MW Jerome Powell says the AI bubble and the dot-com bust are different. He's wrong.

By Brett Arends

The Federal Reserve chair's argument is based on inaccurate - and dangerous - Wall Street conventional wisdom

My apologies to Jerome Powell, but on Wednesday, the chairman of the Federal Reserve was talking out of his hat.

The subject wasn't inflation, or interest rates, or the president's attempts to fire him, but the current stock-market mania over artificial intelligence. This is so obviously a market mania that I honestly don't know how anyone on the Street of Shame can say otherwise and keep a straight face. (But that, I guess, is why they get paid the big bucks.)

Powell insisted the situation today is completely different from the infamous dot-com bubble of 1999-2000, which I remember almost as if it were a few months ago. And Powell's argument was based on the conventional wisdom on Wall Street - a conventional wisdom that is wrong, and therefore extremely dangerous.

"This is different," he said in response to a question from Victoria Guido of Politico. The current AI boom is different "in the sense that these companies, the companies that are so highly valued, actually have earnings and stuff like that. So, you go back to the '90s and the dot-com, they were- these were ideas rather than companies."

In other words, the AI boom is not another dot-com-style bubble because the leading companies today "actually have earnings, and it looks like they have business models and profits and that kind of thing. So, it's really a different thing." By contrast, using Powell's reasoning, the dot-com bubble was all about fake companies with no businesses, consisting of little more than air.

Phooey.

The dot-com bubble was nearing its peak 26 years ago, and on Wall Street, where most careers are not even measured in decades, that's eons ago. So it's no surprise that people today have mistaken memories, or none at all. But let's clarify.

The dot-com or tech bubble of the late 1990s was not dominated by companies with mere "ideas." Yes, of course those existed and, in many cases, raised money at ridiculous valuations. But the dominant companies on the stock market back then, even on the tech-dominated Nasdaq COMP, were "real" companies with "real" businesses, profits and sales, just as they are today.

I checked the data from FactSet: The most valuable companies on the Nasdaq at the peak of the bubble, at the end of February 2000, were tech giants Microsoft $(MSFT)$, Cisco Systems $(CSCO)$, Intel $(INTC)$, Sun Microsystems and Texas Instruments $(TXN)$. The top 100 on the Nasdaq was dominated by technology companies that were making IT infrastructure, and which had businesses, sales and earnings. The cliché of the time was that the dot-com era was like the California gold rush, and the people who were going to make the most money were those selling "picks and shovels" to the prospectors.

Microsoft, the world's most valuable company at the time with a market value of $465 billion, boasted $22 billion in sales over the previous 12 months, and $8.7 billion in earnings. Cisco Systems, No. 2 with a $450 billion market cap, had $15 billion in sales and $2.55 billion in earnings. Intel, No. 3 in value at $380 billion, had $29 billion in sales and $7.3 billion in net earnings - and so on. Of the top 30 companies on the Nasdaq by value, 23 had had sales of more than $1 billion over the previous 12 months, and 25 had positive earnings.

The idea that the dot-com bubble was dominated by Pets.com and eToys is just a myth.

The biggest dot-com, Yahoo, only very briefly saw its market cap break above $100 billion. Crazy, to be sure - but far behind the giants. And among dot-coms, it pretty much stood alone. Amazon.com (AMZN) was valued at the peak at less than $30 billion - about 5% of the value of Microsoft or Cisco, and less than Walgreens. EBay $(EBAY)$ was less than $20 billion.

As for Pets.com, which ended up as the poster child for dot-com insanity? It raised just $82.5 million in its IPO in February 2000, selling stock at $11. The stock hit an intraday peak of $14. As there were 35.4 million shares in issue, it meant the company was valued - very briefly - at $500 million.

Certainly, this was absurd, and Pets.com went bust within a year. But its maximum value was just 0.1% of the value of Cisco's stock.

Fly-by-night ventures got the headlines, but they were small compared to the behemoths. And it was on the behemoths - "real" companies with businesses, and sales, and earnings - that investors then lost their shirts, even when the underlying businesses survived or even flourished.

From the peak of the bubble to the end of 2000, a period of about 10 months, Microsoft investors lost more than half their money.

By the spring of 2001, merely a year after the bubble peaked, Cisco was down 75%. (Yes, really.) By the fall of 2002 Intel, too, had lost three-quarters of its value.

Even though Microsoft today is once again among the world's most valuable companies, it took 13 years after the dot-com bubble burst before investors even got their money back.

It's worth remembering, too, that the investors in the dot-com bubble were directionally correct. They were correctly betting that the emerging internet technology would transform society, business and the economy. They were correctly betting that the winners would make a fortune.

Those betting on AI may be completely correct along similar lines.

But tech investors back during the dot-com bubble were wrong on three key things: timing, valuation, and picking the winners.

On timing, they were way too early, because the internet didn't really start transforming the world until the rollout of high-speed internet, which began several years after the bust, and then the launch of the smartphone, which really began with the launch of the iPhone in 2007.

On valuation, the mistake was obvious. Microsoft ended up a winning company, but people in 2000 were buying the stock for 20 times the trailing 12 month's sales and 50 times trailing per-share earnings. Cisco, absurdly, was selling for 180 times the previous year's sales.

And then there was the question of picking the winners. Few of the tech names that have ended up dominating the new economy were among the big names of 2000. People back then thought Yahoo would dominate search and email. They'd never even heard of Google $(GOOGL)$ $(GOOG)$, which had only recently been started. They'd never heard of Facebook (META), which hadn't been created yet, let alone Tesla $(TSLA)$. Netflix $(NFLX)$ still sent DVDs through the mail.

If you had told anyone in 2000 that the smartphone era would be dominated by Apple $(AAPL)$, they would have laughed in your face. Apple's venture in mobile computing, the Newton, was a byword for failure and embarrassment.

They'd have assured you that the winner would be Microsoft's Windows Mobile - alongside Palm, Nokia $(NOK)$ and Ericsson $(ERIC)$.

Where does that leave us today? Nobody knows how big a bubble or mania will get, or when it is going to burst. The old joke on Wall Street is that this doesn't happen "until the last bear turns bullish."

But as the late Dan Bunting, a terrific money manager over many decades in London, used to say: "Never let a bubble go to waste. You will make the most money from the worst stocks." In other words, the brave can make good money from a mania - so long as they are nimble and remember to sell when things turn south. And as long as they don't mind taking big risks.

Meanwhile, long-term investors can follow the sage advice of Warren Buffett, who famously sat out the dot-com bubble and stuck to what he knew. He looked foolish - but only for a while.

-Brett Arends

This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.

(END) Dow Jones Newswires

October 31, 2025 14:08 ET (18:08 GMT)

Copyright (c) 2025 Dow Jones & Company, Inc.