Stay informed with free updates
Simply sign up to the US companies myFT Digest — delivered directly to your inbox.
Investors are selling out of long-dated AI debt, amid growing fatigue over Big Tech firms’ huge borrowing spree and concerns over whether the multitrillion dollars of infrastructure investment it is funding will ever pay off.
Prices of AI-linked bonds with maturities of 10 years or more have fallen this week, making them some of the worst performers in the investment-grade market, according to MarketAxess data. The yield on a 30-year SpaceX bond has risen to 7.3 per cent, up from 6.7 per cent when it was sold less than two weeks ago.
Bonds issued by the top five hyperscalers — Amazon, Google, Meta, Microsoft and Oracle — are now yielding roughly 0.6 percentage points more than their blue-chip peers with the same ratings and maturities, representing the widest risk premium of any sector in the investment-grade market, according to BofA Global Research.
While the short-term borrowing costs for so-called hyperscalers building vast AI infrastructure remain steady, investors have been demanding more yield to compensate for the risk of holding debt with longer tenors. They point to risks that AI’s eventual profitability may not match current lofty expectations, while future innovation in the fast-moving sector could render this investment obsolete.
“We prefer taking more near-term risks,” said Mariya Entina, portfolio manager at fund firm DoubleLine. “Typically when you buy 30-year bonds you want businesses that have a really solid outlook, such as clear returns on investments. There’s scepticism over the longer-term profitability of AI capital expenditure.”
Pramod Atluri, a portfolio manager at Capital Group, also favours short-dated hyperscaler bonds.
“The technology is evolving so rapidly, which makes it a riskier proposition to lend for a long period of time,” said Atluri. “It’s not clear what the industry landscape is going to look like ten years from now.”
The latest sell-off follows a glut of record-breaking issuance by tech companies to fund the AI arms race. Supply of AI-related high-grade bonds has reached $270bn across currencies this year, almost double the amount for the whole of last year, according to BofA Global Research.
Amazon, which sold $25bn of bonds on Tuesday, saw weaker demand for the longer end of the deal. Its five-year bonds attracted about 20 per cent more orders than the 30-year portion, according to bankers and investors. The 30-year bond was yielding more than 6.1 per cent on Friday, compared with a 4.8 per cent yield on the five-year debt.
The overall order book was just over $60bn, marking a big decline compared with Amazon’s last trip to the capital market in March, when it secured more than $120bn of orders, the people said. So far the tech firm has raised nearly $90bn across different currencies.
Investors needed to offload existing hyperscaler debt to make room for Amazon’s new offering, given that many portfolios were already heavily exposed to AI debt, said John Lloyd, global head of multisector credit at Janus Henderson.
“You’re taking obsolescence and technological disruption risks over time,” Lloyd said. “For us to play a new deal, it has got to come with a pretty big concession to entice us.”
Recent volatility in tech stocks has also damped sentiment, he added. In addition, some investors already have large exposure to the tech sector in their equity portfolios, which may temper their appetite for further exposure in the debt market, according to Amanda Lynam, chief credit strategist at Goldman Sachs Research.
DoubleLine’s Entina added that, as long-dated debts are often purchased by insurance and pension funds that need to match their long-term liabilities, they are often more conservative.
The attraction of long-term AI debt is also lessened by the elevated short-term yields on US Treasuries, due to above-target inflation and expectations that the Federal Reserve’s policy rate will stay higher for longer, particularly following Kevin Warsh’s hawkish tone at his first meeting as the central bank’s new chair last month.
“Why would you add more risks if you have been able to secure attractive yields without extending too far in the curve?” said a credit analyst focused on high-grade credit.
Data visualisation by Ray Douglas
