Market Valuations Approach Dot-Com Peaks Despite Widespread Economic Despair

Wall Street is partying through one of the worst consumer moods on record. The S&P 500 has risen for eight consecutive weeks while the University of Michigan's consumer sentiment index just hit its lowest reading in more than 70 years of surveys.

The Shiller cyclically adjusted price/earnings ratio, which smooths stock valuations against a decade of inflation-adjusted earnings, sits at 40.8. That level has been breached only once before in 145 years of data, and that sole precedent was the dot-com bubble peak around 2000.

Stocks rally as consumers despair

The Dow Jones Industrial Average hit a record close for two consecutive days last week, even as the Michigan sentiment index printed below its prior low from June 2022. That prior low came when inflation was running at the highest level in decades. The gap between where stocks are trading and how households feel is historically without precedent.

Joanne Hsu, director of consumer surveys for the University of Michigan, put the context plainly: "Prices remain extremely high, labor markets have unambiguously weakened in the last four years, and now we're in the middle of a war."

The Iran conflict that began in late February sent gas prices sharply higher. Jet-fuel costs are up more than 70% since hostilities began, enough to push Spirit Airlines into failure. Households in the bottom third of the income distribution are spending 7% less than before the conflict started, per the same MarketWatch report.

The divergence is cleaner at the top. Americans with large stock portfolios are, on average, holding up better than their peers, per the Michigan survey data cited by The Wall Street Journal. But even those investors are less optimistic than investors at prior valuation peaks.

When AI replaced the internet

The combined value of semiconductor companies worth more than $10B has surged 26% since February. The Philadelphia Semiconductor Index is up more than 65% so far in 2026.

Michael Burry, who built his reputation predicting the US housing crash, wrote recently that "stocks are not up or down because of jobs or consumer sentiment. They are going straight up because they have been going straight up. On a two letter thesis that everyone thinks they understand. Feeling like the last months of the 1999-2000 bubble."

Paul Tudor Jones has drawn the same parallel. He told CNBC's "Squawk Box" the bull market may have further to run, noting the environment feels similar to 1999, roughly a year before technology shares peaked in early 2000. The key difference from 2000 is the sentiment backdrop.

At the dot-com peak, according to The Wall Street Journal, consumer confidence was at all-time highs. The economy was adding jobs, inflation was low, and the federal government was running a surplus.

Robert Barbera, director of the Center for Financial Economics at Johns Hopkins University, captures what makes today's setup structurally different: "The stock market on the moon and households in increasing gloom are reflecting on the same thing." That thing is artificial intelligence, where productivity gains for corporations and job-market anxiety for workers are two sides of the same outcome.

What the dot-com bust rewarded

When the bubble broke, even the companies that survived fell sharply from their peaks, per the MarketWatch analysis. MicrosoftMSFT

and AppleAAPL

fell 65% and 91%, respectively, while OracleORCL

and Amazon.comAMZN

fell 88% and 94%.

Recovery timelines ranged from five years for Apple to 16 years for Microsoft. The issue was that valuations had priced in outcomes that took a decade or more to materialize, not whether the technology itself worked.

The regime that rewarded investors in the bust years was one of lower valuations, earnings visibility, and assets that generated income independent of growth expectations. The Buffett Indicator, which measures US stock market capitalization against gross domestic product, now sits at roughly 230%, per the same analysis. In 2000, at the prior CAPE peak, that figure was 170%.

Three moves the cycle supports

The historical pattern from 1999 to 2002 supports reducing exposure to the highest-multiple segment of the market this week. The Philadelphia Semiconductor Index run-up Burry flagged mirrors the SOX trajectory in late 1999.

An investor holding semiconductor-heavy allocations through an S&P 500 index fund might examine how much of that fund's weight sits in the largest AI-related names and whether that concentration is intentional. The Invesco S&P 500 Equal Weight ETFRSP

provides US large-cap exposure with less concentration in the names where valuation risk is most acute.

This month, the cycle supports adding allocation to assets with earnings that aren't tied to AI capital expenditure growth. The MarketWatch analysis points to what it calls HALO stocks, meaning companies that build, own, and operate hard physical assets.

A US-listed vehicle with this profile is the iShares US Infrastructure ETFIFRA

, which holds companies whose revenues depend on assets already in the ground rather than on future AI monetization.

Structurally, across this cycle, the prior bust cycle suggests that cash and short-duration instruments matter more when valuations are at CAPE levels above 40. The iShares Short Treasury Bond ETFSHV

gives a beginner a cash-equivalent position that generates income while the cycle resolves, without requiring a view on when the resolution happens. The prior cycle's lesson is that even a correct long-term thesis, as internet connectivity proved to be, doesn't protect against a decade of flat returns from an overvalued entry point.