Michael Burry Thinks The AI Boom Is Another Dot Com Bubble And Investors Are Sleepwalking To Disaster

Earlier this week, famous investor Michael Burry declared that the US equity market had “jumped the shark,” drawing explicit parallels to the final stages of the 1999–2000 dot-com bubble.

His warning is consistent with several objective valuation and sentiment indicators that suggest stretched multiples and deteriorating consumer fundamentals. Whether or not a crash is imminent, his comments are worth listening to, although it isn’t necessarily the case to just ‘get out’ as they imply at first glance.

Who Is Michael Burry?

Michael Burry is the founder of Scion Capital, best known for identifying and betting against the US subprime mortgage market ahead of its 2008 collapse — a trade later immortalised in Michael Lewis’s book and the 2015 film The Big Short. His analytical approach is idiosyncratic, evidence-driven, and decidedly contrarian. He has correctly called major inflection points in 2000, 2007, and (with COVID assistance) 2019. He has also been wrong a number of times since, predicting crashes that did not materialise in 2021 and again in subsequent years, earning himself, by his own admission, the reputation of the boy who cried wolf.

What Did Burry Say?

In a recent Substack post, Burry described a strong sense of déjà vu when observing current market conditions. His core argument: the market is no longer pricing in economic fundamentals. Instead, it is being driven almost entirely by AI enthusiasm, in the same way the late 1990s market was driven by “dot-com” optimism. An enthusiam that was irrational, reflexive, and disconnected from what is actually happening in the broader economy.

“Stocks are not up or down because of jobs or consumer sentiment,” he wrote. “They are going straight up because they have been going straight up. On a two letter thesis that everyone thinks they understand.”

He acknowledged his imperfect track record candidly, noting that he has become a meme for his repeated crash predictions. But he also pointed to the occasions where he was right — 2000, 2007, 2019, and the meme stock unwind in mid-2021.

Is Michael Burry Right? What the Data Shows

The honest answer is: partially, and possibly more than partially.

1. Valuations are stretched by virtually every measure.

The S&P 500’s trailing twelve-month P/E ratio currently stands at approximately 29x, well above its 5-year average of 24.6x and its 10-year average of 23.3x. The forward P/E is around 21x, still above both the 5-year average (19.9x) and 10-year average (18.9x).

The Shiller CAPE ratio, which adjusts for inflation across a decade of earnings, sits near 39 — a level historically associated with very modest forward returns and elevated drawdown risk. For context, the S&P 500’s long-run historical P/E average is roughly 15–16x. Burry’s argument is that the current market is not pricing in normal conditions.

2. Consumer sentiment is at a 74-year low — and the market is ignoring it.

This is arguably Burry’s most compelling point. We all know consumer sentiment down under is lower than it was even in March 2020 amidst the pandemic panic. But even in America things aren’t smooth sailing.

The University of Michigan’s Consumer Sentiment Index fell to 48.2 in early May 2026, its lowest reading since the survey began in 1952. Three of the four lowest readings in that 74-year history have occurred in the past nine months.

The index’s current conditions component declined to 47.8, driven by growing concern over energy prices (gas above US$4.55 nationally), the inflationary knock-on effects of the US-Israel conflict in Iran, and cumulative price pressures that have lifted prices roughly 25% over five years. Real wages are negative: average hourly earnings rose 3.6% over the past year while inflation is running near 4%.

The labour market added 115,000 jobs in April, and unemployment held at 4.3%. That is not a recessionary reading, but EY-Parthenon economists have flagged a “largely frozen labour market” for the rest of the year, characterised by low hiring and low firing. It is a labour market that supports stability, not expansion — and certainly not the kind of consumer demand that would justify the valuations currently being assigned to US large caps.

3. The AI narrative is real, but so is the risk of overextrapolation.

In our view, the counterargument to Burry deserves to be taken seriously. The dot-com era featured companies with no earnings, no revenue, and no plausible path to either.

The current AI-driven rally is anchored, at least in part, by genuine capital expenditure. Nvidia, Microsoft, Meta, and Alphabet are spending hundreds of billions on data centres and compute infrastructure. Q1 2026 earnings growth for S&P 500 companies came in at approximately 27.7% year-on-year, and analysts are projecting full-year 2026 earnings growth of 21%. Those are not fictitious numbers.

However, the pattern Burry identifies – that equities are rising not because of earnings but because of momentum – is consistent with how late-stage bubbles operate. A market can be both earnings-supported and simultaneously overvalued on a risk-adjusted basis. The two are not mutually exclusive.

Paul Tudor Jones, speaking to CNBC on 8 May, offered a more measured version of a similar concern. He suggested the current rally may have another year or two to run, but warned that if the market rises another 40% from here, the market-cap-to-GDP ratio could approach 300–350%, implying corrections of a magnitude not seen since 1929.

What Investors Should Take Home

Burry has a credible track record at identifying structural vulnerabilities, even if his timing is unreliable at times. Investors should not treat this as a sell signal, but they should treat it as a risk management prompt. Several practical considerations follow.

First, elevated valuations do not require a crisis to produce negative returns, they only require disappointment. If AI capital expenditure does not translate into the earnings growth currently priced in, or if macro headwinds intensify, the margin for error is thin at 21x forward earnings.

Second, the divergence between consumer sentiment and equity prices is historically unusual and historically unsustainable. One of the two will correct eventually. Either sentiment recovers (possible, if inflation eases and geopolitical risk subsides) or equity prices adjust to reflect the economic reality that consumers are experiencing.

Third, geographic diversification has rarely been more relevant. US large-cap valuations are, on almost every measure, more stretched than international peers. Markets in Europe, Japan, and parts of Asia trade at substantially lower multiples with comparable or improving earnings trajectories.

Finally, position sizing matters more in a late-cycle environment. Holding a greater proportion of defensive assets (quality bonds, cash equivalents, or low-beta equities) is not a bet that the market will crash. It is an acknowledgment that the asymmetry of outcomes has shifted. The upside from here, based on historical valuation relationships, is more limited than in prior cycles. The downside, if something goes wrong, is not.

Burry may be early again. He may be wrong again. But the structural case he is making (namely, that markets are pricing optimism, not fundamentals) is supported by data that rational investors should not dismiss. Perhaps the big AI players like Anthropic will continue to make money, but just because your microcap ASX tech stock put out an AI-focused investor presentation, it doesn’t mean it is going to do the same.