Michael Burry Says Anthropic's Private Mark Could Buy 78 S&P 500 Companies

Michael Burry posted a number this week that's worth sitting with. Anthropic's private valuation, he says, could buy 78 profitable companies in the S&P 500. Domino's. Clorox. lululemon. Hormel. That's the shortlist he named.

Notably, Burry didn't specify which valuation he used to run the math. The company's last private round, in May, priced it at $965 billion ahead of a planned IPO. He also drew a parallel to UPS's pre-IPO peak in the 1990s, a comparison that's less about logistics and more about what happens when private marks run ahead of public earnings power.

Here's what matters: Anthropic isn't public yet, so there's no daily price discovery to argue with. That $965 billion figure came from a private round. It reflects a handful of term sheets, not millions of buyers and sellers finding equilibrium. Burry's point isn't that the models are bad. It's that the price being paid for exposure to those models has decoupled from anything you can underwrite on cash flows.

A 78-company basket of Domino's, Clorox, lululemon, and Hormel throws off tens of billions in annual free cash flow, pays dividends, and trades on multiples the street can actually model. Anthropic, at the reported round, is being valued on a thesis. That's not a knock on the company. Early-stage pricing always runs on conviction about the future. But when the number gets this large, the margin for error shrinks fast.

So is Burry right? The reality is nobody knows yet, and that's the problem. Private AI marks have been climbing on limited float and heavy strategic capital. When there's no public tape, there's no forcing function. The UPS analogy is the part worth watching. That company eventually came public and the market repriced it.

From a risk perspective, the number to watch isn't 78. It's the first public filing. That's when the thesis meets a market that can say no.