What Happened?
As a Senior Analyst on Cushman & Wakefield’s Debt & Structured Finance team, I’ve been tracking a structural shift in how the world’s largest technology companies finance themselves, and what it means for the sovereign and domestic debt markets that underpin our borrowing costs. The AI infrastructure supercycle has ended Big Tech’s “asset-light” era. In 2025, major hyperscalers and AI infrastructure borrowers, including Amazon, Alphabet, Meta, Microsoft and Oracle, issued a combined $121 billion of U.S. corporate bonds, up from an annual average of $28 billion between 2020 and 2024. By 2026, AI-related investment-grade issuance is projected to reach $200-$400 billion, while cumulative hyperscaler bond supply could total up to $3.5 trillion by 2030. For context, a $250 billion midpoint for projected AI-related corporate issuance would equal roughly 12.5% of an estimated $2 trillion in annual net Treasury issuance.
To absorb this massive volume, hyperscalers are issuing highly diversified, multi-tranche bond packages that aggressively target both the belly and the long end of the yield curve, reflecting the dual need to finance short-cycle hardware and multi-decade utility assets. In the belly (5-7 years), hardware-related financing achieves exceptionally tight spreads of +32 to +65 basis points (bps) over comparable U.S. Treasuries. By contrast, long-end infrastructure funding (20-30 years) clears at wider spreads of +75 to +95 bps, with issuers such as Alphabet extending duration even further into 50- to 100-year tranches.
What it Means for CRE
When elevated 10-year Treasury yields reflect duration supply pressures rather than Fed policy alone, the total cost of capital rises across all CRE debt executions – even when credit spreads hold flat. Life insurers and debt funds that price loans off Treasury benchmarks are now operating with a base rate that is more volatile and less anchored than at any point in the post-GFC era. But the more immediate effect I'm watching is lender selectivity. The same institutional fixed-income accounts that buy CMBS, REIT unsecured debt, life company paper and other CRE-linked credit are also being asked to absorb record investment-grade corporate issuance. This does not mechanically crowd CRE debt out of the market, but it does raise the relative-value hurdle for lower-liquidity or headline-exposed CRE credit. When investors can earn attractive spreads in highly rated, liquid corporate paper, CMBS and CRE lenders need stronger collateral, cleaner structures, wider spreads or higher risk-adjusted returns to compete for capital. The effect is most visible in office, transitional and non-stabilized assets, where lenders have less tolerance to stretch on proceeds, basis or structure.
What’s Next?
The convergence of three forces I’m monitoring – record hyperscaler corporate issuance, persistent sovereign deficits requiring continuous bond market access, and a Fed no longer absorbing duration through QE – is creating real structural competition for global fixed-income capital. Foreign official demand for Treasuries also appears less dependable than it was in the post-GFC/QE era, as some reserve managers diversify incrementally away from dollar-duration concentration. The result is not simply “higher rates,” but a higher clearing-rate environment for duration, especially at the belly and long end of the curve. For CRE, the implication is that floating-rate pain is not just a Fed funds story, and fixed-rate executions carry more duration risk than the quoted spread alone suggests. Even if inflation moderates or the Fed cuts, borrowers may still face elevated base rates if Treasury supply, term premium and competing corporate issuance keep pressure on long-duration capital. That is a harder environment to time or hedge than a conventional tightening cycle, and it should be reflected today in underwriting, debt sizing, exit cap assumptions and refinance sensitivity analysis.
