AI Boom Under Pressure: Rising Borrowing Costs Signal Funding Risks
A Jefferies report warns of rising credit risks and borrowing costs for AI infrastructure giants like Oracle and CoreWeave.
The meteoric rise of artificial intelligence is hitting a potential financial speed bump. According to a recent research report by Jefferies, financial markets are exhibiting signs of stress regarding the funding of the global AI expansion, with borrowing risks climbing for the very companies tasked with building the necessary infrastructure.
Red Flags in Credit Markets
In a detailed research piece titled "K-shaped stresses, AI tremors and the Houthis," Christopher Wood of Jefferies warns that investors must keep a "laser focus" on credit spreads. These spreads serve as critical indicators of the risk associated with lending; when they rise, it means investors are demanding higher compensation to offset the increased risk of lending to a company.
The report points to a worrying trend among key players in the AI space. The data reveals a sharp spike in credit default swap (CDS) spreads—essentially insurance against a company defaulting on its debt—between September 11 and the date of the report:
- Oracle: The five-year CDS spread jumped from 186 basis points to 288 basis points.
- CoreWeave: The spread surged dramatically from 749 basis points to 1,118 basis points.
These figures represent the highest levels recorded since the AI capital expenditure (capex) cycle began in early 2023. Furthermore, the report notes that the yield on CoreWeave's bond maturing in 2032 has hit 14.8 per cent, highlighting the steep financing costs the company is now facing.
The Infrastructure Challenge
The stakes are incredibly high because the expansion of AI is not just about software; it requires massive physical infrastructure. This includes the construction of sprawling data centres and the acquisition of immense computing capacity. For companies that rely heavily on external funding to fuel this growth, rising borrowing costs could make these essential projects significantly more difficult and expensive to execute.
"Investors should closely monitor borrowing conditions as companies continue to spend heavily on AI infrastructure."
Slowing Growth and Valuation Gaps
Beyond the immediate cost of borrowing, Jefferies is highlighting a potential cooling-off period for earnings. Referencing another report titled "Diversifying Beyond AI," the firm suggests that the explosive earnings growth currently seen in AI stocks is likely to moderate.
The consensus projections for global AI stocks show a steep decline in growth rates:
- 2026: Expected growth around 100 per cent.
- 2027: Expected to drop to 30 per cent.
- 2028: Expected to further slow to 22 per cent.
This projected slowdown is particularly concerning given the current valuation of these companies. Global AI stocks are currently trading at 23 times forward earnings, a stark contrast to the 13 times valuation for non-AI stocks. This significant gap puts immense pressure on AI firms to justify their massive spending and high valuations as growth begins to level off.
An Early Warning, Not a Stop Sign
It is important to note that the report does not claim that funding for AI has ceased. Instead, it frames these rising credit risks as an early warning system. While the investment in AI remains substantial, the combination of expensive debt and slowing earnings growth may force companies to be more disciplined and strategic with their capital expenditures moving forward.