Michael Burry Buys Put Options on Nvidia, Micron, and Palantir as "The Big Short" Investor Predicts a 1987-Style Market Crash

Michael Burry has covered short positions and bought put options on Nvidia, Micron Technology, and Palantir, warning that AI enthusiasm, leverage, and crowded momentum positioning could produce a sharp market correction. His bearish case centers on potentially aggressive depreciation assumptions for AI infrastructure and a future memory-chip downcycle as supply from China, Samsung, SK Hynix, and Micron expands. The article argues that Nvidia, Micron, and Palantir differ from dot-com-era speculative companies because they serve profitable hyperscalers with real demand, while cautioning investors not to follow Burry's public trades without independent diligence.

Analysis

The actionable signal is not the disclosed bearish positioning itself; without strike, expiry, premium, and portfolio-size data, it cannot distinguish a directional view from a hedge. Public attention can nevertheless raise near-term realized volatility in NVDA, MU, and PLTR, particularly if dealer positioning is short gamma into earnings or CPI/FOMC dates. This is a trading-liquidity effect measured in days to weeks, not evidence of a fundamental inflection.

The key fault line is AI infrastructure monetization versus capex: MSFT, AMZN, GOOG, META, and ORCL can absorb investment from internal cash flow, but their returns on incremental compute must eventually appear in cloud growth, advertising yield, enterprise pricing, or retention. If monetization lags, the first adjustment is likely a hyperscaler capex-growth reset, which would pressure NVDA's forward order visibility and the broader semiconductor multiple before it necessarily impairs current revenue. Conversely, sustained cloud backlog growth and raised capex guidance through the next two reporting cycles would invalidate the near-term bubble framing.

MU has more asymmetric cycle risk than NVDA because memory earnings are highly sensitive to incremental supply, but HBM is not a commodity-equivalent market: customer qualification, packaging capacity, and yield curves can delay effective supply despite announced fab additions. The bearish case becomes investable only if HBM pricing or bit-demand guidance weakens while inventory days rise; absent those indicators, a generic memory-cycle short is premature. PLTR remains the cleaner valuation-duration short within the group because a modest deceleration in commercial growth can drive multiple compression even if absolute revenue remains strong.

Key Decisions for Investors

• Do not follow disclosed NVDA/MU/PLTR puts without contract details; set an alert for unusually elevated 30-60 day implied volatility versus realized volatility and consider selling defined-risk call spreads only after earnings, not ahead of them.
• Initiate a 1-3 month relative-value position: long MSFT / short PLTR in equal beta-adjusted dollars. MSFT has funded AI spend and monetization channels, while PLTR has greater sensitivity to a growth or valuation reset; exit if PLTR commercial-growth guidance accelerates materially or MSFT Azure growth decelerates.
• Keep MU as a watch-list short rather than an active short until evidence emerges of weaker HBM pricing, rising inventory, or adverse gross-margin guidance. If two of those occur, express via 3-6 month put spreads to cap squeeze risk from continued supply tightness.
• For portfolio hedging over the next 1-3 months, prefer limited-risk QQQ or SMH put spreads over single-name NVDA puts: a capex-reset shock would likely reprice the AI complex broadly, while idiosyncratic upside from customer concentration, product launches, or supply constraints can defeat a single-name short.