Why U.S. stocks are too pricey to be a great investment by Shawn Tully @FortuneMagazine July 24, 2015, 5:30 AM EDT E-mail Tweet Facebook Google Plus Linkedin Share icons In May, Janet Yellen sounded a warning on stock prices. During an address at a New York City think-tank, the Fed chairman declared that “equity market valuations at this point are generally quite high…there are potential dangers there.” It was an extraordinary statement from the generally ultra-cautious central banker. Yet it also reflected what sophisticated, academic-minded observers — as opposed to the Wall Street cheerleaders — now conclude from the wide variety of convincing, time-tested metrics designed to measure whether stocks are undervalued, or overpriced. Those tools use different methodologies, but they reach the same general conclusion: stocks are indeed extremely expensive. They even agree on the returns investors can expect when they purchase shares at these elevated levels. So if you’re counting on 7% average annual gains, after inflation, that US equities have delivered over the past 125 years, you’ll probably be sorely disappointed. Stocks may still be the “place to be,” at least compared to US bonds, but prudent investors should obey the following maxim implied in Yellen’s admonition, “lower your expectations!” In a recent article entitled “Are Stocks Overvalued? A Survey of Valuation Models,” Chris Brightman — chief investment officer at Research Affiliates — shows how the different methodologies point to remarkably similar outcomes. (Research Affiliates is a research firm that oversees investment strategies for $170 billion in mutual funds and ETFs). Brightman explores two basic types of tools. He calls the first category “Absolute Valuation Models.” These models make a major assumption: That PE ratios, and dividend yields, remain the same over time. In other words, investors put the correct, enduring price on each dollar of earnings, and that metric, and the judgment behind it, doesn’t change. He explores two types of “Absolute” models. Model One is a great Brightman favorite. It forecasts future, long-term returns as the average of the dividend yield, and the earnings yield. The latter is simply the inverse of the PE ratio. This methodology has the advantage of incorporating what analysts usually miss, that earnings-per-share growth is constantly diluted by the issuance of new shares. That dilution has three sources: IPOs that launch fresh competitors, shares issued to buy companies in the same business — think of Cisco CSCO purchasing sundry startups to capture new technologies, and options, as well as restricted shares, awarded to managers. Because the flood of new shares in most industries far exceeds buybacks, EPE waxes well behind growth in overall profits in the economy, a fact investment strategists seldom mention. Today, the S&P 500 dividend yield is 2%, and the earnings yield is 5.4%. You get that earnings yield by flipping today’s PE ratio 18.5 into an earnings-to-price ratio. The average of the two, and the returns Model 1 forecasts for US stocks, is 3.7%, plus inflation, for a total of around 5.7%. Brightman cautions that this methodology, and all the good models, tell you virtually nothing about what will happen in the next year or two, unless by near-total coincidence. But they all provide extremely reliable predictions over a 10-year horizon. In the Absolute group, the second entrant is the “dividend growth model.” It states that the total, future return is the sum of the dividend yield, and the growth in earnings. Over the past century, profits-per-share have been advancing about 1.5% annually, in real terms. That number is so low for precisely the reason we just discussed, heavy dilution, especially the creation and expansion of businesses in the same business, whether it’s software or airlines. So the total real return from the dividend discount model amounts to 3.5%, the 2% dividend yield plus the 1.5% expected growth in EPS. Conclusion: The two Absolute models pretty much converge in the mid-3% range. The second category consists of “Relative Valuation Models.” This group is all about going from abnormal to normal, the gravitational pull that tends to lower or lift lofty or flagging valuations to their historical benchmarks. The most famous is the CAPE, or “cyclically adjusted price-earnings” ratio, developed by Nobel laureate Robert Shilller. Rather than using today’s earnings in the PE ratio, Shiller deploys an average of the past 10 years of profits, adjusted for inflation. That methodology smoothes the numbers so that valuations no longer looks inflated when earnings are depressed, or reasonable when profits are unusually, unsustainably high — more or less the scenario today. At over 28, the CAPE is around 10 points higher than the long-term average. If the CAPE retuned to a more “normal” level of around 18, prices would need to fall some 36%. Brightman doesn’t think that will happen. He recognizes that the CAPE should be adjusted to reflect economic circumstances. Investors find interludes of moderate interest rates and inflation soothing, and hence reward stocks with elevated valuations. That’s the climate today. Still, stock prices relative to earnings, measured by the CAPE, are so extraordinarily high that even favorable rates and inflation won’t likely deliver big returns. Put simply, stocks are just too pricey to be a great investment. Today, the S&P 500 is in the highest CAPE quintile, measured over the past 125 years, and rates in the second lowest. That combination produces annual, real returns of 3% over the following ten years. So the adjusted CAPE portends gains even lower than those from the absolute models. Brightman likes to combine both categories of models to reach a final forecast. He reckons that PE ratios are just too high, and that a snapback to more normal levels will dampen returns for years to come. Hence, stock prices won’t rise as fast as the sum of profit growth and dividend yields, simply because contracting PEs will tug in the opposite direction. Research Affiliates is predicting real gains of less than 1% a year over the next decade. “Are stocks overvalued in the US market? The answer is a resounding ‘Yes!’” Brightman concludes. Whether or not that number seems impossible, it’s best to view equities through his lens, as a world of diminished expectations.