It only takes a cursory glance at the United States bond market to get a meaningful insight into how tech giants are planning to make their grandiose artificial intelligence (AI) aspirations come true. The trouble is that Wall Street is struggling to keep up with the speed at which bonds from these cash-burning tech companies are coming to market.
So much so that the prices of newly issued bonds have slid, which has only fuelled investor anxieties over tech stock valuations, which look stretched relative to historical levels and other market sectors.
Based on The Financial Times estimates, corporate bonds from big tech firms are projected to account for over a quarter of the net supply of U.S. corporate bonds this year.
With AI-driven bond issuances increasing, JP Morgan expects the U.S. investment-grade bond market to expand to a record US$1.81 trillion in 2026.
To put the recent uptick in tech stock bonds in context, since the start of September, so-called AI hyperscalers Amazon, Alphabet, Meta Platforms and Oracle alone have issued around US$90 billion of investment-grade bonds.
According to Dealogic, that’s more than the amount they had sold over the previous 40 months.
Higher yields
While, these companies were able to successfully complete their sales, the higher interest rates they unexpectedly had to pay – to get them away – points to the market’s struggle to digest the sheer volume.
It’s understood that Alphabet and Meta recently raised interest rates by around 0.1–0.15 percentage points compared to previous issuances.
Despite their high credit ratings, the recent surge in bond supply means they must offer slightly higher yields to attract demand.
Equally telling was the sliding price of bonds from these companies, which raises concerns not only about the sheer volume of bonds but also growing investor concern about deteriorating credit metrics.
In other words, markets already unnerved by the sky-high valuations of AI businesses, fear that borrowers' ability to repay debt is declining, and this points to an increased risk of default.
AI growth narrative sours
Meanwhile, the cost of credit-default swaps – which act like an insurance policy against a borrower defaulting on a debt - has gone up, with investor faith in the AI-fuelled growth narrative now turning to scepticism.
“The markets are very interconnected now,” noted John Lloyd, global head of multisector credit at Janus Henderson Investors.
“It will be hard for the credit markets to do well if AI stocks are selling off.”
In short, AI stocks do appear to be selling off, with the powerful AI-led tailwind that underpinned tech stocks in recent years now unravelling.
Nowhere is this collapse more evident than on the tech-heavy Nasdaq Composite, down 6.1% in one month.
The “AI debt investment” trend among big tech firms - which issue bonds to secure investment funds - resembles the semiconductor equipment manufacturers’ large-scale capital expenditure “chicken game”.
This is a game where firms risk market-wide oversupply and reduced profitability in pursuit of market share and a technological competitive edge.
Investor patience waning
According to the Wall Street Journal (WSJ), the cash-to-asset ratios of these tech companies are rapidly depleting.
As a case in point, Microsoft’s cash and short-term investment ratio relative to total assets dropped from approximately 43% in 2020 to 16% as of the third quarter of this year.
Other major tech firms, including Google and Amazon, show similar trends.
“Investors are beginning to signal that they cannot maintain infinite patience toward tech firms that have yet to find a way to generate significant returns from AI investments,” the WSJ noted.
Beyond their AI capital expenditure burdens and potential reductions in shareholder returns, massive borrowing by big tech firms has raised fears that these companies could act as a black hole for financial market liquidity, driving up overall corporate bond yields and harming secondary-tier firms.
The WSJ forecasted, “as the transition to the AI era accelerates, high-cost investment decisions and the resulting uncertainties will only increase.”
Worrisome cash burn
However, recent pressure on tech company bonds is no means a level playing field.
Given that they have been able to fund most of their AI expenditure with the cash they generate quarterly, Alphabet, Amazon and Microsoft haven’t been hit as badly as other big tech issuers.
By comparison, when another tech titan, Meta – with less cash at the ready – issues US$30 billion of bonds at the end of October, it was forced to attract investors with yields comfortably above those of its existing debt.
Unsurprisingly, the prices of some of these bonds were trading lower in the secondary market, pushing yields higher.
Despite carrying double-A ratings, these bonds now yield around the same as IBM’s bonds, which are rated single-A by major ratings firms.
By comparison, Oracle has found itself in a more tenuous position, with plans to burn tens of billions in cash over the next few years in an attempt to morph into an AI cloud-computing giant.
Unsurprisingly, Oracle’s bonds – rated two notches above speculative grade territory – now carry yields that are higher than those of virtually any of its investment grade tech peers.
Some analysts now suspect that rising debt costs could eventually affect investment decisions on the margin for speculative-grade tech companies like data centres and developers.
With investors broadly demanding more of a premium, Will Smith, director's credit at AllianceBernstein, suspects a line of straight growth is less likely.
“People are really going to demand that only really sensible projects - that are structured in a way that works for debt markets, with the right cost of capital – actually get built.”
