AI hyperscalers will drive U.S. corporate bond supply growth in 2026, analysts say

WASHINGTON, Jan 15 (Reuters) – Analysts expect U.S. corporate bond issuance to rise significantly in 2026, driven in part by the need for AI hyperscalers to ramp up.

While pent-up M&A activity and companies’ need to refinance existing debt are likely to contribute to an increase in overall corporate bond issuance this year, the biggest driver is AI-related funding demand, according to a report released Thursday by Barclays.

Overall U.S. corporate bond issuance is expected to reach $2.46 trillion in 2026, an 11.8% increase from $2.2 trillion in 2025, according to analysts at Barclays. It expects net issuance to be $945 billion this year, an increase of 30.2% from last year’s $726 billion.

“The net supply increase is primarily non-financial, with the biggest upside risk being capital expenditures for AI hyperscalers, which may require a larger-than-usual public transaction,” Barclays analysts wrote.

AI companies are rapidly increasing spending and borrowing as they race to expand their data center presence and processor needs.

5 Leading AI Hyperscalers – Amazon (AMZN.O)opens a new tabAlphabet (GOOGL.O)opens a new tab Google, Meta (META.O)opens a new tabMicrosoft (MSFT.O)opens a new tab and Oracle (ORCL.N)opens a new tab U.S. corporate bond issuance totaled $121 billion last year, with average annual issuance of $28 billion from 2020 to 2024, according to a Jan. 9 report from BofA Securities.

BofA analysts similarly expect hyperscaler borrowing to accelerate this year. The five largest hyperscalers are expected to borrow about $140 billion a year over the next three years, which could exceed $300 billion a year, analysts wrote.

The expected increase will bring the five largest hyperscalers in line with the six largest banks’ projected average annual issuance of $157 billion, according to BofA.

Analysts at BofA said: “As the supply of AI funds increases, the five hyperscalers could become the largest issuers of the IG index.”

A December report by MUFG analysts found that hyperscalers accounted for four of the five largest U.S. high-grade bond trades in 2025. Most of them occurred in the second half of this year.

Oracle sold $18 billion in bonds in September. This was followed in October by Meta’s $30 billion deal (the largest single non-M&A sale of high-quality debt in history), and in November by deals from Alphabet ($17.5 billion) and Amazon ($15 billion).

A surge in hyperscalar borrowing has widened credit spreads, and investors are increasingly turning to credit default swaps (CDS) to hedge against AI-related downside risks.

‍According to MUFG, the cost of insuring hyperscalers’ debt through CDS has been rising since June 2025, particularly in the case of Oracle, where five-year CDS have more than tripled since the September deal.

Bondholders sued Oracle on Wednesday, alleging they suffered losses after the company, chaired by billionaire Larry Ellison, failed to disclose that it would need to sell significant amounts of additional debt to build out its AI infrastructure.

Reporting by Matt Tracy in Washington. Editing: Lisa Shoemaker

Our standards: Thomson Reuters Trust Principles.opens a new tab

Purchase license rights

Reports on U.S. credit market activity, including corporate bonds and credit ratings, U.S. Treasuries, commercial mortgages, and ongoing public vs. private financing movements. He frequently reports on other topics and events in collaboration with team members across Reuters. Matt has previously covered regulatory reviews and investigations related to mergers and acquisitions, particularly antitrust, national security, the FCC, and state-level investigations into some of the largest transactions since 2016. He broke the news on government investigations into the AT&T-Time Warner merger, T-Mobile’s acquisition of Sprint, the Bayer-Monsanto merger, and many other multibillion-dollar combinations.

Source link