Michael Burry warns dot‑com vibes

Stocks are back in one of those moods where bad news barely matters and good news gets turned into rocket fuel. That is the setup for Michael Burry’s latest warning. On May 8, he said the market feels like “the last months of the 1999-2000 bubble” — basically the stretch when prices stopped caring much about fundamentals and started feeding on their own momentum. The timing matters because the S&P 500 just pushed to another record even as the labor data softened rather than strengthened. ### What did Burry actually say? Burry’s point was not just “stocks are expensive.” He argued that prices are no longer moving in a normal, data-driven way. His line was that stocks are not rising or falling because of jobs or consumer sentiment — they are rising because they have already been rising. That is a very specific bubble diagnosis. It means momentum itself becomes the story. (cnbc.com) ### Why bring up 1999 now? Because the comparison is not random nostalgia. Burry tied today’s AI frenzy to the final phase of the dot-com boom, when investors crowded into a narrow group of tech winners and treated every dip like a buying opportunity. He also pointed to the semiconductor surge as a close parallel. The Philadelphia Semiconductor Index is up more than 10% this week and roughly 65% so far in 2026 — the kind of vertical move that starts to look less like steady repricing and more like speculative acceleration. (cnbc.com) ### Was the jobs report really that bad? Not really — but it was softer than a market at record highs would usually want. Nonfarm payrolls rose by 115,000 in April, down from 185,000 in March. The unemployment rate held at 4.3%. That is not recession math. But it is also not the kind of hot labor print that screams reacceleration. The more telling detail was underneath: the number of people unemployed for less than five weeks jumped by 358,000 to 2.5 million. (cnbc.com) ### And what about jobless claims? They nudged higher too. Initial claims rose to 200,000 for the week ending May 2 from 190,000 the week before. On its own, that is still a low number historically. But paired with a slower payroll gain, it adds to the sense that the economy is cooling a bit while the market keeps acting like nothing can interrupt the AI trade. That gap is really the whole story here. (bls.gov) ### So is Burry saying a crash is next? Not exactly. He is saying the market’s behavior looks late-cycle and self-reinforcing. That is different from calling the exact day the music stops. Burry has made high-profile bearish calls before, and some landed early. But the reason people pay attention is that he is usually talking about structure, not headlines — crowding, leverage, narrative obsession, and price action detached from the economic backdrop. (fred.stlouisfed.org) ### Why does AI sit at the center of this? Because AI has become the market’s two-letter shortcut for future growth. That is the catch. A real technology boom can also become a valuation bubble. In 1999, the internet was real, but prices still got absurd. Burry seems to be arguing that the same split exists now — genuine technological change on one side, and a market willing to pay almost anything for the winners on the other. (michaeljburry.substack.com) ### Why are investors still buying anyway? Because bubble phases often feel smartest right before they feel dumb. If a narrow group of stocks keeps beating expectations and pulling indexes higher, investors who stay cautious can look wrong for a long time. That creates career pressure, FOMO, and a reflex to chase whatever is already working. Basically, rising prices recruit more believers. That is why Burry’s warning is less about one jobs report and more about market psychology. (cnbc.com) ### Bottom line? Burry is not really arguing over one payroll number. He is saying the market’s reaction function looks broken — softening economic data, rising jobless claims, weak sentiment, and yet stocks keep levitating. When that happens, the danger is not just overvaluation. The danger is that price becomes its own justification, right up until it doesn’t. (cnbc.com)