Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
US companies have sold $1.7tn of investment-grade bonds in 2025, a near-record sum stoked by a rush of borrowing to fund AI infrastructure that has spurred concerns over a debt glut.
This year’s issuance — closing in on the $1.8tn issued in 2020 as businesses rushed to shore up their finances during the Covid pandemic — has come as companies took advantage of relatively low borrowing costs to refinance their debt.
But the debt sales, tracked by trade body Sifma through to the end of November, also reflect an AI borrowing boom, as Big Tech groups including Meta, Alphabet, Amazon and Oracle tapped bond markets to fund data centres and the energy systems needed to power them.
AI-related borrowing now accounts for around 30 per cent of net investment grade issuance, according to Goldman Sachs, and is expected to grow in 2026, despite concerns over the level of debt being taken on by the AI “hyperscalers”.
“This is just the tip of the iceberg,” said Erin Spalsbury, head of US investment-grade bonds at Insight Investment. “It’s very safe to say that we will have more issuance next year. The market is preparing for that.”
Corporate borrowing costs over the summer fell to their lowest level relative to US Treasuries since the end of the last century as trade tensions eased and riskier assets rallied, with additional interest rates paid by top-rated companies falling to as low as 0.74 percentage points over government debt.
They have since increased slightly to just above 0.8 percentage points as investors respond to what has been called a “flood” of AI-related issuance.
Doubts about the mismatch between AI borrowing and current revenues — for example at Oracle’s quarterly earnings, where revenue undershot expectations and data centre spending overshot — have sparked sell-offs in tech sector equity and debt.
Investors say issuance in 2026 is likely to break the Covid record and put further upward pressure on companies’ borrowing costs.
Next year’s debt sales will be driven not just by the AI build-out — though JPMorgan estimates that sector will need to borrow $1.5tn by 2030.
More than $1tn in debt is maturing in each of the next three years, which will require large refinancing deals, according to Dan Mead, head of the investment-grade syndicate at Bank of America. A busy pipeline of M&A deals should also lead to big bond sales to fund takeovers.
“We expect a similarly busy 2026, with the potential for it to be the largest issuance year ever for investment-grade bonds,” said Mead.
Top-rated companies’ borrowing costs over Treasuries could rise by around 0.2-0.3 percentage points next year, estimated Hans Mikkelsen, a US credit strategist at TD Securities, as the extra issuance puts pressure on investors’ appetite for US corporate debt.
Recommended
In a separate note, Mikkelsen wrote that the amount of financing the tech sector needs next year is so big that life insurance companies — large buyers of long-dated bonds — may exceed their internal limits on exposure to a single bond issuer.
Some investors have already begun to hedge against the risk that the AI boom turns into a credit bust, if returns do not come through quickly enough to justify the surge in borrowing.
Trading in credit default swaps, derivatives that pay out if a company defaults, has jumped sharply in recent months for the largest technology issuers. Oracle’s CDS reached its highest level since 2009 earlier this month.
Volumes in single-name CDS tied to a handful of US tech groups have climbed about 90 per cent since early September, according to data from clearing house DTCC.
Additional reporting by Ian Smith