What the dot-com bubble can teach investors worried about AI

There’s a bubble in the number of people calling for bubbles.

That is how Zor Capital Managing Director Joe Fahmy described the current environment on Wall Street. It’s never been more popular, in other words, to claim that stocks — specifically those betting big on artificial intelligence — are too high.

Here’s the argument from those concerned about a potential bubble: First, valuations are high. The forward 12-month price-to-earnings (P/E) ratio of the S&P 500 is near 23. That’s higher than it’s been in five years and 24% more expensive than its 10-year average (18.6).

Second, a disproportionate amount of the market’s returns has been concentrated in a small number of stocks. Data from J.P. Morgan suggests that since the October 2022 low, 75% of the S&P 500 gains, 80% of the earnings growth and 90% of the capital spending growth have come from the “Magnificent Seven” (Nvidia, Microsoft, Apple, Alphabet, Amazon, Meta and Tesla) plus 34 other companies tied closely to AI data centers.

Third, much of the $400 billion per year (!) spent on AI infrastructure might not prove to be profitable. It’s too early in the cycle to quantify “how much is too much,” but the magnitude of spending doesn’t leave much margin for error.

On the other hand, plenty of people near the epicenter of the AI boom suggest we could be underestimating the positive effects AI will have on economic growth, labor efficiency and corporate profitability. The truth is, it’s too early to know for sure.

There are, however, several lessons learned from the dot-com bubble of the late 1990s that can help guide investors through the uncertainty ahead. Here are a few:

The internet changed the world and the global economy as much, if not more, than predicted. The NASDAQ still fell nearly 80% from its peak in early 2000. For every corporate winner, there were a handful of startups that went bankrupt and left a sizable hole in investment portfolios. It’s not enough to bet on the right trend. You need to buy the right companies. Or at least avoid the wrong ones.