(Fortune) -- In two tumultuous weeks in October 1987, the stock market shed nearly one-third of its value in perhaps the second most notorious crash in U.S. history. It could happen again. Don't be deceived by the rebounding economy, any more than the bulls should have been misled by the balmy climate during the late Reagan years. Right now, stocks are extremely vulnerable to the same scenario. The reason: The market is even more overpriced than when thunder struck on that distant Black Monday.
That doesn't mean that a giant correction is inevitable; far from it. But the quasi-bubble that followed the big selloff in late 2008 and early 2009 makes the probability of sudden downward swing far more likely. And today's high prices make it practically certain that investors can, at best, expect extremely low returns in the years ahead.
Let's gauge where the market stood just before the shock of October '87. For this analysis, we'll use the price-to-earnings multiples developed by Yale economist Robert Shiller, who smoothes the erratic swings in profits by calculating PEs based on a 10-year average S&P earnings, adjusted for inflation.
In the fall of 1987, stocks were on a tear. The Shiller PE had doubled to 18.3 in just four-and-a-half years. That's far above the historical average of less than 14 (though pales in comparison to recent giant PE numbers). The dividend yield was also an extremely narrow 2.6%, well below the norm of around 4.5%. That's important to remember, because over long periods, dividends are by far the biggest source of investor returns from equities. Then the bottom fell out of the stock market.
When the carnage ended, the PE had dropped to 13.3, around its 60-year average, and the dividend yield was approaching 4%. In other words, the fall hardly made stocks a bargain. But with brutal speed and efficiency, it chopped valuations to levels that made sense.
The slow crash of 2008-2009: a temporary return to sanity
From early 2008 until March of 2009, stocks suffered a steep decline that echoed the 1987 correction, though over many months versus a few days. During that period, PEs fell from the low 20s to 13.3, precisely their level following the '87 crash. Once again, dividend yields approached 4%. For a few brief weeks, it appeared that sanity had returned to the capital markets. Suddenly, equities--overpriced for 20 years-- looked like a good buy, just as they had in at the close of 1987.
But the rational interlude didn't last. By early May of this year, the PE had soared back to almost 22, shrinking the dividend yield to 1.8%. In recent days, stocks have dropped over 7% amid giant gyrations. That decline isn't nearly enough to restore equities to anything like their fair value. So what do the current, extremely price PEs tell us about the future of stock prices?
The best bet is always that equity values, like most investments, "revert to the mean." So following a bubble or bubblette, PEs and dividend yield will inevitably fall back to their historic averages. The big question is how long it will take. In 1987, it took a few days. In the 1990s and early 2000s, stocks remained overpriced for years.
The problem is that the returns investors can garner at today's lofty valuations are just too low. They're also too low to last. If PEs stay constant, the total return to investors is the dividend yield plus the growth in earnings per share, with an added bump for inflation.
Today, the yield is 1.8%. Earnings per share typically grow at a "real" rate of around 1.5%, way below the pace that Wall Street advertises. So the total gain from investing at today's prices is 3.3% or so, plus around 2.5% for inflation, for a total of between 5.5% and 6%.
That's not enough. Stocks are far too risky for investors to accept those puny rewards, as the wild swings of the last two years, and last two weeks, make abundantly obvious. The only way for future returns to rise is for PEs to fall.
How far must prices fall to get back to basics? For the S&P to return to a PE of around 14, the index would need to drop by around 33% to less than 800, its range in early 2009. That would substantially raise dividend yields, and raise future real returns into the high single digits, where they belong.
The drop is going to happen. Here's how:
Here's how I see the odds. The chances are about one in three that we suffer a huge, wrenching correction in the next year or two similar to the one in 1987. That possibility is so high because stocks are so startlingly expensive. Another high probability event is that markets go on a long sideways grind, with smaller drops along the way. What's extremely unlikely is that the market rises substantially from current levels and stays there for any extended period.
Whatever happens in the next couple of years, the odds are overwhelming that investors who buy stocks today will reap puny returns for 10 years. For example, if you'd purchased shares at today's PE of 22 in early 2003, you would have gotten a return of around 3% a year, barely enough to compensate for inflation, let alone buy the blood pressure medication you'd need to survive the scary ride of stock ownership.
Now let's look out a decade or two. The evidence is extremely strong that price matters, and matters a lot: except in rare cases, buying stocks when they are pricey -- when the Shiller PE exceeds 20 -- leads to puny returns ten years later.
Not that you'd ever know that from the happy talk from Wall Street. So screen the noise out, and follow the numbers. They'll eventually get better for investors. But to get back there, we may revisit October of 1987.