Palantir’s $40 billion wipeout exposes the fragility of AI stock valuations when growth narratives collide

Palantir’s stock lost $40 billion in market value in three days because a single tweet reframed its growth story. Michael Burry didn’t just criticize the company — he asserted that Anthropic, a competitor, scaled from $9 billion to $30 billion in annualized revenue within months, while Palantir took 20 years to reach $5 billion. That comparison ignited a selloff, dragging the stock from $148 to $127, its steepest weekly drop in a year.

Burry didn’t walk back the call. He confirmed holding put options betting Palantir will fall below $50 by 2027 — a 65 percent drop from current levels. He called the stock “wildly overvalued,” even after its retreat from a $207.52 high. The market listened. Palantir now trades at $139.53, down 28 percent year to date.

The valuation remains extreme: 99 times expected 2026 earnings. That’s nearly five times the software sector median. The bet on Palantir isn’t just about growth — it’s about exclusivity. The company’s deep integration in classified government systems, like the Pentagon’s Maven Smart System, creates a moat commercial AI firms can’t cross. Anthropic’s API can’t operate in secure environments. That’s why bulls, including Wedbush’s Dan Ives, dismiss Burry’s thesis as fiction.

But the threat isn’t in government contracts. It’s in corporate boardrooms. Burry’s point is structural: why pay millions for Palantir’s on-site engineers when a CIO can access powerful AI through a few hundred thousand dollars in API calls? Cathie Wood’s ARK Invest bought $11.15 million in shares, betting the commercial momentum — which saw US revenue surge 137 percent in Q4 2025 — will hold.

The market is split. Consensus targets imply 40 to 50 percent upside. But if the cost-benefit calculus shifts toward lightweight AI, Palantir’s growth, and its premium, won’t survive.