US stock market valuations have reached levels approaching those seen during the dot-com bubble of 2000, raising concerns about potential significant losses for investors. What exacerbates the risk is that these elevated valuations coincide with company earnings that are also well above their historical norms. If valuations were to revert to average levels simultaneously with a decline in earnings, investors could face compounded losses.

The cyclically adjusted price-to-earnings ratio (CAPE), popularized by Yale University’s Professor Robert Shiller as a tool for assessing long-term stock market prospects, has once again drawn attention. In May, the CAPE for the S&P 500 surpassed 40 times earnings, a level not reached since the height of the late 1990s tech boom. This comparison has prompted discussions about similarities between the current AI-driven market surge and the internet-driven expansion at the turn of the millennium.

While current supporters argue that the companies at the center of today’s boom – including major hyperscalers and semiconductor firms – are profitable, unlike many internet startups during the previous bubble, some analysts remain skeptical. They note that these firms are not just profitable but are earning what can be considered supernormal profits.

An analysis of S&P 500 earnings per share (EPS) indicates that current earnings are approximately 59% above the long-term exponential growth trend, a significant jump from 14% above trend at the start of 2023, when the AI boom gained momentum. Most of this earnings acceleration has occurred in the last year.

Using a statistical measure called the z-score, which compares current values to historical averages adjusted for volatility, the CAPE for the S&P 500 now stands at 2.9—well into bubble territory. While this is slightly lower than the 3.3 measured at the peak in December 1999, it exceeds the 1.8 reading at the peak of the 1929 market bubble.

However, the key difference lies in the earnings z-score. In early 2000, the EPS z-score was a moderate 0.6, indicating elevated but not extreme earnings, mostly outside of the tech sector. Today, the EPS z-score is at 1.8, approaching bubble levels, with earnings heavily concentrated in companies linked to the AI market surge.

This situation suggests that the current market is experiencing both a valuation bubble and approaching an earnings bubble. Some of the largest tech companies have maintained elevated profit margins and earnings for more than a decade due to their capital-light business models and unique product offerings. Yet, if earnings were to regress to the long-term trend amid rising capital expenditures by tech firms, the CAPE could effectively jump to 64 times earnings, yielding a z-score of 4.6 — a level considered extraordinarily rare in statistical terms.

While a market crash is not imminent, and tech booms can persist for extended periods, this analysis emphasizes elevated risks for long-term investors. Market observers recommend vigilance and readiness to reduce exposure should market conditions shift.