According to Business Insider, Michael Burry has doubled down on his criticism of AI giants and revealed he’s actively betting against both Nvidia and Palantir. In a Tuesday Substack post titled “Unicorns and Cockroaches: Blessed Fraud,” the Big Short investor called Nvidia’s recent memo to Wall Street analysts “disappointing” and full of “straw man” arguments. Burry disclosed he holds put options on both companies with a combined notional value of $1.1 billion, though the positions only cost him around $10 million each. Nvidia shares have already slumped 14% from their November 3 high as AI concerns mount. The investor specifically questioned AI companies’ depreciation accounting practices, suggesting they’re artificially inflating short-term profits by spreading chip costs over 5-6 years instead of 3.
The depreciation time bomb
Here’s what’s really interesting about Burry’s argument. He’s not just saying AI stocks are overvalued – he’s pointing to something much more specific in the accounting. Companies can make their current numbers look better by claiming their AI chips will last longer. But if those chips actually become obsolete faster than expected? You get massive writedowns down the road. And Burry thinks that’s exactly what’s coming between 2026 and 2028. Basically, he’s saying the math doesn’t add up when you’re dealing with technology that evolves this quickly.
nvidia-s-questionable-response”>Nvidia’s questionable response
What’s really telling is how Burry describes Nvidia’s memo. He calls it “disingenuous on the face” and says it “almost reads like a hoax.” That’s pretty strong language from someone who usually lets his trades do the talking. The memo apparently tried to rebut claims Burry never made – like suggesting Nvidia was dragging out depreciation of its own equipment. But as Burry points out, Nvidia is primarily a chip designer, not a manufacturer, so their own depreciation isn’t even the issue. It’s the hyperscalers like Microsoft and Google who are buying these chips by the truckload. When even Microsoft’s CEO is talking about slowing data center buildouts because he’s worried about chip generations becoming obsolete, maybe Burry has a point.
The bigger AI bubble concern
Look, Burry has been wrong before – his track record since The Big Short has been mixed at best. But he’s putting real money behind this thesis, and the timing is interesting. AI stocks have been on an absolute tear, and any hint of skepticism tends to get shouted down. Yet here we have one of the most famous contrarian investors basically saying the emperor has no clothes. The fact that his relatively small $20 million in option premiums could potentially pay out on $1.1 billion in notional value tells you how leveraged this trade is. He’s betting that the AI bubble isn’t just going to deflate slowly – he’s positioned for something much more dramatic.
What happens now?
So where does this leave us? Nvidia and Palantir are two of the most loved names in the AI space, and having Michael Burry publicly short them definitely adds to the bear case. But remember – being early is the same as being wrong in markets. Burry was famously early on the housing crash too. The real question is whether his depreciation argument holds water. If AI progress actually does slow down, and chips don’t become obsolete as quickly as he fears, these extended depreciation schedules might look perfectly reasonable. But if he’s right? Well, let’s just say those $10 million bets could end up looking pretty smart. Either way, it’s going to be fascinating to watch.
