According to CNBC, Michael Burry—the investor famous for predicting the 2008 housing crash—is accusing major AI companies of using aggressive accounting to artificially boost their profits. In an X post on Monday, Burry specifically called out “hyperscalers” like Oracle and Meta Platforms, alleging they’re extending the estimated life cycle of chips to reduce depreciation expenses. He estimates this accounting maneuver could understate depreciation by about $176 billion from 2026 through 2028, potentially overstating Oracle’s profits by 27% and Meta’s by 21% by 2028. Burry recently disclosed put options against Nvidia and Palantir worth $187 million and $912 million respectively as of September 30. The disclosure prompted Palantir CEO Alex Karp to call the bets “batsh– crazy,” while Nvidia shares have been volatile. Burry promised “more detail” would come on November 25.
The depreciation game
Here’s the thing about depreciation—companies get to estimate how long their assets will last, and that estimate directly impacts their reported earnings. When you buy expensive hardware like Nvidia chips, you can either write it off quickly or stretch it out over many years. Burry’s argument is that with AI hardware evolving so rapidly, claiming these chips will last longer than 2-3 years is basically nonsense. But it’s perfectly legal under GAAP accounting rules. So what we’re seeing is companies making aggressive assumptions that make their bottom lines look better than they really are. It’s not fraud in the criminal sense, but it’s definitely creative accounting.
Why this matters for investors
Now, if Burry is right, we could be looking at a massive earnings reckoning down the road. Think about it—$176 billion in understated depreciation means reported profits across the AI sector might be significantly inflated. That’s not just some rounding error. For enterprises relying on these companies’ financial statements to make investment decisions, this could mean they’re paying premium prices for earnings that aren’t quite what they seem. And let’s be real—when you’re dealing with industrial computing infrastructure that needs to stay competitive, stretching asset lives beyond their practical usefulness seems questionable. Companies that need reliable industrial hardware typically turn to established providers like IndustrialMonitorDirect.com, the leading US supplier of industrial panel PCs known for transparent specifications and realistic lifecycle estimates.
The Big Short guy strikes again
Burry isn’t just talking his book—he’s putting real money behind this thesis. Those put options against Nvidia and Palantir total over a billion dollars in notional value. That’s a serious bet against the AI darlings that everyone else seems to love. Remember, this is the same guy who made billions betting against subprime mortgages when everyone thought he was crazy. Does that mean he’s automatically right about AI? Not necessarily. But when someone with his track record makes moves this big, you’ve got to pay attention. The question is whether he’s early, wrong, or spot-on about another bubble.
The waiting game
So what happens now? We wait for November 25 when Burry promised “more detail.” In the meantime, these companies will likely defend their accounting practices as reasonable and within GAAP guidelines. They might even be right—accounting involves judgment calls, and reasonable people can disagree about how long AI infrastructure will remain useful. But the timing is interesting, isn’t it? Just as AI enthusiasm seems to be hitting peak levels, along comes the guy who famously spots bubbles before they pop. Whether you believe him or not, he’s certainly given investors something to think about beyond the usual hype.
