Michael Burry Disclosed Put Options Worth an Estimated $1.1 Billion Against Nvidia and Palantir Before Closing His Hedge Fund. Was That Bearish Bet Ahead of Its Time?

Michael Burry maintained bearish exposure to AI leaders Palantir and Nvidia after previously disclosing put positions with roughly $912 million and $186,580 in underlying notional value, respectively. He sold December-expiring puts but retained short positions in both stocks and Palantir puts expiring in 2027 as part of a portfolio-wide risk reduction. The thesis centers on Palantir's valuation above 100x earnings and Nvidia's customer-concentration risk should hyperscaler AI capital spending weaken.

Analysis

This is primarily a positioning headline, not new fundamental information. The key market implication is asymmetric: PLTR is more vulnerable than NVDA because its valuation leaves less room for even a modest deceleration in commercial growth or operating-margin progression, while NVDA has a broader earnings-support mechanism through AI infrastructure demand and a deeper institutional ownership base. A high-profile bear can increase demand for downside hedges, but without a negative revision to forward estimates it is unlikely to create durable selling pressure by itself.

The more relevant near-term signal is the removal of short-dated downside exposure. That reduces the probability of a forced year-end catalyst and suggests investors should not infer conviction about an imminent drawdown from a reported notional figure; option notional is not capital at risk and regulatory disclosures do not establish current strikes, premiums, or net exposure. Over the next 1-3 months, the decisive variables are hyperscaler capex guidance, NVDA supply/lead-time commentary, and PLTR's commercial bookings and remaining-performance-obligation trajectory. A broad risk-off move could compress both multiples, but PLTR's lower margin of valuation safety should produce greater beta.

Contrarian view: consensus may be too focused on whether AI spending is a bubble and too little on the split between infrastructure monetization and application-software monetization. NVDA can sustain earnings even if customers rationalize spend, provided compute intensity and replacement cycles remain high; PLTR needs evidence that deployments convert into repeatable, large-scale revenue rather than remaining concentrated in a limited set of programs. The clean falsifier for a PLTR-underperformance thesis is sustained acceleration in commercial revenue and raised multi-year margin targets; for NVDA, it is continued capex increases from major customers alongside stable gross-margin guidance.

Key Decisions for Investors

• Do not trade solely on the disclosed bearish positioning; treat it as a volatility/hedging flow watch item rather than a directional catalyst until earnings estimates or customer capex plans change.
• Initiate a 1-3 month relative-value position: short PLTR / long NVDA in dollar-neutral sizing after any PLTR relief rally. Target 10-15% relative downside in PLTR versus NVDA; exit if PLTR raises commercial-growth guidance or if NVDA faces a material gross-margin or hyperscaler-demand reset.
• For portfolios with concentrated AI exposure, buy PLTR put spreads rather than outright puts, using 3-6 month maturities. This limits premium decay risk while protecting against a valuation compression; avoid naked long-dated puts until implied volatility and strike-specific pricing are reviewed.
• Set earnings alerts for hyperscaler capex revisions and NVDA customer-concentration disclosures. A single major customer reducing AI infrastructure budgets would be a reason to add NVDA downside hedges; absent that evidence, maintain NVDA exposure rather than extrapolating PLTR's valuation risk to the full AI complex.