A biblical legend in the Epistle of James provides an allegory for today's stock market. In that moral fable, a merchant sends his servant to the Baghdad market for provisions. "Death," the narrator, is roaming the stalls as well, and jostles the servant. Death recoils and throws up his arms as if to strike, terrifying the servant, who flees to borrow the merchant's horse for a quick getaway. "I will ride away from the city to avoid my fate," he tells the merchant, then gallops for Samarra. Later that day, the merchant sees Death in the marketplace. “Why did you threaten my servant?” demands the merchant. "That wasn't a threatening gesture, it was a gesture of surprise!" Death replies. "I was surprised to see him in the city, because I had an appointment with him tonight in Samarra."

Big-cap stocks may well face their own appointment in Samarra. That's the menacing signal coming from the famous cyclically adjusted price/earnings ratio, or CAPE, developed by Yale economist Robert Shiller. The metric is favored by many of the best minds in the markets, from Research Affiliates’ Rob Arnott to AQR Capital’s Cliff Asness.

The CAPE is a much more reliable measure of when equities are cheap or superexpensive than using the standard P/E. That's because earnings are extremely volatile, so that a sharp drop like the one we witnessed last year makes multiples look inflated, while a temporary, unsustainable surge gives the illusion that the S&P 500 is a bargain, the situation in late 2019. Instead, Shiller averages profits over the previous decade to smooth the peaks and valleys.

As of midday on April 8, as the S&P hit a record 4094, the CAPE stood at 36.8. It's been that high during only one period in the 139-year history tracked by Shiller. Even at the verge of the 1929 crash, it hovered below 33. That CAPE-at-37 span bracketed the interlude when prices were most inflated during the turn-of-the-century Internet bubble. Exactly 23 years ago, in April 1998, the CAPE hit 37.4, a record high at the time. So here's the question that should be foremost in investors' minds: Where did prices go from there, and what does rescaling that 1998 milestone tell us about where they'll go from here?

What's interesting about the dotcom craze is that even when the CAPE started signaling danger, stocks roared for a long time. Over the 28 months starting when the CAPE nearly hit 37, the S&P jumped from 1112 to the frenzy's peak of 1485, reached in August 2000. Hence, the index gained 33.5% after notching a multiple that in retrospect should have petrified investors.

As it turned out, the 37 CAPE in April 1998 already foretold a steep fall in prices. The higher stocks climbed from there over the next two-plus years, the sharper the eventual fall ahead was likely to be. A wipeout ensued. By September 2001, the index had made a roundtrip, dropping below the 1112 mark when the CAPE first reached 37. The descent didn't stop there. At the low point in February 2002, prices sank 23% under the benchmark in May 1998, and they didn't get back to even until August 2004. So for over six years from the month the CAPE hit 37, the S&P delivered zero capital gains. Investors did collect 1.4% annually from dividends, a return that didn't even match inflation.

How did shareholders fare over the full decade that started with a CAPE of 37? By April 2008, the S&P had returned 2.0% a year in price appreciation, and another 1.4% in dividends, for a paltry 3.4%. If you had bought back in the spring of 1998 and never sold, as of today you'd have 6% in annual capital gains. Add dividends, and the total return would be an unimpressive 7.4% a year.

But that 7.4% number is probably illusory, if you're planning to hold. In our scenario, in which you bought into the index in 1998, you would only have achieved even those mediocre, single-digit returns because the S&P is once again highly inflated. Today's CAPE of over 37 is forecasting a big fall, just as it was almost a quarter of a century ago.

So should you sell now?

That's a tougher question than it appears. The truth is, these fevers can last a long time and send shares from extremely to wildly overpriced. We saw that phenomenon in the Internet boom, when the CAPE started at 37, and the S&P proceeded to climb over 30%. It could happen again. The problem is that when the CAPE shows that prices are extra rich, they usually don't keep getting richer for an extended period. The longevity of the dotcom explosion was the exception. In May 2008, with the CAPE looking pricey in the mid-20s, stocks dropped in six months by over one-third. Message: If you think you can time the market and predict when already overpriced stocks will eventually peak, you risk getting caught in the deluge.

It impossible to say when the deluge will come. But one thing is certain: The grim reaper is waiting in Samarra.