Kevin Warsh said higher borrowing costs can reach AI companies in places Wall Street may be underestimating. His Kevin Warsh AI stock market comments came after the October FOMC meeting, where he and his Federal Open Market Committee colleagues held rates steady on Oct. 28.
Warsh put it in six words: “interest rates get in the cracks.” The warning landed against a market backdrop in which AI stocks have helped drive the Dow Jones Industrial Average, S&P 500 and Nasdaq Composite higher over the last four years, even as some AI hardware expansion is being financed in part with debt.
Oct. 28 FOMC and Warsh
On Oct. 28, Warsh and his Federal Open Market Committee colleagues voted to hold interest rates steady. He followed that decision with remarks that pointed to the pressure higher rates can put on financing, not just on headline valuations. That is the part investors have to watch, because the AI story has not only been about demand for chips and software; it has also been about capital spending.
AI hardware expansion is being financed in part with debt, and that leaves borrowing costs directly in the path of future build-outs. If financing gets pricier, the pace of data-center expansion can feel the strain before the market narrative does.
CME Group rate pricing
The bond and rate market is already leaning the other way. The CME Group’s FedWatch Tool shows a 72% chance of a 25 basis point rate hike in December, with a nearly 50% chance of 75 basis points in cumulative hikes by Sept. 15, 2027.
That pricing matters because inflation has stayed above the Fed’s long-term 2% target for 67 consecutive months. The Fed and the FOMC are expected to raise interest rates several times beginning in December, so Warsh’s warning arrives before the next move is even made.
Debt, yields and AI
The broader debt picture is even heavier. U.S. debt stands at $40.1 trillion, and the last time U.S. Treasury yields were this high, total U.S. national debt stood at $8.9 trillion. Every 1 percentage point increase in the average cost of servicing the debt now translates to about $401 billion per year in interest expense, compared with about $89 billion in 2007.
That gap leaves an additional about $312 billion in annual interest expense for every percentage point increase in borrowing cost. For AI companies leaning on debt to fund expansion, the message is plain: the rally can keep running, but rates still get in the cracks.
