Burry's 'I told you so': Nvidia, CoreWeave, Nebius down as much as 65%
The 'Big Short' investor says he gave AI infrastructure buyers the information they needed to sell, and that the selloff runs deeper than Jensen Huang's exit and politics.

Michael Burry says he warned investors that Nvidia, CoreWeave and Nebius were headed for trouble, and the three are now down roughly 42%, 55% and 65% from their October 6 intraday highs. For executives and boards, the selloff is a live test of how much of the AI buildout's value rests on a single supplier's capital discipline and on depreciation assumptions Burry argues are too generous.
Michael Burry has a message for everyone who bought AI infrastructure stocks and watched the value slide: he tried to warn them. The investor made famous by "The Big Short" posted on X on Thursday that CoreWeave, Nebius, Nvidia and semiconductors broadly were headed for trouble, and the tape has been unforgiving since. Nvidia shares opened about 4% lower on Friday, while CoreWeave and Nebius each fell around 7%. Measured from their intraday highs on October 6, the day before the current AI drawdown began, the trio are down roughly 42%, 55% and 65% respectively. The NYSE Semiconductor Index has slumped about 30% over the same period.
"CoreWeave, Nebius, Nvidia, semis, I tried to warn you all," Burry wrote. "I didn't say short. But I provided the information one would need to sell." That second line is the whole argument. Burry is not claiming a victory lap on a short position; he is claiming he handed readers the analysis and let them decide. It is a subtle but deliberate framing from an investor who has spent the past year publishing his personal portfolio thinking on Substack rather than running other people's money, and who now needs the record to speak for itself.
The record he points to is specific. On November 11, the day before Jensen Huang resigned as Nvidia's chief executive, Burry used a Substack post to urge shareholders to "get out" of the AI chipmaker's stock. His reasoning: without the prolific dealmaker in charge, Nvidia would grow more slowly and its profit margins would shrink. A day later, on November 12, he widened the warning to Nvidia's partners, singling out Nebius and CoreWeave, the so-called "neoclouds" that build data centers and rent access to Nvidia chips and other AI infrastructure. Those companies, he wrote, have "benefited greatly from Nvidia's market power and largesse, which seems to have entered a new paradigm now."
That is the structural worry in one sentence. Neoclouds are, in effect, leveraged bets on someone else's balance sheet. They buy or lease scarce chips, build capacity around them, and sell access by the hour. When the supplier's pricing power, allocation strategy or capital discipline shifts, the downstream economics move fast and in one direction. Burry's point is that the market has been pricing these businesses as if Nvidia's largesse were permanent, and that the assumption is now being tested in public.
Burry's public skepticism is not new. On August 12 he declared that "Nebius is what the top of a boom looks like." The stock closed at $259 that day, 62% higher than its November 19 close. A week earlier, on August 6, he said he could not resist shorting Nebius and Oracle, calling it "a bit like shooting fish in a barrel." He has also shorted Nvidia, CoreWeave and chip indexes this year. The contrarian investor, one of the few people to predict and profit from the collapse of the mid-2000s housing bubble, has been loudly skeptical of the stratospheric rise in AI stocks since he pivoted from running a hedge fund to writing about his personal investments a year ago.
Strip out the scorekeeping and the substance of his bear case is a balance-sheet argument. Burry has warned that AI companies are overinvesting in fast-depreciating assets such as microchips and data centers, where the useful life assumed on the books may run far longer than market reality. He has called out aggressive accounting, circular-financing deals, and dangerous levels of debt and off-balance-sheet liabilities. For anyone underwriting data center capex, those are not vibes; they are the inputs that determine whether a project earns its cost of capital.
Burry also made clear he thinks the market is misreading the cause of the selloff. In a late Thursday post on X, he argued his concerns about Nvidia and its partners and peers run far deeper than Huang's alleged extramarital affairs and the prospect of a less AI-friendly environment with Democrats in charge of the House and Senate. "If you think this is still all about an affair and some politics, you're not thinking hard enough," he wrote.
The stakes for operators and boards are straightforward. If Burry is right that the correction reflects financing structure rather than headlines, then the most exposed companies are the ones whose growth depends on a single supplier's goodwill, whose assets depreciate faster than their models assume, and whose debt was raised against both. That is a diligence checklist, not a prediction. The counterargument is equally simple: Nvidia's opening dip of about 4% is a fraction of the 42% drawdown in its most levered partners, and the market may be repricing concentration risk rather than the AI thesis itself. Either way, the burden of proof has shifted to the borrowers.
This story's Key Insights and Take-aways are locked.
Create a free account to unlock Executive Actions for one credit.
Register to UnlockAlways free for Executives Club members. Join the Club
More in Business
S&P 500 Enters Correction, Down 10% From Its High on AI Worries
The index's slide past the 10 percent line is a symbolic verdict on investor mood, and the AI trade that powered the rally is now the thing testing it.
Dow and S&P 500 enter correction as Nasdaq logs worst week since 2020
All 11 S&P 500 sectors fell more than 1% in the same week for the first time since April 2025, a breadth signal that matters more than any single index level.
Chip stocks lose $2.3 trillion in a week, worst run in 25 years
A $2.3 trillion weekly wipeout in semiconductors is forcing every board with an AI story to ask whether it owns a thesis or just a crowded trade.




