CoreWeave Bond Yield Surges Past 16% as AI Infrastructure Debt Stress Mounts

Credit markets are sending an increasingly sharp warning to AI data center builders, as the yield on CoreWeave's 2032 bond — which carries a 9.625% coupon — climbed above 16%, reflecting deepening investor concern about the cost of financing the AI infrastructure boom.

CoreWeave's five-year credit default swaps spiked above 1,289 basis points on Thursday before easing roughly 8% on Friday to around 1,155, according to Deutsche Bank — still above their October peak and a level that underscores the stress accumulating in AI-linked debt markets.

The anxiety is spreading beyond neocloud operators to companies with far stronger balance sheets. CDS on Nvidia debt rose 76% in the last month, while Broadcom's climbed 49% and Amazon's gained 44%. On an absolute basis, however, the spreads remain far apart: Nvidia's stood at roughly 139 basis points, compared with approximately 298 for Oracle and around 1,100 for CoreWeave.

"We've had plenty of warnings from the credit space," said Jim Reid, Deutsche Bank's global head of macro and thematic research, in a note on Friday, adding that debt and equity markets are now moving more in sync.

The analogy framing Wall Street's concern comes from Jefferies global head of equity strategy Christopher Wood, who wrote in a note Friday that "the most plausible way the AI capex arms race will end is that the credit card will be taken away."

That moment has not arrived. Goldman Sachs tallies five AI bond deals totaling $67 billion since October 29, pushing total hyperscaler debt issuance this year to $296 billion. Goldman forecasts $440 billion of hyperscaler debt issuance next year, plus an additional $300 billion of project-level issuance across investment-grade and high-yield markets.

Still, the political landscape has shifted materially. Democrats flipped both chambers of Congress and won key gubernatorial races, including in Ohio, Texas, and Georgia — three states that together hold much of the planned U.S. data center capacity. Goldman Sachs estimates that potential moratoriums in those states could affect 45% of planned U.S. capacity.

Texas governor-elect Gina Hinojosa has said she is not opposed to data centers but wants a moratorium until the state legislature enacts community protections, adding a layer of regulatory uncertainty that credit markets are now pricing in.

National Economic Council Director Kevin Hassett said Friday that he expects data center growth to continue as forecast, despite the election results — signaling the outgoing Trump administration's intent to support the buildout in its remaining time.

The underlying financial pressure is compounding the political headwinds. Goldman analysts noted that hyperscalers collectively reported negative free cash flow in the third quarter, as capital expenditures exceeded operating cash flow. "Delayed monetization and growth in purchase commitments — which we view as additional and sizable uses of cash, separate from Capex — are the key risks to more debt issuance, subject to market conditions," the bank said.

Bank of America analysts project AI-related credit issuance will rise 31% from 2026 to 2027, reaching $685 billion across public and private debt, before falling 23% to $531 billion in 2028. The bank flagged explicitly that a slowdown in AI infrastructure projects stemming from the election outcome represents the primary downside risk to that forecast.

For the neoclouds specifically, any sign of retrenchment by hyperscalers in their capital spending plans could translate directly into fewer or shorter compute contracts and delays in prepayments — a feedback loop that would further pressure already-stretched balance sheets.

Whether the AI debt cycle has reached its inflection point, or whether the market is repricing a temporary political shock, will likely become clearer as incoming Democratic governors detail their regulatory intentions in the weeks ahead.

Disclaimer