Michael Burry, famous for shorting U.S. subprime mortgages in 2008, claims the bubble formed by AI spending will burst earlier than his original benchmark
Scion's Burry: Bubble timing moved up
Michael Burry, who gained fame for shorting U.S. subprime mortgages in 2008, stated that the bubble formed by AI spending will burst earlier than his original benchmark, expressing greater confidence than before, with a downturn possibly arriving within a year. The original benchmark year was 2028.
In an article from late September 2026, he wrote that he is moving the timeline forward, thus providing more leverage for short positions. A better timeline makes leverage more acceptable. He cited analysis from Ares Management, pointing to a fragile connection between large capital expenditures and actual revenues, stating that a significant portion of data center capital expenditures is financed by debt.
Positions have shifted from direct shorting to put options. Micron's options have a strike price of about $500, expiring in June 2027, with an implied decline of over 50%. Nebius's strike price is in the double digits, also pointing to June 2027. Nvidia's options expire in September 2027, with a strike price in the mid-$100s, while the stock price was above $220 at the time of position building. Palantir's put options have been added with a strike price of about $100, expiring in September 2027. The semiconductor index fund SOXX has a strike price at a low of $400. Oracle and the Nasdaq 100 have also been included in similar positions.
On September 29, 2026, he wrote again that for the benefit of humanity, the market should decline significantly to prevent OpenAI and Anthropic from going public. He claimed these companies would siphon off and destroy trillions of dollars in capital, which is still a smaller part of the damage. OpenAI secretly submitted its IPO documents in June. Sam Altman stated that this year is not a good time for an IPO, with the target pushed to 2027.
Scion Asset Management is no longer reporting positions as a hedge fund. Semiconductor stocks and the Nasdaq have remained resilient even after he increased his put options. Micron fell after disclosure but remains well above the $500 strike price. The report did not provide the premiums and contract quantities for these options.
Burry, expressing a downturn with put options, contrasts with buyers who are still paying high valuations for chip, cloud, and model companies. The event-driven narrative arises from the mismatch between capital expenditures and revenues, not from the macro data of the day. The benefiting narrative falls on the premiums collected by option sellers, as long as the stock price does not fall to the strike price. The pressured narrative falls on Micron, Nvidia, Palantir, and model companies preparing to go public. He has previously claimed that downturns would occur earlier, but the market did not decline at those times.
Source: Public information
ABAB AI Insight
Burry's rise to fame was from 2005 to 2007 at Scion, shorting subprime mortgages using credit default swaps, with profits in the hundreds of millions. The 2015 film "The Big Short" portrayed him as a character played by Christian Bale. He has repeatedly applied the same method to the public market: in 2022, he compared that year to watching a plane crash, and publicly called for selling in early 2023, after which the index rose. In 2026, he shifted his focus from mortgage-backed securities to AI capital expenditures, moving the benchmark from 2028 to within a year.
Money is now placed in options rather than short selling. Directly shorting Micron and Nebius has been replaced with put options expiring in June 2027, with strike prices around $500 and in the double digits, while Nvidia's options are set for September 2027 at mid-$100s. His motivation is expressed in his own words: the timeline has moved closer, making leverage worthwhile. Put options cap losses at the premium paid, while retaining odds of a decline exceeding 50%. Another expression is the hope for a significant market drop to block the IPOs of OpenAI and Anthropic, preventing trillions of dollars in capital from being siphoned off.
The analogy is to his own subprime positions in 2008, also close to shorting internet infrastructure capital expenditures in 2000. The difference is that back then, mortgage defaults had monthly data, while this time he uses Ares's research on the disconnect between capital expenditures and revenues, along with data center debt. The current position is not a crisis confirmation but rather a narrative control phase: the index remains above his strike price, and Micron is also well above $500.
Structurally, this represents a transfer of pricing power. Model companies want to use IPOs to sell private valuations to the public market, while Burry demands the market to drop first to close this exit. The mechanism is that if capital expenditures rely mainly on debt and revenues do not keep up, the pricing power of valuations will shift from the funds buying into the narrative to the marginal sellers holding put options. The premium has already been paid, and pricing will depend on whether the stock price hits those strike prices before the options expire in 2027.
ABAB News · Cognitive Laws
Put options buy the timing, not the direction itself.
Someone who calls early ten times may still hit once.
An IPO is an exit; a crash closes that exit.