Why a mean reversion would be mean to the stock market

March 26, 2014, 11:17 PM UTC

FORTUNE — Seldom have so many of the metrics that influence stock prices strayed so far from the long-term trends we’ve come to consider “normal.” Corporate earnings are far above what’s normal historically, interest rates are way below normal because of Fed intervention, and bond prices that wax when rates wane are hovering at seemingly unnatural heights. The typical investor might echo something Yogi Berra could have said: “I’m confused by the ‘new normal’ because it’s so unusual.”

Because so many influential factors are so unusual, investors struggle to determine if stocks are really cheap or at least reasonably priced, or extremely expensive. Wall Street mavens generally argue that equities are a bargain, and urge you to buy more, while usually prescient, quantitative benchmarks show extra-rich valuations, and urge caution. What to believe?

To clear the confusion, let’s compare the likely future equity returns from two scenarios. In the first, we assume that the current metrics will remain in place, so that the “new normal” persists for years to come. In the second, we predict that the metrics that have prevailed during most of the past half-century return in force. That’s called “mean reversion,” the tendency for market rates and ratios to go back to their historic norms, tugged by a kind of gravitational economic force.

To set the framework, it’s important to gauge where earnings, dividends, and prices, the determinants of future returns, sit right now. Then we’ll examine where they’re likely to go in 10 years under the two sets of assumptions. An excellent starting point is the CAPE, or cyclically adjusted price-earnings ratio, developed by 2013 Nobel laureate Robert Shiller of Yale. Instead of using current profits, a highly erratic measure, Shiller employs a 10-year average of inflation-adjusted earnings per share that smoothes the constantly-shifting peaks and valleys — right now we’re at a historic peak. He divides that adjusted earnings number into the current S&P average to arrive at the CAPE.

MORE: Fed: Taxpayer subsidy for “Too Big To Fail” banks? $8.5 billion

Why trust the CAPE? Two of the best minds investing, Rob Arnott of Research Affiliates and Cliff Asness of AQR Capital agree with Shiller that it’s either the best, or one of the best, measures of whether stocks are cheap or dear, and an excellent guide to future returns. Put simply, the lower the CAPE, the better the gains in the decade to come, WI.

Today, the CAPE stands at 25.4. Like almost every market metric out there, the current CAPE is way out of the ordinary — on the high side. The average CAPE over the last two decades is around 18, and for the past century, a couple of points lower.

For our first, “now is our future” assumption, we’ll project that the CAPE stays at its current level for the next decade. That means investors will get an inflation-adjusted return equal to the inverse of the CAPE, or the “earnings yield,” of 3.93%, let’s call it 4%. (If a company’s PE or CAPE is 25, its earnings yield must be 4%.)

The total expected return is that 4% plus inflation of around 2%, or a total of 6%. That return comes in two parts. The first is the dividend yield of around 1.6% a year (large companies today pay out about 40% of their profits), and the second is earnings-per-share growth of 4.4% annually. Keep in mind that earnings per share increase at a far slower rate than overall corporate earnings that over long periods track GDP, because companies typically issue large numbers of shares each year, in excess of buybacks, to fund their plans for expansion.

That’s hardly a wonderful outlook, and it’s a long way from bountiful future Wall Street expects. But the second scenario is far more daunting. It demonstrates the dastardly meanness in mean reversion.

Now we’ll project that the CAPE reverts to recent average of 18 in by 2024. EPS will be 57% higher than today at our 4.4% growth rate, but because the multiple will drop, the stock will post a capital gain of only 8% over that entire period. Dividends, collected each year, will deliver another 16% or so, for a total return of 24%. That’s an annual gain of just 2.15%, a number that should pretty much match inflation, and nothing more.

MORE: Investors lukewarm on King IPO

Our exercise reveals a big surprise. The most important measure in all of equity analysis is actually normal. It’s the equity risk premium, the extra return investors demand over and above the rate on U.S. government bonds — it amounts to the compensation for the additional risk of holding stocks. Today, the ERP is a robust 3.5%. That’s the 4% earnings yield, minus the inflation-adjusted rate on 10-year treasuries of around 0.5%. “We’re in a situation where the level of earnings yield is extraordinarily low, but because of extraordinary monetary policy, the real rates are extraordinarily low,” says Chris Brightman, head of investment management at Research Affiliates, which oversees strategies for more than $160 billion in investment funds. “Subtract the low real rate from the extra-low yield, and you have a perfectly good looking ERP.”

Blame it on the Fed. In the “new normal,” no mean reversion scenario, the total return is the risk premium of 3.5% plus the real yield of 0.5% plus inflation. For that 6% scenario to triumph, the real yield needs to stay puny. Staying puny means more of the same, so that low-yielding bonds will continue providing weak competition even for expensive stocks. Hence, investors will make do with low returns on stocks because the alternatives are so unattractive.

In fact, for real rates to remain that low, the U.S. would need to go into a prolonged period of economic stagnation such as the malaise that’s long inflicted Japan, says Brightman. In that case, it’s highly possible that the mid-single digit returns we forecast could be far too optimistic.

In the second scenario, everything returns to normal. But the risk premium already is normal. What’s far from the mean, and threatens to revert, is the real interest rate. Over the past two decades, inflation-adjusted rates have averaged around 2%. If they return to that level, then our mean, mean reversion scenario is the one that wins. It’s important to point out that the change may not come quickly. But if the U.S. economy shows any kind of vitality, it will arrive eventually. Then returns will barely beat inflation. And 10 years in the wilderness is a long, parched journey.

The abnormally low real rates are the work of the Fed. They cannot last. Once demand for capital supplants money supply creation as the principal force driving rates, as it must, real rates are bound to rise sharply, restoring the trend that began in mid-2013. That’s why the mean-reversion scenario is the most likely, if not inevitable, outcome. One scenario is fair, the other is poor, and the poor one will probably reign. The Wall Street pundits can’t stop thanking the Fed. They should reconsider.