Nvidia wants AI chips to become fungible asset, but Wall Street isn’t convinced yet | Here's why

Nvidia wants its artificial intelligence chips to become something more than hardware powering the AI boom. CEO Jensen Huang wants “compute” — critical processing power provided by Nvidia’s GPUs — to become a productive, durable and fungible asset that can support long-term financing’.

Wall Street, however, is taking a more cautious view and said they have some doubts about whether the chips can act as long-term collateral, as Nvidia says

The debate has become important as Nvidia pursues a mammoth financing initiative to help AI developers access its compute through chip-backed loans. The plan depends on lenders being comfortable treating Nvidia’s GPUs as valuable collateral for years to come.

Why does the value of AI chips matter?

When a bank lends against an asset, it needs to estimate the asset's value if the borrower cannot repay the loan.

Nvidia argues that its most advanced GPUs can continue generating revenue for as long as a decade. But some banks and credit investors are underwriting the chips over just three to four years, according to people familiar with the matter cited by Reuters. That gap is crucial.

If lenders believe a GPU will lose much of its economic value after three or four years, they are likely to demand greater protection before lending against it. That means investors are likely to seek higher interest rates, bigger financial cushions and stronger repayment protections before backing loans secured by AI chips.

“Wall Street is much more conservative,” Tony Trzcinka, a senior portfolio manager at Impax Asset Management, told Reuters.

Why are banks cautious about GPUs?

One reason is the lack of historical data.

GPUs are evolving rapidly, with newer and more powerful generations regularly entering the market. That makes it difficult for lenders to determine how much an older chip will be worth several years from now.

"Nvidia would imply that the GPUs work well north of five years, and that actually has been proven to be true thus far," said Andrew Chang, a director at S&P Global Ratings. Yet "we take a conservative view of the value of those chips," he said.

Addressing the concern, Nvidia pointed to several third-party studies that showed major cloud companies extending the depreciation periods for servers up to five to six years, from three to four. It also cited a finding from Barkr, a firm that provides valuations for AI collateral such as GPUs, indicating that the company's most recent GB300 NVL72 systems could have a useful life of 9 to 10 years.

What does Nvidia’s financing plan involve?

Nvidia announced the $500 billion initiative in August with financial partners including Blackstone, Apollo and KKR. The idea is to use Nvidia chips as collateral so AI developers can access financing to build out computing capacity.

Initially, Nvidia indicated that some transactions could have residual-value guarantees of no more than 25%. But bankers and asset managers now expect some deals to carry stronger guarantees or be supported by customer contracts and Nvidia guarantees.

That would give lenders greater certainty over debt repayment.

What do existing deals show?

Recent GPU-backed financing deals have generally relied on strong customer revenues to reduce the risk for lenders.

CoreWeave, in which Nvidia has a stake, secured an $8.5 billion investment-grade GPU-backed loan, with lenders relying heavily on contractual payments from Meta.

Broadcom, meanwhile, backstopped more than 80% of a $35 billion financing structure linked to AI computing capacity for Anthropic.

These deals indicate what Wall Street may demand as Nvidia attempts to build a broader market for chip-backed financing.

(With inputs from Reuters)