I was reading Michael Burry’s latest Substack post when the market chart blinked red across my screen. You could feel the air tighten: a man who called 2008 is now shortening the clock on AI’s boom. I want you to hear what he’s doing and why it matters to any position you hold.
Traders watching tickers noticed Nvidia’s gains hit new highs this week.
I’ve followed Burry since The Big Short made him a household name, and his timing is the story here. He says the “bubble in AI may burst sooner than later” and has shifted short positions into near-term put options that expire within a year. That change in timing is not a shrug — it’s a deadline.
You should understand why: Burry is betting that stocks tied to AI — from Nvidia to Palantir and Oracle and the Nasdaq 100 index that houses the hyperscalers — will slide from current peaks before his options run out. He told readers on Substack and signaled the moves again on X, and Business Insider picked up the details.
Will the AI bubble burst soon?
Yes, he believes it could happen within the next year. I don’t say that to alarm you, but to frame the trade: when a veteran who foresaw a systemic collapse starts shortening horizons, probability shifts from theoretical to tactical.
Accountants at cloud giants noticed cash flow strain this quarter.
Google reported negative quarterly free cash flow — a first for the company as a public entity — largely driven by massive AI spending. When hyperscalers ramp capex and R&D at scale, their balance sheets can look healthy in revenue terms while cash actually drains.
If you follow the cash, you see the stress points. That spending is meant to purchase compute and chips, much of it supplied by Nvidia, which now sits at the center of an intricate financing loop. I’ll call it what it is: a structure that moves enormous sums around to create and sustain demand.
Why is Michael Burry betting against AI?
Burry sees three red flags. First, capex growth that mirrors the late-1990s tech bubble. Second, increasingly circular dealmaking where suppliers, customers, and financiers effectively prop one another up. Third, growing reliance on debt and leverage across the system. If spending slows, that web could unravel.
Analysts on desks noticed Nvidia’s new share buyback announcement made headlines.
Nvidia announced an additional $150 billion (€138 billion) in share buybacks — a jaw-dropping authorization it calls the largest increase in history. Jensen Huang framed it as confidence in long-term opportunity, and buybacks usually signal that a company sees value above current market prices.
Yet Burry has put options on Nvidia that expire next September. He’s adding leverage to a short view even as the company repurchases stock and posts strong results. I’d describe this tension as a sentinel warning: corporate signals of strength versus investor bets that the larger market dynamic will change.
How could Nvidia be affected if AI spending slows?
If hyperscalers trim commitments, demand for accelerators would drop quickly. Nvidia is the primary supplier for training and inference infrastructure; a meaningful slowdown would compress future revenue expectations and reverberate through its customers and partners.
Portfolio managers noticed funding flows become circular and concentrated.
Those circular flows — dollars moved among vendors, customers, and financiers — can skew demand and concentrate risk. I’ve seen systems like this before where paper demand masks fragility.
If breakthroughs slow or a major project fails to deliver, capital that looks committed can evaporate, and positions funded by debt will force re-pricing. Burry warns that a pause in spending could make “it all come apart.”
I’ll offer one blunt observation: when a market’s tail starts wagging the dog — suppliers financing their own demand, lenders inflating expectations — you’re closer to a correction than a continuous ramp. Think of it like a house of cards: one misplaced piece and the stack collapses. At the same time, the industry’s dealmaking resembles a spiderweb, drawing everything toward a single node.
I’ve told you what Burry is doing and why a short-term timeline matters. You should watch option expiries, capex reports from hyperscalers, Nvidia’s buyback execution, and cash flow statements from Google and peers. Tools and platforms that matter here include Substack for his notes, X for his earlier posts, Business Insider for coverage, and earnings portals for realtime cash-flow reads.
So what will you do with this window of risk — hedge now, hold through, or bet with Burry’s timeline?