The S&P 500 (^GSPC 1.27%) is in rarified air. After climbing for three straight years, the benchmark stock index is trading near an all-time high as we kick off 2026. Investor optimism surrounding artificial intelligence (AI) has been the driving force behind the bull market, with the tech-heavy Nasdaq Composite (^IXIC 1.50%) outperforming the broader S&P 500.
While the market has faced several hiccups during its current bull run, including a massive sell-off in April after President Donald Trump announced sweeping tariffs on imports, both indexes have quickly brushed them aside to continue moving higher. In fact, the S&P 500 has climbed so much so quickly that it just did something seen only twice in the last 45 years. Both prior occurrences had clear market reactions, including one of the most memorable decades in stock market history.
Image source: Getty Images.
A historical run-up in stock prices
Three years into the current bull market, investors have bid up the prices of stocks at a rate much faster than the underlying fundamentals of most S&P 500 companies have improved. Moreover, the fastest-growing businesses also happen to be some of the largest companies in the index already, which carry more weight. As a result, the S&P 500's forward price-to-earnings ratio (P/E) has climbed from about 15 when the market bottomed in 2022 to over 23 just over three years later.
That's only the second time the S&P 500's valuation has climbed to that level in the last 45 years. The most recent was in 2020 as the markets recovered from the shock of COVID-19, and it previously reached that level (and higher) in the run-up of the dot-com bubble.
The high stock valuations in the market today are another manner in which the current excitement and optimism around artificial intelligence echoes what we saw in the 1990s with internet and telecom companies. Many have compared the way big tech companies are spending so much to build the necessary infrastructure to power their artificial intelligence models to the way telecom companies spent copious amounts to build out internet infrastructure 30 years ago. While there are promises of big returns, many companies have yet to show meaningful earnings growth from AI innovations.
SNPINDEX: ^GSPC
Key Data Points
That has led many to look to history to project what the next decade could bring in terms of market returns -- and history doesn't paint a bright picture. Every single historical data point indicates that when the S&P 500 forward P/E ratio exceeds 23, the subsequent 10-year returns for the S&P 500 are negative. In other words, investors can expect their portfolios to decline in value over the next decade if they're fully invested in S&P 500 stocks, based on the historical data.
Many smart people are suggesting that the next decade could be rough for investors, particularly those focused on U.S. stocks. Economist Robert Shiller projects negative real returns for U.S. stocks over the next decade, as well, based on his CAPE model. Vanguard's analysts project returns that just barely beat inflation for U.S. stocks.
However, before you sell everything and run for the hills, there's a big grain of salt that readers need to consider about what history can tell us.
The bigger historical trend to consider
The big problem with all the economists and stock market analysts warning investors that the stock market is set to collapse, based on historical data, is that there's a very limited sample size.
As mentioned, the only other times the S&P 500 forward P/E topped 23 were in the dot-com bubble and in late 2020. It's only been five years since the 2020 instance, so we only have a handful of overlapping 10-year periods from the 2000s to examine. As it turns out, that decade was mired by a couple of uncorrelated events leading to what's known as "the lost decade."
The dot-com bubble reached its peak in March 2000. A decade later, the S&P 500 was down about 20% and the Nasdaq Composite was down over 50%. However, stocks had steadily recovered until the financial crisis in 2008. Importantly, the financial crisis that led to the Great Recession had absolutely nothing to do with the dot-com bubble that formed in the '90s and popped at the start of the decade.
Therefore, focusing on that period of returns to draw historical relevance isn't the best use of data. A small sample size can distort a lot of projections, and that's exactly the issue here.
In statistics and probability, there's a concept called the "base rate." It's the probability of something before factoring in additional evidence. Projections that overly rely on specific pieces of evidence can be guilty of base-rate neglect, leading to wildly inaccurate projections.
If you look at the average rolling 10-year returns for the S&P 500 throughout its history, you can get a pretty good idea of what the "base rate" is for the stock market over the next 10 years. The average annualized 10-year total return for the index is 10.6% over the last 100 years.
While the current market valuation may result in lower total returns over the next decade than the historical average, it's likely investors will see returns closer to the base rate than the single sample decade we have from the last time valuations were this high. Investors should remain mindful of valuations for both the overall index and individual stocks, but abandoning stocks altogether would likely be detrimental to their portfolios over the long run.
