Burry Slams Nvidia's $50B AI Financing: A 2008-Style Wall Street Trick
Nashnova编辑部
Michael Burry — the investor who called the 2008 subprime crisis — publicly labeled Nvidia's $500 billion AI financing plan a '2008-style Wall Street trick.' This means → the debate over hidden risk inside AI's funding boom is now front and center.
What exactly is Nvidia's $500 billion financing plan?
Nvidia teamed up with six Wall Street firms to create a $500 billion+ capital pool dedicated to financing its clients' data-center projects.
The money comes mainly from insurers, pension funds, and other long-term institutional investors. In plain terms = Nvidia isn't writing the checks — it built a platform that channels other people's capital toward buyers of its own products.
Nvidia also pledged up to 25% residual-value support on select projects — meaning if a project loses value, Nvidia absorbs part of the loss — evaluated case by case.
Why does Burry call this a "2008-style trick"?
Burry drew a direct line between this "private credit + asset-backed securities" structure — ABS packages loans into tradeable financial products — and the structured-finance instruments that preceded the 2008 financial crisis.
This means → he believes the layered complexity masks real risk; if AI demand slows, the chain reaction could exceed expectations.
Burry is the real-life figure behind *The Big Short* — he profited by shorting subprime in 2008, and that track record gives his warning outsized market attention.
What do analysts say about this "circular financing"?
UBS analysts had already flagged Nvidia's move as "further raising questions about AI circular financing" — in plain terms = the supplier funds its own customers to buy its own products; the demand numbers look strong, but the money loops inside the system.
Seaport Global analyst Jay Goldberg was blunter: "Someone at some point is going to say, 'Wait, what are we doing?'" He sees Nvidia pledging its own credit behind client debt as a sign the boom is turning irrational.
This reflects a now-public split inside Wall Street over whether the AI investment cycle has overheated.
Is there a counterargument?
Cantor Fitzgerald analyst CJ Muse offered a different read: Nvidia wants clients to use its GPUs instead of competitors' products, and financing support is a tool for building a competitive moat.
This means → seen through a competition lens, this is not "inflating a bubble" but "locking in customers" — an entirely different logic.
What does the market's immediate reaction tell us?
On the day the plan was announced, Nvidia's stock was roughly flat; competitors like Broadcom and hyperscalers such as Alphabet and Amazon fell.
At the same time, alternative-asset managers KKR, Apollo, and Blackstone rose, along with data-center equipment makers and construction-related firms.
In plain terms = money shifted from "the people buying chips" to "the people lending money and building buildings" — capital is betting on the financing chain itself, not on AI demand growth.
What is the unresolved question?
The exact trigger conditions, project scope, and risk-sharing mechanics of the residual-value guarantee remain undisclosed.
Whether Burry's "2008 analogy" holds comes down to one thing: can this financing structure withstand an AI demand slowdown?
This reflects the real core of the current market debate — not whether AI has a future, but whether the way AI is being financed could itself become the source of risk.
Content is for reference only, not financial advice.