On the final trading day of May 2026, the S&P 500 closed above 7,365 for the first time in its history. The milestone, confirmed through daily closing data published by FRED (the Federal Reserve Bank of St. Louis, sourcing S&P Dow Jones Indices LLC), puts the most widely followed stock index on the planet at a level that would have seemed fantastical just three years ago. But the number that should command more attention is buried in a Yale economist’s spreadsheet. The cyclically adjusted price-to-earnings ratio, known as CAPE, sits at approximately 39. The last time it stayed this high on a sustained basis was March 2000, weeks before the dot-com crash began erasing nearly half the index’s value.
For the tens of millions of Americans whose 401(k) balances, IRAs, and brokerage accounts rise and fall with this index, the collision of a record price and a historically extreme valuation ratio is not abstract. It raises a pointed question: is a CAPE of 39 a warning, or has the market evolved enough to justify the price tag?
Where the numbers come from
The CAPE figure of approximately 39 is drawn from the monthly dataset maintained by Nobel laureate Robert Shiller at Yale. His ie_data.xls spreadsheet is the primary upstream source for CAPE calculations, providing columns for monthly price, earnings, and the cyclically adjusted ratio itself. That same dataset shows CAPE reaching roughly 44 in March 2000. Within months of that peak, the S&P 500 began a decline from about 1,527 to roughly 776 by October 2002, a loss of approximately 49% over two and a half years.
The intellectual foundation for this valuation gauge traces to a 1988 paper by John Y. Campbell and Shiller titled “Stock Prices, Earnings, and Expected Dividends,” available through Harvard’s DASH repository. That work, also circulated as NBER working paper w2511, argued that averaging inflation-adjusted earnings over a long window (eventually standardized at ten years) provides a more reliable lens for judging whether stock prices reflect underlying fundamentals than a single year’s profits. The CAPE ratio investors watch today grew directly out of that research.
Why this time feels different, and why that argument has limits
A CAPE near 39 tells investors that the S&P 500’s price is elevated relative to its ten-year average of inflation-adjusted earnings. What it does not tell them is when, or even whether, a correction follows. The ratio has a mixed record as a timing tool. In March 2000, the crash arrived within weeks. In late 2021, CAPE climbed to a similar neighborhood near 39 before a sharp drawdown in 2022, yet the market recovered and pushed to new highs by 2024. Investors who sold on the signal alone missed the rebound.
The strongest bull case for why today’s CAPE may not carry the same omen centers on the composition of the index itself. The S&P 500 of 2026 is far more concentrated in technology and platform companies than the index of 2000. The five largest stocks now account for a share of total market capitalization that dwarfs anything seen during the dot-com era. These firms, many generating substantial revenue from artificial intelligence products and cloud infrastructure, carry higher profit margins and more intangible assets than the industrial and financial companies that once dominated the benchmark. If those margins prove durable, a higher baseline CAPE could be structurally justified.
That argument, though, rests on assumptions that deserve scrutiny. Elevated margins can compress as quickly as they expanded, particularly if regulatory scrutiny, antitrust enforcement, or competitive pressure intensifies. And the interest-rate backdrop matters enormously. In March 2000, the 10-year Treasury yield sat near 6%. As of recent Treasury auction data in spring 2026, yields remain above 4%, meaning stocks face real competition from bonds for investor capital. A CAPE of 39 when risk-free rates are elevated leaves less room for error than the same ratio in the near-zero-rate environment of 2020 and 2021.
Much of the current rally rests on the assumption that AI will deliver a sustained acceleration in corporate earnings. Yet no widely cited institutional research, from bodies like the NBER or major academic journals, has directly addressed how AI-driven earnings growth might alter the historical relationship between CAPE and future returns. The academic literature has not caught up to the market’s thesis, which means investors are, to some degree, pricing in a productivity revolution that remains unproven at scale.
What CAPE actually predicts, and what it does not
Shiller’s dataset does not include a column labeled “sell signal.” The 1988 Campbell-Shiller paper frames the relationship between valuation ratios and future returns in probabilistic terms, not as a binary alarm. Historically, buying the S&P 500 when CAPE has exceeded 35 has been associated with below-average returns over the following decade. But “below average” is not the same as negative. Investors who exited U.S. stocks entirely in late 2017, when CAPE first crossed 30, missed years of strong gains before any meaningful pullback materialized.
It is also worth noting what the official sources do not say. There is no statement from S&P Dow Jones Indices or the Federal Reserve on the significance of the 7,365 close. The FRED database provides raw data without commentary. Shiller’s spreadsheet offers the ratio without forward-looking interpretation. Any specific return forecast tied to a CAPE level is interpretive, not deterministic, and anyone claiming certainty about what happens next is working beyond what the data supports.
What this means for people with money in the market
For anyone managing a retirement portfolio or making allocation decisions in June 2026, the practical takeaway is narrower than the headline might suggest. A CAPE of 39 does not predict a crash on any specific timeline. It does suggest that the margin of safety embedded in current prices is thin by historical standards, and that expected long-term returns from this starting point are likely lower than the market’s long-run average of roughly 10% nominal per year.
The risk is not evenly distributed. An investor in their 30s with decades of accumulation ahead can absorb a drawdown and continue buying at lower prices. Someone within five years of retirement, or already drawing down savings, faces a fundamentally different calculus. A sharp decline from these levels would compress the window for recovery, and sequence-of-returns risk, the danger that early losses permanently impair a portfolio being withdrawn from, becomes the dominant concern.
One comparison adds perspective: developed international markets, as measured by indexes like the MSCI EAFE, currently trade at CAPE ratios well below the S&P 500’s level. That gap does not guarantee international stocks will outperform, but it does highlight that the valuation stretch is concentrated in U.S. large caps, not spread evenly across global equities.
A reasonable first step for anyone reassessing their exposure is mechanical, not speculative: compare your current stock allocation against your target and rebalance if the rally has pushed equities well above their intended weight. That is not a market call. It is the same discipline that applies whenever one asset class outperforms for an extended stretch.
Why a CAPE of 39 warrants attention even without predicting the next crash
At its core, the CAPE ratio is a tool for calibrating expectations about future returns, not for timing the next week or month. At 39, it calibrates those expectations lower than usual. The last comparable sustained reading preceded one of the most painful bear markets in modern memory. That does not make a repeat inevitable. But it does mean that the penalty for complacency, should earnings disappoint or rates stay elevated, would be steeper from here than from almost any other starting point in the past quarter century. The price of the index is a fact. The question is whether the earnings will eventually justify it.
