Who will pay for the AI revolution? Retirees

Tech companies are facing an enormous investment need that will outstrip even their considerable existing resources. Morgan Stanley analysts in July estimated that of the roughly $3 trillion that is expected in global data-center capital expenditures through 2028, only about half of that could be funded by projected cash flows. That would leave a roughly $1.5 trillion financing gap.

To meet such a tremendous need, the companies will need to turn to the biggest funding markets. And when it comes to corporate borrowing, those include the mainstream, high-grade bond market. Issuance in the investment-grade corporate bond market represented around two-thirds of the more than $2 trillion sold this year through October across the U.S. corporate bond and asset-backed securities markets, according to figures compiled by Sifma.

There have recently been several big bond offerings from tech companies in the AI race, including Oracle, Meta Platforms and Google-parent Alphabet. JPMorgan Chase analysts recently forecast that the high-grade bond market could absorb $300 billion of AI-data-center-related issuance over the next year.

The mainstream corporate bond market is among the biggest, and often one of the cheapest, ways for companies to borrow. Historically, though, that market also has some rules of the road: It is for companies with the highest credit ratings and has been focused on bonds featuring public reporting and fairly straightforward structures. Many bonds mature in less than 10 years.

But the more a company borrows beyond what its cash flows support, the more its credit rating can suffer. And AI still has lots of questions around it: How much end-user demand will there ultimately be? What is the lifespan of the chips needed to power it? Will there be enough cheap electricity?

Meanwhile, life-insurance companies need to invest a growing wave of retirement money to help people generate income when they aren’t working. This year represents the peak for the number of people turning 65 in the U.S., which has helped fuel another record for sales of U.S. annuities, with $345 billion sold through the first nine months of the year, according to industry group Limra.

Demand from insurers has broadly been a driving force in the credit markets. It is one factor helping drive spreads on investment-grade corporate bonds—or the premium they pay above the yield on benchmark Treasurys—to around their tightest levels since the 1990s.

“The biggest change in the technicals of credit markets over the last two to three years has been the emergence of U.S. life insurance companies as the largest marginal buyer," says Vishwanath Tirupattur, chief fixed income strategist and director of quantitative research at Morgan Stanley. “This has driven credit spreads across the board tight."

A recent report on life insurers by economists at the Swiss Re Institute said that “in advanced and rapidly aging emerging markets," increasing demand for retirement income “will likely shift the liability structure towards longer duration." The so-called longevity risk that retirees will live longer—stemming from advances in biosciences and healthcare—means that insurers need to be on the hunt for longer-term assets, too.

In their investments in recent years, insurers have been willing to turn to less-vanilla instruments. A research analysis published by the Federal Reserve Bank of Chicago found that life insurers had been investing more in “higher yielding, but more complex, private placements."

So with ready potential buyers that can stomach additional complexity or duration in exchange for more yield and size, even in the investment-grade market, more offerings to fund AI’s build-out seem likely.

Analysts at JPMorgan in October wrote that the public, high-grade bond market “has grown increasingly comfortable absorbing unconventional financing vehicles tied to data center growth."

“You’ll see more of this from hyperscalers," says Ben Hunsaker, head of structured credit at Beach Point Capital Management. “The expectation is that if they win at AI today, they can make trillions. So what if they pay a bit more in interest."

Everyday investors, though, might in the future need to spend more time evaluating what they might have previously looked at as a highly straightforward market.

Those who turn to corporate bonds mainly because they want something with as little risk as possible might need to be more selective about what part of the market they buy. Active bond fund managers won’t mind if investors might want more careful selection in their portfolios.

People nearing the end of their careers might never have to reckon much with the challenges and opportunities of AI in their jobs. But it might still play a big role in their retirement years, in a very different way.

Write to Telis Demos at Telis.Demos@wsj.com