In the last few months, many influential voices have raised concerns about the risk of a stock market bubble building from the sky-high valuations of artificial intelligence (AI) stocks. Some people are comparing the current moment to the late 1990s, when “irrational exuberance” led to the dotcom crash in 2000. There is, however, a different group of people, including some Wall Street analysts, who say that this bull market still has legs to run.

The biggest companies in the AI space, such as Nvidia, Microsoft, Alphabet, Amazon and Apple, are led by some of the greatest visionaries on the planet. Their genius is what enabled those individuals to build the amazing enterprises that have made them centibillionaires. They and their companies are currently investing hundreds of billions of dollars in the technologies that will power the AI revolution. Obviously, they are doing so because they strongly believe in the enormous potential of AI to transform how the world operates.

Those who are raising alarms about the emergence of an AI-related stock market bubble do not question the technology’s vast potential benefits. They rather argue that the returns on all that invested capital are unlikely to be as high as the current stock valuations suggest. Although there is little chance of that happening, I wonder what the tech visionaries themselves would say if any of them could volunteer to offer an opinion. Would they agree, if they thought so, that their company stocks are excessively overvalued? Or would they side with the analysts who say the valuations are justified and that the stock prices could go even higher?

Nvidia earnings announcements, in particular, have lately become eagerly awaited events because of the significant weight the stock carries in the indices that include it as a component. On such corporate earnings calls, company bosses provide data to show how their firms performed over the reporting period, and what they expect in terms of operational and financial performance in future periods. After that, it is up to Wall Street analysts, institutional fund managers and retail investors to decide what to do with that information. That is where the problem lies.

In hot sectors such as AI, it is typical for a company stock price to move steeply higher on the slightest hint of positive news. All it takes is one or a handful of analysts placing a buy rating on the stock. Herd mentality instantly kicks in, causing the price to jump because it becomes a party that everyone wants to attend. The same thing happens when the news is negative. Investors frantically dump the stock because no one wants to be caught holding it after it has lost much of its value. They sell first and ask questions later.

Corporate bosses sometimes grumble when they think that some analyst has taken an unfairly dim view of their company stock. An analyst’s sell or hold rating can cause a significant portion of a company’s wealth to be wiped out. In the case of the AI company bosses, they are probably quietly happy, even if they think the values of their stocks are overinflated. Whatever overvaluation exists helps provide some of the capital they need for all those massive investments they are trying to make.

One of the prominent voices sounding the alarm about the risk of an AI-related stock market bubble is Gita Gopinath, former first deputy managing director of the IMF and a previous chief economist of that institution. She is currently a professor of economics at Harvard. In her recent article in The Economist, she writes that over the past decade and a half, American households and foreign investors have significantly increased their holdings in the U.S. stock market, lured by strong returns and the dominance of American tech firms. Given the scale of the exposure, she worries that a bursting of the bubble would have seismic implications for the global economy.

Gopinath calculates that a stock market correction of the same magnitude as the one that occurred in the dotcom crash could wipe out over $20 trillion in wealth for U.S. households. That amount is about 70 percent of U.S. GDP in 2024. She estimates that foreign investors could lose more than $15 trillion, about 20 percent of the rest of the world’s GDP. According to Gopinath, overseas investors lost roughly $2 trillion ($4 trillion in today’s money) in the dotcom crash, a sum that was less than 10 percent of the rest of the world’s GDP at the time.

Some economists have said in recent weeks that a bursting of an AI-related stock market bubble would almost certainly lead to a global recession. But they think that it would be a shallow one. That tempered view isn’t much comfort however, given the frightening picture that Gopinath painted in her article.

By their very nature, stock market manias require piercing instruments from the outside to deflate the bubbles. Someone or something has to scare the herd to force it to jump off the bandwagon. Those balloon-popping tools often come in the form of contrarian investors who take short positions in stocks, or authoritative voices who “talk down the market,” as is currently happening with the warnings about overstretched valuations. But what if it is the party poopers who are wrong in the case of this AI-driven stock market?

It could very well be that there is still a good bit of upside potential in AI company stocks, and the market overall, and that those pessimistic voices are talking it down prematurely. If that is the case, they could be robbing investors of the additional wealth they would have accumulated had the party not been stopped too early. Is this one of those instances though where it is better to be safe than sorry?