- Investment-grade companies have sold nearly $1.5 trillion of bonds this year — a 36% jump from a year earlier and on pace to eclipse the 2020 record — with tech giants’ borrowing alone equivalent to roughly 25% of the Treasury’s net note and bond issuance to private investors, five times more than in 2025.
- Bank of America estimates the AI corporate bond surge has pushed up 10-year Treasury yields by about 0.3 percentage points this year, as investors sell Treasuries to buy higher-yielding corporate debt: Alphabet’s recent 30-year bonds priced at ~6.4%, while a Meta data center bond paid over 7.5% — yields that are hard to pass up relative to Treasuries.
- Treasury Secretary Bessent has tried to contain long-end yield pressure by shifting issuance toward short-term bills — trimming net supply of 10- and 30-year bonds by ~$440 billion versus 2025 — but the AI debt wave has more than plugged that gap, with corporate net bond supply expected to swell by $474 billion, much of it long-dated tech issuance.
- This dynamic is still in its early innings: Amazon and Alphabet have both raised spending forecasts, Nvidia is working to raise another $500 billion for AI infrastructure, and JPMorgan estimates total AI infrastructure spending will reach $5.5 trillion through 2030 — meaning the crowding-out pressure on Treasury markets is a structural feature of this cycle, not a temporary spike.
What Happened?
The AI infrastructure buildout has triggered a corporate bond issuance boom of historic proportions. Investment-grade companies have already sold nearly $1.5 trillion of bonds in 2026 — 36% more than the same point last year — putting the market on pace to break the 2020 record. The biggest tech companies’ roughly $200 billion in borrowing alone now equals about 25% of the Treasury’s own net issuance to private investors, up from just 5% of that figure in 2025. The result is a market flooded with long-dated paper competing for the same pool of capital — and bond funds have responded by trimming Treasury allocations and increasing corporate holdings to a three-year high of 30% of portfolios. “AI has been the biggest story in markets for the last couple years,” said Vanguard’s Alex Payne. “It just touches everything.”
Why It Matters?
The conventional “crowding out” theory holds that heavy government borrowing squeezes out private borrowers. What’s emerging now is an inversion: massive private AI borrowing is squeezing out the government, forcing investors to demand higher yields on Treasuries to compensate for the more attractive alternatives in corporate credit. Bank of America estimates the AI bond wave alone has added about 0.3 percentage points to 10-year yields this year. For Bessent, this complicates the task of managing borrowing costs: even as he reduces long-bond supply, AI companies fill the gap and keep duration pressure elevated. “Whoever’s issuing, be it a government or a hyperscaler, is now competing with more borrowers,” said Nuveen’s Tony Rodriguez. “And therefore yields have to be higher.”
What’s Next?
There are no signs of deceleration. Amazon and Alphabet have both raised capex forecasts, Nvidia is actively raising another $500 billion for AI infrastructure, and JPMorgan projects total AI infrastructure investment will reach $5.5 trillion through 2030. As PGIM co-CIO Greg Peters put it: “That is a crowding-out effect. We are just starting. This hyperscaler debt issuance story has really just begun.” For Treasury markets, this means the AI crowding-out dynamic is not a temporary shock but a durable structural feature of this rate cycle — one that keeps long-end yields elevated regardless of Fed or Treasury actions, and could eventually force the Treasury to further reduce its own long-bond auction sizes to avoid being priced out of its own market.
Source: Bloomberg
