Legendary Contrarian Burry Escalates Massive Bet Against AI Sector Valuations

A Widening Wager on AI Excess

Michael Burry, the investor celebrated for correctly anticipating the 2008 housing market collapse, is doubling down on his skepticism toward the artificial intelligence trade. Through an August 26 Substack post titled “Nvidia, Friends, and the Rebel Alliance,” Burry revealed that he has substantially increased his short positions across a range of technology companies integral to the AI infrastructure buildout. His expanded short stock holdings—spanning Nvidia, Oracle, Palantir, Nebius, and additional firms—now represent over 21% of his total portfolio when excluding put options. This significant allocation demonstrates the depth of his conviction that current market valuations in the AI sector have become fundamentally disconnected from sustainable earnings.

Separating Operational Strength from Stock Value

Crucially, Burry does not challenge Nvidia’s operational performance or financial results. Instead, his thesis centers on a critical distinction: exceptional business performance does not automatically translate to appropriate stock valuations if that performance cannot persist indefinitely. Burry observes that while market participants already recognize Nvidia’s commanding position, the stock has failed to sustain the explosive growth trajectories it achieved in earlier periods. His analysis suggests that Nvidia’s dominant market share and extraordinary profit margins represent a temporary economic window rather than a permanent structural advantage.

Burry’s concern extends beyond any single company. By shorting Oracle, Palantir, and Nebius alongside Nvidia, he is essentially wagering that the market has miscalculated the longevity of the broader AI economic cycle. Each company embodies a different dimension of the AI infrastructure ecosystem, yet Burry views them collectively as overpriced relative to the realistic duration of their competitive advantages. He also questions whether the substantial profits these companies generate will ultimately accrue to shareholders or be captured through other mechanisms.

Managing Tail Risk in a Crowded Trade

Aware of the dangers inherent in betting against the market’s prevailing enthusiasm, Burry has implemented protective measures. He purchased December call options with strike prices in the mid-to-high $200 range ahead of Nvidia’s recent earnings report, allocating 3.5% to 4% of his portfolio to these positions. By explicitly characterizing these calls as hedges rather than bullish speculation, Burry acknowledges the possibility of being wrong—a pragmatic approach to managing the execution risk of a contrarian position.

This hedging strategy reveals the internal tension within Burry’s argument: he can simultaneously hold that Nvidia represents a genuinely exceptional company while maintaining that its stock price does not adequately reflect the temporary nature of its advantages. What concerns him is not Nvidia’s current strength but the implicit assumption baked into current valuations—that today’s dominant AI economics will persist for decades. Burry contends that investor expectations have become disconnected from realistic timelines for technological disruption and competitive erosion in this space.

Source: Michael Burry’s Substack post “Nvidia, Friends, and the Rebel Alliance” (August 26), via U.Today. Not financial advice.