The market might be all-in on the AI trade, but a growing chorus of traders and analysts is changing strategies.
Mega-cap tech companies like Microsoft, Meta, and Amazon have been hit particularly hard in recent weeks as their plans to spend billions on AI infrastructure have attracted growing worries.
That money has had a destination: The pinch points of the boom. The rotation could forebode trouble for Wall Street’s favorite trade.
They’re not bluffing, they’re bragging
While computer hardware companies have been soaring all year, Mag7 money has been rotating into companies like memory giant Micron Technology, which boasted 85% margins amid massive data center spending by Big Tech.
Micron hasn’t been alone. For weeks, robust results in the semiconductor and hardware corner of the AI boom have appeared to precipitate a pullback in hyperscalers that are committed to big capital investments to further their AI product ambitions.
The results suggest it’s full speed ahead for big spenders. Demand is intact, supply is limited, and pricing is not on their side. Investors are taking note.
In particular, Magnificent 7 names like Microsoft and Meta have fallen over 30% from their 52-week highs. Other names, such as Nvidia, Tesla, Amazon, and Alphabet, have seen similar pullbacks, with their stocks burning trillions in market value.
FinViz.com
Some hyperscaler hopefuls are in even worse shape: Oracle is down nearly 60% from its all-time high last September. Market newcomer CoreWeave is off 45%. Bitcoin miners-turned-hyperscaler hopefuls IREN Limited and MARA Holdings are both off over 40%.
Data center investments are in the crosshairs
Mega-cap hyperscalers aren’t the only pocket affected by the pullback. Companies downstream of the data center theme are also getting a live repricing.
While a number of data center projects are going ahead, many are not happening at the scale first thought. Due to community opposition and long lines to get connected to the electrical grid, over 75 builds worth a combined $130 billion were canceled in the first quarter of 2026, per data from Data Center Watch.
Alternative investment and private credit companies backing the theme have been among the hardest-hit parties, owing to rising worries about withdrawals and the quality of their underwriting. Prominent data center backers such as Blue Owl, TPG, Ares Capital Management, KKR, and Blackstone are among the steepest decliners. They’ve fallen between 40% and 60% from all-time highs.
There have also been pullbacks in Independent Power Producers (IPPs), a once hot-button portion of the market that exploded on the backs of big data center electricity deals. Constellation Energy has fallen 37% from its 52-week high. Vistra Energy is down over 26%, and the Sam Altman-backed Oklo, which hopes to one day build small nuclear reactors to power data centers and other infrastructure, is off over 70%.
All it takes is one
I suspect many technology operators are very confident in their vision of going “all in” on AI. At this juncture, I’d bet that companies would be more likely to embrace layoffs to improve free cash flow (FCF) or free up dollars for AI investment. In fact, that has been the play that Meta has been running this whole time, even as it lacks a cogent AI strategy.
But remember, none of this solves the core problem: The ROI equation simply does not math out at this point. AI labs and tech giants must prove that inference is not just profitable but will also pay back the hefty cost of investment. The stock declines are the first pushback the market has given to say, “Actually, we don’t believe it.”
Why this could change the AI trade
Operators are driven by the moves in their stocks, in great part because their fortunes are tied to performance and stock rewards. And with many of the aforementioned companies having lost a quarter of their value from all-time highs, you have to suspect it’s only a matter of time until one company blinks and changes its spending plans.
If one does, experts warn that it could bring on even more turbulence. A recent report warns that a reversal of “excessive” tech investments could risk a “global financial crash.” The market’s recent appraisal of hyperscalers already supports some of this worry: Through the drawdowns from recent highs, Mr. Market appears to be saying that AI spending is a hazard, not a help.
Big Tech names might find a way to make inference profitable. There are a growing number of AI labs willing to do that, even if their reputations are tarnished in the process.
But for now, money is rotating into other parts of the market. Maybe the wrong lessons will be gleaned from the recent pullback, but at this stage, you’d have to guess that investors would reward any sort of walk-back of the capex that has been promised.