Michael Burry, the investor who famously profited from the 2008 housing collapse, is now putting real money behind his belief that the AI-fueled stock rally has gone too far. His fund, Scion Asset Management, held put option positions on Nvidia as of September 30, 2025, according to the firm’s quarterly filing with the Securities and Exchange Commission. In a Substack post, Burry warned the market has “jumped the shark,” drawing a direct comparison to the dot-com peak of 2000, a warning later summarized in Bloomberg coverage of his remarks.
Scion’s Nvidia put options and the dot-com parallel
The tension behind Burry’s warning is straightforward: one of the most closely watched contrarian investors in the world has placed documented bearish bets against the same AI chip stocks that have driven broad market gains over the past two years. Nvidia, in particular, has become a symbol of the enthusiasm around artificial intelligence infrastructure, with its valuation and index weight rising alongside demand for high-performance graphics processors.
According to the Bloomberg account of his Substack essay, Burry pointed to the rapid rise in a semiconductor benchmark and argued that the surge in chip names resembled the final, euphoric phase of the late-1990s technology boom. Back then, investors extrapolated early internet growth into perpetuity; now, he suggested, markets may be doing something similar with AI-related earnings and spending forecasts.
That comparison carries weight because Burry built his reputation on a willingness to bet against consensus during the mortgage crisis, when he shorted subprime securities long before most investors recognized the risks. At the same time, his track record since then has included several high-profile bearish calls that arrived early or failed to play out. The current wager on Nvidia therefore sits at the intersection of his history as a crisis-era skeptic and the reality that even skilled macro calls can be mistimed for years.
For investors whose portfolios are heavily tilted toward growth stocks or tech-weighted index funds, the question is not whether Burry will be exactly right, but how much risk they are assuming if he is directionally correct. His stance underscores how concentrated recent gains have become in a small group of mega-cap technology and semiconductor names, leaving diversified investors exposed to a narrow set of themes.
What Scion’s SEC filing actually shows
Scion Asset Management filed its quarterly Form 13F report on November 3, 2025, covering the period that ended September 30, 2025. The information table lists a put option position tied to Nvidia Corporation, with the security type labeled as an option and the put designation clearly indicated.
The accompanying plain-text filing confirms the same entry, showing Nvidia alongside the notation “PUT” in the relevant column. That detail establishes that Scion held bearish exposure to Nvidia at the quarter’s close, rather than merely trimming a long position or rotating into other growth names.
However, the 13F format leaves out many of the specifics that traders might want. It discloses the notional value of the position as of the reporting date but omits the strike price, expiration date, and the exact number of option contracts. Without those fields, outside observers cannot determine how far Nvidia’s share price would need to move, or over what time frame, for the position to generate significant profits or losses.
That limitation is important because it constrains the conclusions investors can draw. The filing confirms Burry’s directionally bearish stance on Nvidia at a point in time, but it does not reveal whether he is targeting a sharp near-term correction, hedging broader exposure, or expressing a longer-dated view that AI valuations will normalize over several years. Any claims that tie his position to a precise price target or a specific percentage decline in a semiconductor index are therefore speculative.
Scion’s next 13F, covering the quarter ending December 31, 2025, would show whether Burry increased, reduced, or exited the Nvidia puts. As of now, that subsequent report has not appeared in the SEC’s EDGAR system, making the September 30 snapshot the latest official record of the fund’s disclosed U.S. equity holdings and listed options.
Gaps in the record and what investors should watch
Several pieces of the puzzle are missing. The full text of Burry’s Substack post is not included in any regulatory filing, and only secondary descriptions summarize his comments. That means important details – such as the exact semiconductor index levels he cited, the valuation metrics he emphasized, or any time horizon he attached to his warning – are not available for independent verification.
There is also no public visibility into any over-the-counter derivatives, non-U.S. securities, or cash positions Scion may hold, all of which could materially change the risk profile of the Nvidia puts. A concentrated put position, for example, might be offset by long exposure elsewhere in the capital structure or by stakes in other beneficiaries of AI adoption.
For individual investors, the most practical takeaway is not to mirror Burry’s trade, but to use his move as a prompt to reassess concentration and downside scenarios. Portfolios dominated by a handful of AI and semiconductor names are effectively making a similar bet in reverse – that the current optimism will persist and that short-term volatility will be tolerable.
Watching future 13F disclosures, monitoring how semiconductor valuations evolve relative to earnings, and paying attention to whether AI spending meets lofty expectations will matter more than any single headline. Burry’s Nvidia puts are one data point in a broader debate over whether the AI boom represents a durable shift in corporate profitability or another cycle of overreach that will eventually correct.



