On March 10, 2000, the Nasdaq index peaked at 5,048 units. It would not see that level again for fifteen years, reaching a new all-time high only on April 23, 2015. In between lay the wreckage of the dot-com bubble, which is why investors now arguing about artificial intelligence and data centres keep reaching for it as a yardstick.
How the boom was built
The raw material was real. By the mid-1990s the cost of sending and storing information had fallen sharply, personal computers were spreading, and the World Wide Web was changing how trade, finance and services worked, Goldman Sachs recounts in its firm history. Entrepreneurs drew up plans for online businesses. Venture capitalists, drawn by the prospect of profit, were keen to pay for them.
The economy cooperated. The expansion was the longest in the United States since the Second World War, with inflation and unemployment falling. Money was cheap, too: the Federal Reserve had cut interest rates after the collapse of the hedge fund Long-Term Capital Management in 1998. From October of that year, markets cheered a seemingly endless run of dot-com flotations and paid little attention to whether the business models worked. Most of these startups, Goldman notes, would never produce revenue or profit.
The weight of the Nasdaq tells the story. In 1990 stocks traded there were worth 11% of those on the New York Stock Exchange. By December 1999 the figure was 80%. The index rose 86% in 1999 alone, and the AOL and TimeWarner merger seemed to confirm that a “new economy” had arrived.
Britannica describes the mindset. Many investors came to believe that the old yardsticks, balance sheets, profits and cash flow among them, no longer applied to internet companies, so money kept flowing even into debt-laden firms with no realistic route to profit. Alan Greenspan, then chair of the Federal Reserve, is credited with the phrase “irrational exuberance” for that overconfidence.
The unwinding
Early in 2000 the Federal Reserve announced a modest rise in interest rates to head off inflation. Dot-com investors began dumping their holdings in a panic. Cash-strapped startups turned worthless within months, and the market for new share issues froze, according to Goldman. By October 4, 2002, the Nasdaq stood at 1,139.90, a fall of 77% from the peak. Britannica notes that most publicly traded dot-com companies had failed by then.
The damage did not stop at the border. Goldman lists tech markets in Tokyo, Seoul, Frankfurt, London and Paris that were dragged down in the same reversal. The economy slid into recession in 2001.
The comparison with AI
Is the data-centre build-out a rerun? Jeff Buchbinder, chief equity strategist at LPL Financial, said in comments reported by Yahoo Finance in May 2026 that the comparison between the AI boom and the dot-com era is reasonable, and that the patterns in the Nasdaq 100 are broadly similar if measured from the launch of Netscape and from the launch of ChatGPT. But the scale differs. The index has gained more than 140% since ChatGPT arrived. At the March 2000 peak, its return stood at 1,090%.
Buchbinder argued the two moments are fundamentally different. The leaders of the AI build-out are paying for growth out of internal cash flow rather than “speculative capital raising”, he said, and the tech sector trades at roughly 25 times forward earnings against 58 times at the 2000 peak.
Not everyone is so relaxed. Strategists cited by Yahoo Finance pointed to a surge in debt issued by hyperscalers such as Alphabet and Oracle, which are funding $725 billion in capital expenditures projected for 2026. Buchbinder himself was careful. “We’re not saying history will repeat,” he said, adding that the current market may be “more like 1997 than late 1999 or early 2000.”
Sources
- The Late 1990s Dot-Com Bubble Implodes in 2000 | Goldman Sachs
- Dot-com Bubble & Bust | Definition, History, & Facts | Britannica Money
- Why the Nasdaq 100's AI-driven stock boom still looks tame compared to the dot-com era | Yahoo Finance
