CNBC's Jim Cramer warned Monday that Nvidia's potential $250 billion backstop for OpenAI's Ohio data center project revives the same circular financing patterns that unraveled during the dot-com crash. The warning carries weight for anyone tracking how long the AI infrastructure boom can sustain itself on customer debt that depends on continued access to capital.
"I lived through 2000," the "Mad Money" host said. "I don't want the sequel."
The deal that sparked the warning
The Wall Street Journal reported Sunday that Nvidia was discussing a $250 billion guarantee to help finance a planned 10-gigawatt AI data center campus in Ohio for OpenAI. CNBC confirmed the report Monday. The proposed backstop would support the project's lease and construction debt - not the Nvidia chips deployed inside the facility. Nvidia declined to comment. Shares of Nvidia fell more than 4% on Monday, pulling many semiconductor stocks down with it.
Circular financing and the dot-com parallel
The discussions are the latest example of the increasingly circular nature of AI financing. Nvidia has invested in several companies that are also major customers for its chips, including a $30 billion investment in OpenAI in March and a $10 billion investment in Anthropic last year. The chipmaker has also backed multiple neocloud providers that rent Nvidia-powered computing capacity to customers. Nvidia has said those investments support the growth of the AI ecosystem while offering attractive long-term returns.
Cramer drew a direct line to the late 1990s, when telecom equipment makers helped customers finance major purchases to fuel growth. Those deals initially boosted sales, but many unraveled when cash-strapped buyers could no longer pay, inflicting heavy losses on suppliers and investors alike.
"What we learned in 2000 is that you don't lend to companies who buy your goods," Cramer said.
What happens if the buyer can't pay
Cramer stressed that he still views Nvidia as an exceptionally strong company and is not predicting a repeat of the dot-com crash. Rather, he said history shows investors can quickly lose confidence when suppliers become too reliant on customers whose massive spending depends on continued access to capital.
"If the buyer, in this case, OpenAI, can actually afford to pay for these chips, perhaps because it comes public ... then Nvidia's in terrific shape," Cramer said. "If the buyer can't pay, well, that's a different story."
OpenAI confidentially filed for an initial public offering in June but has not announced a timeline for its debut. The company was valued at more than $800 billion by private investors in March as it races to expand computing infrastructure while competing with Alphabet and Meta.
Cramer said the risks extend well beyond Nvidia because a growing number of companies now depend on continued investment in AI infrastructure. "There are so many companies counting on the data center for their earnings," he said. "If the market decides it doesn't want to fund any more data centers, and the companies themselves don't have the money, or they don't get paid, then we're back in 2000."
He added that strong balance sheets alone have not always been enough to shield companies from the fallout when customers become overextended. "Nvidia shouldn't make these guarantees even if it has all the money in the world. Just history, that's all, just history," he said.
Why this matters for finance professionals
The warning highlights a concentration risk that doesn't show up on a standard balance sheet. When a supplier backstops its own customer's debt to drive revenue growth, the credit exposure loops back to the supplier - and to every investor holding that supplier's stock. Finance professionals evaluating AI-sector positions should examine not just a company's direct debt load but the off-balance-sheet guarantees and equity investments tied to its largest customers. Cramer's point is not that Nvidia is weak. It's that the financing structure itself echoes a pattern that ended badly the last time it scaled this far.
Your membership also unlocks:
