Michael Burry Is Bearish on Palantir, Which Is Up 19% Over the Past Year, but His Oracle Short Just Got a Boost From a 53% Stock Drop. Is His AI Bubble Bet Still Worth Following?

Michael Burry argues that borrowed money is propping up a market in “denial” and maintains put positions against Nvidia, Palantir, and Oracle, with expirations in September 2027 for Nvidia and Palantir and December 2027 for Oracle. His bearish bets have so far been right on Oracle but wrong on Nvidia and Palantir; he says he is holding the options because he believes the denial phase could continue. The article presents his view as a contrarian warning, not evidence that a market correction is imminent.

Analysis

The investable signal is not Burry’s crash call; it is the accounting-to-economics question around AI hardware. Longer chip lives and resilient resale values would support customer returns and repeat orders. If either assumption proves too optimistic, customers may stretch replacement cycles, weakening future GPU demand before current-period revenue visibly rolls over. That would pressure NVIDIA and could spill into AI infrastructure spending, including Oracle, while Palantir faces a separate risk: a valuation reset if growth or monetization fails to meet elevated expectations.

The timing is the problem. Long-dated puts can be directionally right and still lose through premium decay or a delayed catalyst; a historical analogy and disclosed positions do not establish attractive entry prices. The market can also keep rising while breadth deteriorates, so an index-level short is less clean than targeting company-specific expectation risk.

Over 1–3 months, focus on earnings commentary on customer returns, depreciation assumptions, order timing, and cloud/AI spending. Over 6–18 months, watch for lengthening replacement cycles or evidence that used accelerators retain value; either could validate the bear case. Conversely, sustained customer spending and stable resale economics would undermine it. The contrarian risk is that Burry’s thesis may be directionally plausible but already well known, while the timing premium remains expensive. Without option-implied volatility, strikes, and valuation context, this is a hedge/watchlist signal—not a justified broad short.

Key Decisions for Investors

• Do not copy disclosed long-dated puts on headline alone. Before considering a defined-risk NVDA put spread, verify implied volatility, skew, strike, and premium at entry; size it as a hedge rather than a core short.
• Use upcoming NVDA and ORCL disclosures as a catalyst check: flag weaker customer ROI, delayed infrastructure orders, or lengthening replacement cycles; thesis is weakened by sustained spending and credible evidence of durable accelerator resale value.
• Treat PLTR separately from the hardware thesis. Consider bearish exposure only if growth or monetization indicators disappoint relative to expectations; a general AI-capex slowdown does not by itself establish PLTR-specific downside.
• No immediate trade in MELI or DECK based on their mention as Burry holdings. Reassess only on independent fundamentals or relative valuation; their inclusion is not evidence they hedge AI exposure.