Mark Zandi, chief economist at Moody's Analytics, said on LinkedIn Post on Sunday that tech companies are now issuing more debt than before the dot-com crash as the AI boom drives rapid infrastructure development.
Even after adjusting for inflation, big tech companies are issuing more corporate bonds than they were in the late 1990s. And companies aren't just refinancing existing debt, they're taking on additional debt.
“While the increasingly aggressive (and creative) borrowing by AI companies is not a negative for them, their debt could quickly become a problem if they underperform investors' expectations and their stock prices decline,” Zandi wrote.
“Borrowing by AI companies should be noted as a growing potential threat to the financial system and the economy as a whole.”
Zandi said in a LinkedIn analysis last week that the 10 largest AI companies, including Meta, Amazon, Nvidia and Alphabet, will issue more than $120 billion this year.
And he pointed out that this is different from dot-com-era bond issuances because Internet companies back then didn't have a lot of debt. Instead, it received equity and venture capital funding.
“That is not the case with the AI boom,” Zandi added.
While hyperscalers such as Amazon, Google, Meta and Microsoft can pay for AI buildouts with their own profits, issuing debt is the “cheapest and cleanest” way to fund infrastructure builds of this scale that will likely last more than a decade and be worth trillions of dollars, said Shay Boroa, chief market strategist at Futurum Equities. luck.
“These companies are much more comfortable issuing, for example, 10- to 40-year securities at very low spreads, because the market now views them as quasi-utility companies, not just technology companies, because they're building all this infrastructure,” Boroa said.
He added that over the past six months, tech companies have shown “virtual evidence” that future demand for AI will skyrocket.
Despite concerns about an AI bubble, Nvidia released a strong third-quarter earnings report last month, announcing a 66% year-over-year increase in AI data center revenue.
Still, critics warn that the augmentation may not be able to keep up with the rapid development of AI.
George Calhoun, director and professor at the Hanlon Center for Financial Systems at the Stevens Institute of Technology, said computer hardware, which makes up the bulk of the cost of AI data centers, may become obsolete during the AI boom and be more likely to be replaced by more advanced technology, as opposed to wireless and internet enhancements, much of which is still in operation today. luck.
“The innovation cycle in the chip industry is much faster than in wireless technology or fiber optics,” he explained. “There is a real risk that much of that hardware could become competitively disadvantaged by new technology in a shorter period of time before it is fully repaid.”
At the same time, Calhoun said the big players in the AI boom, namely OpenAI, don't currently have the returns to cushion their huge investments, increasing risk.
“If OpenAI fails, the snowballing effect will be significant,” Futurum Equities' Boloor said. He added that while large technology companies would be less affected by a potential OpenAI collapse, companies that rely heavily on OpenAI's business, such as Oracle, could be affected.
Still, Boror is optimistic about the development of AI, saying the main bottleneck to its success is America's energy capacity.
“I think the risk is that trillions of dollars of AI capacity will be built faster than the North American power grid can support it, potentially delaying its delivery,” he warned.
