June 3 - The wave of investment into artificial-intelligence infrastructure that has helped drive a record run in technology equities is also playing a visible role in the rise of long-term Treasury yields, market participants say. Firms including Meta Platforms and Oracle, among others, have tapped global debt markets to the tune of $250 billion so far this year, according to Morgan Stanley, borrowing at a scale that would have been difficult to imagine only a few years ago.
Analysts point to the AI-related buildout as a partial explanation for the selloff in May that pushed 30-year Treasury yields to their highest level since 2007. While yields have retreated from those peaks, they remain above where they started the year, and the increase in corporate bond issuance tied to AI projects is one of several forces that helped lift long-term rates.
Corporate financing for long-lived assets
Data center investments often combine components with vastly different economic lives. High-performance AI chips may need replacement every few years, but buildings, land, power connections and other site infrastructure typically last two to three decades. That mismatch makes long-term, fixed-rate financing attractive to companies building such assets because it locks in funding costs for the life of the durable infrastructure.
Some of the largest technology issuers have shifted toward providing duration - that is, long-term interest-rate exposure - in the investment-grade market. Srini Ramaswamy, a senior financial economist at the Federal Reserve Bank of Dallas, notes that Oracle, historically a modest long-term borrower, has become one of the bigger suppliers of duration risk to the market. Ramaswamy estimates that AI-related issuance accounts for roughly 15% of the duration supplied by total Treasury issuance. "That’s a huge number," he said.
Those moves have also been reflected in credit-derivative pricing. The cost of Oracle’s five-year credit default swap climbed to 150 basis points from about 30 basis points over the past year, a sign that investors are factoring in the firm’s sharply increased debt load.
Short paper, long exposure - and the role of swaps
Even as publicly reported long-term borrowing has surged among large technology firms, institutional constraints and lender preferences can mean the headline figures understate the effective increase in long-duration supply. Many institutional investors face internal limits on exposures to any single issuer and on overall portfolio duration. Private lenders frequently favor floating-rate loans.
To obtain long-term fixed-rate funding while accommodating these constraints, borrowers often issue shorter-dated or floating-rate debt and then use interest-rate swaps to convert that exposure into longer-term, fixed-rate financing. Ramaswamy estimates that such swap activity alone generated about $50 billion in 10-year-equivalent supply in the fourth quarter, and that the number is likely higher now.
By comparison, the Treasury sold $540 billion in 10-year notes in the past year. The additional duration coming from corporate issuance and swap-converted liabilities complicates a simple narrative that Treasury yields moved chiefly because of shifts in expectations for Federal Reserve policy or inflation.
Real yields, inflation expectations and capital demand
Movements in real yields - which abstract from inflation expectations - help illustrate the dynamics. Jonathan Hill, head of U.S. inflation market strategy at Barclays, observes that in a typical inflation scare, long-term inflation expectations climb sharply. This episode, he says, has been different: real yields have risen while inflation expectations have remained relatively contained. Hill interprets that pattern as consistent with an AI-driven investment wave that is raising capital demand now, even though its eventual effect could be to boost productivity and ease inflation over time.
Other supply-side pressures are also at work. Higher yields increase government debt-service costs, which in turn require more issuance; more issuance can then put further upward pressure on yields. Hill summarizes the situation by noting the scale of corporate issuance tied to AI is running alongside unusually heavy government borrowing. "The issuance numbers on this are dramatic," he said. "It’s running in the background and it truly is the big story."
Fundraising channels and corporate optionality
Corporations have multiple ways to raise funds. Hyperscalers have been issuing sizable sums to finance data centers, power systems and computing capacity. Long-term financing carries clear benefits when companies are erecting long-lived infrastructure, but some firms also have other options: Alphabet’s $80 billion stock-sale announcement shows that equity offerings remain an alternative path to raise capital.
Thomas Urano of Sage Advisory in Austin framed the magnitude of corporate capital spending associated with AI as comparable, in its market effect, to a federal infrastructure program. "We’re talking $750 billion, $850 billion of capex spending on an annualized basis, and it’s expected to ramp up to close to a trillion next year," he said. "It’s kind of like thinking about a federal stimulus package or some kind of infrastructure spend at a federal level." Those estimates underscore why corporate issuance connected to AI can materially affect the demand-supply balance in the Treasury market.
Where this leaves Treasuries
None of the market participants interviewed says AI borrowing is the single most important determinant of Treasury valuations. Fed policy, inflation data, fiscal deficits and global appetite for safe assets continue to be the primary drivers. Yet the corporate buildout is introducing a significant new source of duration supply at a moment when the market is being asked to absorb heavy government issuance.
Insurance companies and other long-term investors have historically sought duration, and increased corporate supply can appeal to such buyers. At the same time, swaps and the conversion of floating- or short-term issuance into long-term exposure mean the actual amount of duration arriving in the market may exceed what headline issuance figures alone suggest.
The combination of large-scale corporate borrowing for AI infrastructure, active swap markets that extend effective maturities, and a substantial government funding program has therefore contributed to higher long-term yields. How long that pressure persists depends on future corporate financing needs, swap activity, fiscal issuance and developments in inflation and monetary policy.