'It's not that different': why is Michael Burry comparing the bubble in AI to the dotcoms?

Investor Michael Burry has published the first post on his new blog titled "The main sign of a bubble: oversupply". In it, he explains why he thinks it is a mistake to claim that the current hype around artificial intelligence is different from the dot-com bubble that collapsed in the early 2000s.

Burry is convinced that many investors "don't remember or simply don't know" what triggered the dot-com bubble and its subsequent collapse. At the center of the excitement, he writes, was not .com sites, but the rapid construction of data infrastructure. Burry calls it a "mania" - companies sought to transmit data "at ever-higher speeds around the world, to every home." Businesses - from telecom carriers like AT&T to pipeline company Williams - were spending tens of billions to develop networks, including laying undersea cables.

In 1999, Cisco's stock, which produced much-needed routers for networks, soared 125%, and in 2000 the company briefly became the world's largest company by market capitalization. Executives argued that Cisco "was at the center of an economic revolution that was changing not only the economy but the entire society," Burry cited company press releases from the time. In 2000, he said, shortages of communications equipment were "the norm," and investors thought demand was huge. Cisco executives also claimed that they saw no signs of slowing demand and that, on the contrary, "the radical transformation of the Internet business was accelerating," the publication said.

But the reality was different: broadband adoption was very slow, Burry writes. According to the data he cites, by 2002, less than 5 percent of the infrastructure that had been rapidly built during the bubble was functioning in the United States. Eventually, Burry says, the bubble burst because of "catastrophic oversupply and grossly insufficient demand."

In 2001, Cisco started losing money and its stock collapsed 78%. According to Burry, investors didn't sell the stock until the last minute, hoping to get a signal of slowing demand first, but they ended up losing money.

According to Burry, the largest companies "have found that planning for grandiose spending is a powerful incentive to grow shareholder value." The five publicly traded leaders of the current AI boom - Microsoft, Google, Meta, Amazon and Oracle - and non-public startups like OpenAI have pledged to invest nearly $3 trillion in AI infrastructure over the next three years, and "investors are thrilled," Burry writes.

At the center of the current bubble, he says, there's a Cisco of its own, "with picks and shovels for everyone and a broad vision for the future." "Its name is Nvidia," Burry writes. Like Cisco 25 years ago, Nvidia claims there is no bubble in the artificial intelligence market, the investor reminds us.

Burry became famous for predicting the collapse of the mortgage bubble in the U.S. housing market, leading to the 2008 global financial crisis, and making money from it. His bet against the market was the basis for the book "The Downgrade Game" and the movie of the same name.

This fall, Burry returned to social media for the first time in two and a half years to warn of a new market bubble. He accused hyperscalers including Meta and Oracle of artificially inflating profits, and bet Palantir and Nvidia's stock prices would fall. In November, the investor closed his hedge fund Scion Asset Management, obviating the need to disclose his trades publicly, and started a paid newsletter, promising to share investment ideas.

This article was AI-translated and verified by a human editor