On Friday, June 11, the S&P 500 closed at a record 4,239, having climbed almost 13% this year alone. Because earnings are highly volatile, a temporary spike can make P/E multiples look artificially cheap, while a short-lived fall can make S&P valuations seem inflated. To correct for that problem, Shiller deploys a 10-year average of earnings per share, adjusted for inflation. Using that formula for smoothing the peaks and valleys makes the CAPE ratio a great predictor of the returns shareholders can expect in the years to come. Starting at an ultrahigh CAPE ratio portends poor to mediocre performance five or 10 years hence.

Shiller puts “normalized” earnings at $114 a share. So at the S&P “price” of 4,239 on June 11, the CAPE ratio stood at 37.2. That’s the highest reading since December of 2000, when the index had just begun its descent from the heights of the tech bubble. A half-decade later, the S&P had dropped from 1,331 to 1,262, shedding 5%. Ten years from the last time the CAPE ratio hit 37, in December of 2010, the index closed the month at 1,241. So market-weighted, big-cap portfolios that investors bought at the close of the millennium were worth nearly 7% less at the end of 2010.

It’s necessary to add two caveats before positing that if you buy at today’s elevated prices, you’re likely to live through a repeat of 2000 to 2010. First, these crazes can last a long time. In the dotcom interlude, the CAPE ratio first reached 37 in April of 1998 and, amazingly, jumped as high as 44 in subsequent months. Stocks rose by a third from what looked like the danger zone to gigantic multiples that Wall Street justified by scrapping the traditional valuation tools and claiming we had entered a new era of profitability. Indeed, the CAPE ratio stayed above 37 for well over two years. It’s possible that today’s momentum could keep chugging away, making what’s exorbitantly priced more exorbitantly priced.

Second, if you had bought at a CAPE ratio of 37 in late 2000, and fallen asleep for the ensuing nearly 21 years through mid-June of 2021, you wouldn’t be doing too badly, though not nearly enjoying the double-digit gains the perma-bulls are always predicting. In that period, the S&P tripled, rising 218%. That jump equates to capital gains of around 5.7% a year. Add 1.2% in dividends and your total return would be in the 6.9% range.

The problem is that you would only get that relatively modest return because valuations have recently surged to probably unsustainable heights. So to garner a 7% return over the next decade or two you would have to be betting that the basics no longer count and that big-caps, on average, will keep selling at gigantic multiples in the years ahead, and that earnings per share will far exceed their historic growth rates.

Here’s my view on where things will stand in June of 2031. The Shiller earnings per share figure is somewhat understated because it includes increases for inflation but not for “real” growth that usually lags a bit behind GDP. A rule of thumb is that adding 10% to the Shiller number provides a more realistic estimate of today’s normalized earnings. Bumping up by a tenth lifts Shiller’s $114 in earnings per share to $125. At $125 in earnings per share, the adjusted P/E is not 37 but a bit lower at 34.

If the P/E remains at 34—which is around double the 70-year average—a good estimate of future gains is the inverse of the multiple, or the dollars in earnings investors get for each dollar they pay for their shares. It’s like the yield on bonds, except that you would also get an extra 2% or so for inflation, since companies raise prices for their cars or groceries to match the rise in the consumer price index (CPI). Those producers effectively dictate the CPI. So the absolute best investors can expect is a 2.9% “real” or inflation-adjusted return, plus 2% for inflation, for a total of just under 5%.

They are unlikely to pocket nearly that much. The reason: If history is any guide, the multiple will shift downward to a much more ordinary level. That reversion to the mean will probably erase all “real” gains. Investors will be fortunate to keep up with inflation. It’s tempting to buy into this pricey market because so many other assets, from bonds to real estate, look highly expensive as well. As they say about the oil markets, the remedy for overpricing is overpricing. Eventually, the true believers will no longer believe, and stocks will tumble to once again offer good buys. It happened briefly last spring. A Shiller P/E of 37 is a compelling sign that it will happen again. We just don’t know when.