A deluge of hyperscaler debt creates risk that investors will run up against concentration limits in their portfolios from a single company or industry
A deluge of AI-related debt this year could be reaching a tipping point.
Investors are no longer finding it easy to digest the torrent of debt funding the artificial-intelligence buildout.
Tech borrowing only kicked off in earnest last fall, when the megacap Big Tech companies known as hyperscalers started to voraciously issue debt. Now, hyperscalers are the largest contributors to the 32% increase in U.S. investment-grade corporate bond supply in 2026 through July 10, according to a new BofA Global report.
"This is a generational shift in a technology," said Tom Murphy, global head of investment-grade credit at Columbia Threadneedle Investments - noting that it has significant implications for both equity and debt markets.
But right now, the magnitude of public and private debt for AI is "just kind of a shock to the system," Murphy said.
Amazon.com (AMZN) has already issued $92 billion of bonds across currencies this year, the most among the hyperscalers. It's followed by Alphabet $(GOOGL)$ $(GOOG)$, Meta Platforms (META) and Oracle $(ORCL)$, as the below chart shows.
Corporate debt issuance this year from Big Tech hyperscalers.
The deluge means investors could risk running up against concentration limits in their portfolios from a single company or industry. Those limits can create selling pressure when new bond deals are announced and investors need to make room. It also can limit how much of any new deal they can buy, which can increase borrowing costs.
That's why Amazon's $25 billion bond deal last week might be a watershed moment for the market. The financing surprised investors because it came before the company reported second-quarter earnings, when plans for borrowing, AI capital expenditures and monetizing tend to be updated. Amazon is slated to report quarterly results on July 30.
Unlike previous Amazon debt deals, last week's didn't breeze through the market without friction. Of note, it followed on the heels of SpaceX's $(SPCX)$ $25 billion post-IPO bond deal in June, which sold off sharply after it priced.
That sort of weakness, if it persists, could put pressure on the rest of the U.S. corporate bond market, where borrowers have benefited from spreads near record lows, according to Murphy. But those days may be numbered.
"There's been hand-to-hand combat around AI," Murphy said, noting that most hyperscalers still put up robust earnings and have high credit ratings and low leverage. But if weakness in their bonds continue, "the rest of the market has to reprice wider."
Wider "spreads" mean investors are getting extra compensation in addition to a benchmark rate, which is often the yield on low-risk 10-year Treasury note BX:TMUBMUSD10Y. That can help compensate investors for increased market jitters or credit risks, but it also can bump up borrowing costs for companies.
Corporate bond spreads have been exceptionally low in recent years, in part because Treasury yields have held at high levels relative to several years ago. The 10-year yield was up 5 basis points to 4.62% on Monday, near a one-year high for the yield, according to FactSet. In 2020, during the height of the COVID-19 pandemic, yields were below 1%.
U.S. stocks DJIA SPX COMP, bonds and gold (GC00) were all selling off Monday on renewed tensions over the Strait of Hormuz in the U.S.-Iran conflict and a sharp increase in global oil prices (BRN00) to about $83 a barrel.
Neil Sun, a portfolio manager at RBC Global Asset Management, said his team has been looking recently for opportunities in investment-grade corporate bonds, given the prospect of an AI supply overhang and wider spreads.
Overall, AI-related corporate debt supply has already jumped more than 99% from a year ago to $270 billion, as the below chart shows.
BofA and Goldman Sachs both expect $2.1 trillion of total U.S. investment-grade bond supply this year, which would rival 2020 pandemic peak levels.
With the benchmark Bloomberg U.S. Corporate Bond Index around $8 trillion in size, the corporate debt sector is expected to finance the bulk of the AI financing need. Leverage loans, private credit and the asset-backed markets also have been smaller parts of the financing pie.
Major U.S. banks, which kick off their earnings on Tuesday, have exposure to AI in their capital markets and lending activities. They tend to borrow in the bond market right after reporting earnings.
"From an investor perspective, you need to understand your true exposure to AI," RBC Global's Sun said.
-Joy Wiltermuth
This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.
(END) Dow Jones Newswires
July 13, 2026 15:30 ET (19:30 GMT)
Copyright (c) 2026 Dow Jones & Company, Inc.
Comments