
Newsletter
Six stories and one chart, in your inbox by 6:30 a.m. Eastern. Written by the desk editors, not a machine.

The massive borrowing spree fueling the artificial intelligence boom is beginning to test the appetite of bond investors, with major buyers demanding higher yields as technology companies flood the market with debt.
Investors remain broadly comfortable with the financial strength of companies such as Amazon and Alphabet, but the sheer volume of new borrowing is forcing companies to offer increasingly attractive terms to get deals sold.
Neil Sutherland, head of U.S. fixed income at Schroders, said signs of "indigestion" are already emerging in technology credit markets.
The concern is not necessarily that major technology companies are becoming financially unstable. Instead, investors are questioning how much additional debt the market can absorb as spending on AI infrastructure continues to accelerate.
Corporate bond spreads measure the additional yield investors demand compared with U.S. Treasury bonds. When spreads widen, investors are effectively demanding more compensation for taking on corporate debt.
Technology bond spreads have now reached 89 basis points, about 9 basis points above the broader investment-grade market, according to Karen Choi, a portfolio manager at Capital Group.
That marks a significant shift for a sector that has historically enjoyed some of the tightest spreads in corporate credit because of strong balance sheets and relatively low borrowing needs.
Amazon's recent $25 billion long-term bond sale offered investors roughly 120 basis points above comparable Treasuries, analysts said. That premium was about twice what the company would have paid for similar debt a year earlier.
"Tech has gone from trading materially through the market to actually trading wider than the market," Sutherland said.
The change reflects growing investor sensitivity to the enormous funding requirements of the AI race.
The scale of new borrowing has increased dramatically.
According to BNP Paribas data through August 10, AI hyperscalers have issued approximately $220 billion in debt so far in 2026, compared with just $12.5 billion during the comparable period last year.
That represents an increase of more than $200 billion in a single year.
Alphabet's bond offering earlier this month was reportedly well received, but still required an estimated 10- to 15-basis-point concession compared with the company's existing bonds.
George Catrambone, head of fixed income for the Americas at DWS, said investors are beginning to demand larger concessions as issuance reaches record levels.
Deals that were easily absorbed earlier in the year are now requiring higher yields to attract sufficient demand, suggesting that some traditional bond investors are becoming more cautious.
"The issuance in January versus August looks different," Catrambone said, adding that signs of fatigue are beginning to emerge.


The biggest concern may not be whether investors want technology debt at all, but how much of it institutional portfolios can realistically absorb.
Pension funds and insurance companies often limit exposure to individual issuers to roughly 2% to 3% of assets, according to Choi.
That creates a potential problem as the same group of major AI companies repeatedly returns to the bond market to finance massive infrastructure investments.
If borrowing continues at the current pace, some institutional investors could eventually reach their internal exposure limits even if they remain confident in the underlying companies.
Foreign investors, pension funds and insurers have so far helped absorb the surge in investment-grade debt. The broader investment-grade corporate bond index currently yields around 5.4%, roughly in line with long-term averages, helping maintain demand.
But investors warn that the current appetite should not be mistaken for an unlimited supply of capital.
For now, the situation is not considered a major credit crisis. Major AI companies continue to have strong ratings, substantial cash flows and enormous balance sheets.
The issue is increasingly about price and capacity rather than credit quality.
As technology companies issue more debt, investors are demanding higher yields and larger concessions. That could eventually increase borrowing costs and force companies to reconsider the pace at which they tap debt markets.
The surge in AI-related borrowing has also contributed to upward pressure on Treasury yields, as investors seek greater returns to absorb the growing supply of bonds.
If technology companies eventually slow their borrowing, that could reduce pressure on the long end of the Treasury market.
But if the AI spending race continues unchecked, investors are warning that financing conditions could become increasingly expensive.
"It's not a blank check," Catrambone said. "If these companies keep tapping the market over and over again, concessions are going to get larger and spreads are going to get wider."
The message from bond investors is becoming increasingly clear: Wall Street still believes in Big Tech, but financing the AI boom is no longer coming without a price.
