Michael Burry, the man who famously shorted subprime mortgages, is back in the spotlight. This time, his target isn’t housing, it’s AI. Over the past weeks, Burry has called out hyperscalers like Nvidia, Microsoft, Oracle, and Meta for what he sees as aggressive accounting tricks: stretching out depreciation on expensive AI chips and servers to make earnings look smoother than reality allows.
In plain terms, he’s saying profits are inflated and the AI trade is overvalued.
Burry’s disclosed bets are eye-catching. He’s spent millions buying put options on Palantir and Nvidia, signalling he expects a meaningful correction. With Palantir trading at $178 and Burry’s put strikes at $50 for 2027, the gap tells you just how bearish he is. The market, it seems, is starting to pay attention: AI stocks have been hit this week as investors re-evaluate sky-high valuations.
What makes this moment particularly interesting is that Burry has closed his hedge fund↗︎, Scion Asset Management, echoing the move he made in 2008 before the housing crash. History suggests that when Burry steps away from managing outside capital, it often signals he’s preparing for a significant structural risk in the market, something he clearly sees in AI.
His critique isn’t purely about short-term market swings. It’s about structural risk. AI infrastructure evolves fast. Chips bought today can be obsolete in a few years. If companies depreciate them over five or six years instead of the two or three that matches reality, earnings are artificially propped up. Burry estimates this could inflate sector earnings by roughly $176 billion through 2028. That’s a lot of optimism built on stretched accounting.
Critics may argue Burry is early, or even wrong. AI adoption is accelerating, and revenue growth could keep valuations elevated for longer than expected. Momentum alone can carry stocks higher, leaving early shorts nursing losses for years. But Burry isn’t dabbling; he’s signalling conviction. This is the kind of calculated, contrarian move that made him a legend in 2008.
So, is he right? On fundamentals, his argument makes sense. Rapid tech cycles, enormous capital expenditures, and stretched depreciation could create a reckoning. On timing, markets may have already started questioning the AI frenzy. But as with any contrarian bet, patience is critical. Burry may be early, but in the long run, history suggests the structural risks he highlights won’t stay hidden forever.
For investors, the takeaway isn’t just whether to buy or sell AI stocks. It’s a reminder to look past hype, understand the economics under the flashy headlines, and consider the cracks that even the most celebrated narratives can hide.
