Data centre bond investors pulled back hard this week, with the Financial Times describing a 'Darwin moment' for AI borrowers as credit markets reprice risk. The NASDAQ 100 fell 0.83% to 29,702 while the VIX climbed 3.56% to 17.47, and the US 10-year Treasury yield sits at 5.24%, raising the cost of capital for leveraged AI infrastructure names just as doubts over spending intensify.
Data Centre Bonds Face a 'Darwin Moment'
The Financial Times reports that investors are exiting data centre bonds as scrutiny of AI-linked borrowers sharpens. A separate FT piece on CoreWeave's bonds describes signs of financial distress in the debt of one of the sector's highest-profile borrowers. Neither report details specific spread levels, but the direction is clear: credit investors who financed the AI build-out are now demanding a higher premium for the risk.
This matters because the AI infrastructure boom has been funded substantially with debt, not just equity. With the US 10-year yield at 5.24%, the base cost of long-duration capital is already elevated before any company-specific risk premium is added. Borrowers that locked in financing assumptions during an earlier, lower-rate period now face refinancing math that looks far less favourable.
Tech Stocks Slide as AI Spending Questioned
The selloff in credit has a mirror image in equities. The New York Times reported that chip and technology stocks plunged as doubts over the scale and payback period of AI spending intensified. The NASDAQ 100's 0.83% decline stands in contrast to the Dow's 0.76% gain, a split that suggests capital is rotating out of concentrated AI exposure and into broader, less leveraged names.
A bond market that once treated AI infrastructure debt as a near-certain growth story is now asking the same question equity analysts have asked for months: how long can spending outpace revenue before credit quality becomes the binding constraint.
The pullback is not confined to equities and bonds. Bitcoin fell 0.43% to $81,500 and Ethereum dropped 1.49% to $2,385, both easing in step with the risk-off tone in tech. That correlation is worth tracking in its own right, and readers following how digital asset prices move with broader macro stress can find ongoing coverage at Coin Hours.
Central Banks, Yields and the Rate Outlook for AI Credit
The rate backdrop is central to this story. The Fed Funds Rate sits at 3.75%, yet the US 10-year yield at 5.24% shows a term premium that keeps long-duration borrowing costs high regardless of where the Fed sets its overnight target. For AI borrowers who financed data centres with long-maturity debt, this gap between policy rate and market yield is the real constraint. Rate cuts at the short end do little to help if the long end stays anchored near 5%.
Global central banks are not moving together. The RBA cash rate holds at 4.6%, the ECB at 2.5%, and the Bank of Japan at just 0.98%. This divergence shapes currency and capital flows as much as it shapes borrowing costs. Meanwhile, the BBC reported the UK economy grew just 0.1% in August ahead of Chancellor Healey's first Budget, a soft print that contrasts with the FTSE 100's 0.9% gain, underscoring how equity markets and underlying growth data can tell different stories in the same economy.
For AI-linked credit specifically, the message from the Fed Funds Rate versus the 10-year yield is that financing conditions will stay tight even if the central bank holds steady. Markets are not pricing imminent relief at the long end, and that keeps pressure on any borrower whose business case depended on cheap, abundant debt.
Background: Commodities and the Search for Ballast
As credit and tech equities wobble, some capital has rotated toward hard assets. Gold USD trades at $4,160, up 1.1%, and silver sits at $57.10, up 2.19%, both benefiting from demand for assets outside the leveraged tech-credit complex. Readers wanting a grounding in how raw materials function as investable assets, including the mechanics of futures pricing and the structural demand drivers behind cycles like this one, can review Commodities Explained — How Raw Materials Become Investment Assets. For readers tracing how modern central banking got to where it is today, including the shift toward inflation targeting that underpins current rate settings, The History of Central Banking — Part 2 offers useful context.
None of this means the AI investment cycle is over. It means the financing side of that cycle is being tested in a way the spending headlines have not yet forced investors to confront. Credit markets, not earnings calls, are where that test is currently playing out.
Investors holding AI-adjacent credit or equity exposure should watch the 10-year yield and data centre bond spreads together, since a sustained widening alongside yields above 5% would signal the Darwin moment the Financial Times describes is just beginning.