
Warning signs lighting up
In 1999, Warren Buffett publicly warned that stock prices could start to fall. The market ignored him — at least temporarily. Now, several of the same indicators he highlighted at the time are pointing to comparable or even more extreme values, according to an analysis published by Nasdaq Markets.
The so-called Buffett Indicator, which measures the ratio of total stock market capitalization to gross domestic product, has reached 238% in the US as of June 2026. During the dot-com bubble, it peaked at around 148%. The total value of the US stock market is now estimated at $75.7 trillion, against an economy of approximately $31.8 trillion.

Shiller P/E: Most expensive market in 25 years
Another widely followed valuation measure, the Shiller CAPE ratio (which adjusts for cyclical fluctuations in earnings over ten years), crossed 42 for the first time since July 2000 during July 2026, according to the research material. During the dot-com bubble, this ratio peaked at 44.
David Rosenberg of Rosenberg Research stated in July 2026 that the indicator, outside of the dot-com period, shows that "this is the most expensive S&P 500 market in history." Russ Mould of AJ Bell emphasizes that US equities look expensive "no matter how you slice the numbers," and that most valuation measures are sitting in the "top 10 percent of their historical ranges."

Concentration and momentum at extreme levels
Market concentration is another standout feature. The ten largest companies in the S&P 500 now account for 39% of the index's total weight, compared with 27% during the dot-com bubble. The five largest alone represent 30%, up from 17% at that time.
The S&P 500 Momentum Index rose 34% during April and May 2026 — the strongest two-month period in over three decades. Morningstar Wealth's CIO Americas, Philip Straehl, points to this as a sign of market extremes. The last time a comparable surge was recorded was in late 1999.
Key differences: This is not 1999
Despite the striking similarities, there are significant structural differences between today's market and the dot-com bubble — which makes the picture more nuanced.
The leading technology companies today, often referred to as the "Magnificent Seven," are among the world's most profitable businesses. Average profit margins in the technology sector now stand at around 26%, more than double the level seen in 2004. By comparison, many dot-com companies were characterized by negative cash flows and speculative growth with no earnings.
NASDAQ's P/E ratio today is around 35 — markedly lower than the 100 the index reached during the dot-com bubble. JPMorgan has concluded that the AI sector does not meet the classic criteria for a financial bubble, and Goldman Sachs believes the current situation more closely resembles 1997 than 1999, with forward-looking technology P/E ratios of 25–30, compared with 50–58 in the early 2000s.
AI boom: Real value creation or a new bubble?
The AI wave is today the primary driver behind the market rally, much as the internet was in the late 1990s. Ray Dalio of Bridgewater said in early 2025 that investment levels in AI are "very similar" to the dot-com bubble. Sam Altman of OpenAI expressed in August 2025 that he believes an AI bubble exists, while Jamie Dimon of JPMorgan stressed that "AI is real," but that some of the capital being invested now will be lost.
One important distinction is that the current AI investment wave is largely being financed by cash-rich companies — not by debt and venture capital flowing into immature businesses, as was the case during the dot-com era. Analysts estimate that cumulative debt tied to data center construction and AI infrastructure will reach $1.5 trillion by 2028.
What does this mean for investors?
The source material does not provide grounds for claiming that a crash is imminent. What can be said is that valuation levels for the US stock market are in historically extreme territory across multiple metrics — and that market dynamics, including heavy concentration and elevated momentum activity, display characteristics that have been observed ahead of previous major corrections.
The decisive factor is whether the underlying earnings expectations will materialize. If they do, high multiples may prove justified. If they do not, the downturn could be severe. History offers no guarantee of the outcome — but it does give reason to ask questions.
This article was written using large language models under editorial supervision by Aprex. Content is source-verified and auditable. Read our method →