Nvidia’s financial backing has helped AI cloud projects get off the ground. That support could also determine whether they can keep raising money. Apollo Global Management CEO Marc Rowan told CNBC on November 17 that many specialist AI cloud companies could lose access to capital if Nvidia withdraws its financing or guarantees.
Speaking on Squawk on the Street, Rowan described Nvidia as more than a supplier to these businesses, known as neoclouds.
The scale of that role is substantial. CNBC reported that Nvidia’s July quarterly filing disclosed $36 billion in off-balance-sheet AI cloud agreements, or backstops, for neoclouds. Those commitments provide context for Rowan’s concern: the chipmaker’s financial involvement extends beyond selling the equipment used in AI infrastructure.
If that banking function, that guarantor function, gets taken away, access to capital for a large number of neoclouds is going to disappear.
Marc Rowan, Apollo Global Management CEO, on CNBC’s Squawk on the Street
Debt markets show the pressure
Rowan’s warning comes as investors scrutinize the debt used to build AI data centers. CoreWeave and Nebius, which use financing for that construction, had each lost more than 50% of their share value over the preceding three months, CNBC reported. Concerns about neocloud borrowing have also pushed bond yields higher.
CoreWeave’s 9.75% bond due in 2031 was yielding 15.66% on November 17. The two percentages describe different things: 9.75% is the bond’s stated interest rate, while 15.66% was its market yield. Bond yields rise when prices fall, so the higher yield reflects pressure on the price investors will pay for that debt.
Another measure pointed to concern about repayment. Spreads on CoreWeave’s five-year credit default swaps were around 1,113.5. These contracts let investors pay for protection if a borrower cannot repay its debt. Wider spreads signal greater perceived risk; CNBC noted that readings above 1,000 are generally considered a sign of distress.
Rowan sees room for a different financing model
Rowan said Nvidia had a strong business incentive to help projects get started. But he also expects new ways of financing AI infrastructure to emerge. He compared the opportunity with the automobile industry, where the arrival of cars was followed by financing businesses such as GMAC and Ford Motor Credit.
For AI, Rowan said he would not be surprised to see a more finance-company-like approach develop, including a role in judging creditworthiness.
If Nvidia does pull back, Rowan sees the advantage going to companies with the financial strength and locked-in funding to build their infrastructure.
Reader comments
Newest comments first. Replies stay oldest first.