Creators revive dot‑com bubble talk

Stocks are back at record highs, and the internet has reached for the oldest tech-market analogy it knows. Dot-com bubble. That talk got louder on Friday, May 8, when Michael Burry said the market felt like “the last months of the 1999-2000 bubble” just as the S&P 500 and Nasdaq closed at fresh highs. The reason this landed is simple — the rally looks powerful, but it also looks narrow, fast, and very AI-heavy. ### Why are people saying “dot-com” again? Because the setup feels familiar. A new technology story takes over everything, investors decide the winners will be huge, and money rushes into the obvious names first. Burry’s point was not that every AI company is fake. It was that price action has started to detach from normal inputs like jobs data or consumer sentiment, with stocks rising because they have already been rising. (cnbc.com) ### What actually happened this week? The S&P 500 and Nasdaq both hit record closes on May 8. The Nasdaq finished at 26,247.08 and the S&P 500 at 7,398.93. Tech earnings helped, the April jobs report came in stronger than feared, and AI-linked names kept doing the heavy lifting. Nvidia rose, while Micron and Sandisk each surged more than 15% on demand tied to AI data-center buildouts. (cnbc.com) ### Why do semiconductors matter so much here? Because chips are the picks-and-shovels trade for AI. If investors believe every company will need more computing power, they buy the companies selling GPUs, memory, networking gear, and storage first. That is why Burry focused on the Philadelphia Semiconductor Index. It was up more than 10% in that week alone and 65% for 2026 by May 8. When one part of the market runs that hard, people start asking whether they are looking at genuine earnings power or classic melt-up behavior. (cnbc.com) ### Is this just hype, though? Not exactly — and that is the catch. The dot-com era had plenty of companies with weak businesses and no profits. Today’s AI leaders are mostly giant, profitable firms with real cash flow, real customers, and real infrastructure demand behind them. This rally is speculative at the edges, but it is also being fed by actual spending on data centers, memory, and software. (cnbc.com) That makes the comparison useful as a warning, not as a perfect match. ### So why are people still nervous? Because concentration cuts both ways. A handful of AI-linked stocks can drag the whole index higher, but that also means the index is more exposed if those names wobble. Reuters noted that even on one of these record days, most S&P sectors were down. That is the classic narrow-rally problem — the headline index looks healthy, but leadership is clustered in a small group. (money.usnews.com) ### Does everyone bearish think the top is here? No. Even some veterans who see 1999 echoes do not think the end is immediate. Paul Tudor Jones said the backdrop feels similar to 1999, but he also argued the bull market could keep running for another year or two before a bigger correction. Basically, “bubble” is not a clock. It is a description of behavior. Markets can stay euphoric much longer than skeptics expect. (virginiabusiness.com) ### What should investors take from this? Probably not “sell everything.” The more useful takeaway is to stress-test what you own. Ask which positions depend on real earnings growth and which ones depend on the story staying hot. In a narrative-driven market, both can rise together for a while. But when the mood shifts, the story stocks usually find out first. ### Bottom line? The dot-com comparison is back because the market is acting like a market in love with one idea. (cnbc.com) But this time the biggest winners also have real businesses. That does not remove the risk — it just changes where the weak spots are.