According to Business Insider, Michael Burry issued a new warning about AI stocks in a cryptic X post on Monday, specifically targeting hyperscalers like Meta, Oracle, and Microsoft that are spending massively on AI infrastructure. He estimated these five companies would understate depreciation by around $176 billion between 2026 and 2028, with Oracle potentially overstating earnings by 26% and Meta by 20% by 2028. Burry’s concerns mirror those of Jim Chanos, who recently warned that Amazon, Meta, Microsoft, Alphabet, and Apple are on track to spend $349 billion on capex this year, much of it directed toward AI hardware. Both investors worry that the rapid depreciation of semiconductor chips on a 2-3 year product cycle creates a potential time bomb for earnings calculations. Burry promised “more detail coming November 25th” in his post.
The Burry-Chanos Convergence
It’s pretty striking when two legendary short-sellers start singing the same bearish tune. Burry and Chanos are basically pointing at the same fundamental problem: these tech giants are spending like crazy on AI infrastructure that’s going to become obsolete faster than they can monetize it. The chips have a 2-3 year lifespan, but the accounting might not reflect that reality. And here’s the thing – when you’re talking about $349 billion in collective capex this year alone, even small miscalculations in depreciation schedules can create massive earnings distortions.
The Depreciation Time Bomb
So what’s really happening here? Basically, companies are treating these massive AI infrastructure investments as capital expenditures that get depreciated over time. But if the useful life of that equipment is actually much shorter than they’re accounting for, they’re effectively overstating their earnings. Burry’s suggesting this could amount to nearly $200 billion in understated depreciation across just five companies. That’s not just accounting noise – that’s the kind of discrepancy that can crater stock prices when it gets corrected. Remember Enron? Chanos certainly does.
The Monetization Reality Check
Here’s the fundamental question nobody seems to be asking: When exactly are these companies going to start making real money from all this AI spending? We’re seeing absolutely massive investments in data centers and chips, but the revenue streams remain pretty fuzzy. I mean, look at the numbers – AI spending is growing faster than income or revenue. At some point, shareholders are going to demand to see the return on investment. And if these companies can’t deliver, we could see some very uncomfortable decisions being made about cutting back on the AI gold rush.
Wall Street’s Quiet Concern
What’s really interesting is that this isn’t just two bearish investors shouting into the void. Business Insider notes that Wall Street analysts have been quietly flagging this depreciation issue since at least last year. They’re calling it a potential “time bomb” at the heart of the AI trade. The market’s been riding this AI hype wave for years now, but doubts are starting to form. Valuations are looking pretty lofty, and investors are getting nervous about whether the monetization plans will actually materialize. When you combine that with potential accounting issues, you’ve got the recipe for some serious volatility ahead.
