The red-hot market for financing AI data centers is cooling. Banks are 'filled up' with exposure and becoming more selective, while bond yields for these projects have risen from ~6% to 8-9%. This dual-market pressure makes it harder and more expensive for developers to fund new capacity.

Jeffrey Schmid suggests the massive capital investment required for the AI and data center build-out is creating significant new demand for credit. This demand competes directly with public sector borrowing and other commercial needs, which in turn puts upward pressure on bond yields as part of a classic supply-and-demand dynamic for money.

A financial flywheel, reminiscent of the pre-2008 crisis, is fueling the AI data center boom. Demand for yield-generating securities from investors incentivizes the creation of more data center projects, decoupling the financing from the actual viability or profitability of the underlying AI technology.

Data center projects are often built to a target 'yield on cost,' a key real estate metric. With bond yields rising to 8-9%, the cost of financing inflates the project's denominator. This squeezes profit margins to the point where new projects may become unaffordable for borrowers.

AI data center investments are viable because long-term leases cover initial costs, leaving developers with a valuable future asset. Rising rates increase lease costs and decrease the present value of that future asset, threatening the financial viability of the entire AI infrastructure buildout.

Unlike hyperscaler facilities locked into 10- to 15-year tenant agreements, colocation data centers rely on shorter leases across diverse tenants. With vacancy rates near 1%, colocation operators can reprice contracts upward more rapidly, allowing them to absorb elevated borrowing costs.

The massive capital demand for the AI buildout has resulted in debt issuance from hyperscalers and NVIDIA that now rivals U.S. Treasury volumes. This intense competition for capital from what are effectively "nation scale" AI projects is a significant and novel factor putting upward pressure on global interest rates.

Surging borrowing requirements and elevated interest rates are squeezing capital-heavy AI cloud infrastructure providers. When CoreWeave's secondary bond yields jumped to 15% to 16%, credit markets began pricing in significant default risk, serving as a reminder that debt capital for data center buildouts is finite.

Tech leaders argue that the AI buildout is a key driver of rising interest rates. The demand for capital from hyperscalers and data center projects is so immense—borrowing at a 'nation scale'—that it creates a highly attractive alternative to government debt, forcing yields higher to compete for investment.