market reactions as an example of that dynamic. He noted that the S&P 500, Nasdaq 100 and Russell 2000 each declined between 1% and 1.6% during Monday's trading session.
At the center of Wolfe's argument for 2026 is the durability of hyperscaler capital expenditures. The note observes that "over the course of 2025, hyperscalers capital expenditures surprised to the upside versus consensus estimates," but adds that "the key question this year is whether they will continue this furious rate of spending."
Wolfe Research also flagged growing constraints that could begin to impede the buildout of large-scale AI infrastructure by the second half of this year. Those constraints include limitations tied to power generation, increases in material costs and potential regulatory hurdles, any of which could pressure project timelines and execution.
The firm laid out the market implications of a change in spending behavior. It argued that "a delay or cut to capital spending would be a very positive catalyst for areas of the market that have been most impacted by AI," while at the same time warning that such a shift could precipitate "a large drawdown for the most crowded stocks levered to AI spending (e.g., semis, industrials)."
In short, Wolfe's view is that the combination of heightened sensitivity to AI headlines and uncertainty around whether hyperscalers will sustain above-consensus capex raises the odds of meaningful market moves if spending falters or slows.
Clear summary
Wolfe Research cautions that investor expectations for continued aggressive AI investment may be overly optimistic. The firm points to last year's upside surprises in hyperscaler spending and warns that constraints - including power, materials and regulation - could hinder further buildouts, with meaningful ramifications for sectors most exposed to AI-driven capital expenditure.
Key points
- Investor behavior has become more reactive to AI headlines, amplifying moves in affected industries.
- Hyperscalers surprised to the upside on capital expenditures in 2025; the critical question for 2026 is whether that pace can be maintained.
- A shift in capex trends would materially affect sectors tied to AI spending, notably semiconductors and industrials, and broader market segments sensitive to tech investment.
Risks and uncertainties
- Operational constraints - including limitations in power generation - that could slow large-scale AI infrastructure projects.
- Rising material costs that may increase project expenses and delay timelines.
- Regulatory hurdles that could create impediments to deployment and affect capital spending plans.