Practical analysis for investment professionals
19 November 2012

Asness: Shiller P/E is Signaling Risk of Another Lost Decade for US Stocks

In Security Analysis, Benjamin Graham and David Dodd first proposed that investors examine an average of earnings over “not less than five years, preferably seven or ten years.” By smoothing out the effect of the business cycle on corporate earnings, investors would get a truer picture of how expensively or cheaply stocks are priced. This advice was later popularized by Yale professor Robert Shiller, and the Shiller P/E ratio, alternatively known as the cyclically-adjusted P/E (CAPE) ratio, has become a widely followed and efficacious stock market valuation measure.

With the Shiller P/E for the S&P 500 Index currently standing at a 21.4 (approximately 30% higher than its long-term average), many value investors, including Cliff Asness of AQR Capital Management, have adopted a cautious stance toward US stocks. In “An Old Friend: The Stock Market’s Shiller P/E,” Asness presents his most recent analysis of the Shiller P/E ratio, which suggests that the next decade could very well be another period of subpar returns for equity investors.


Shiller P/E Ration for the S&P 500


The Shiller P/E divides the inflation-adjusted price level of the S&P 500 Index by its average real reported earnings over the prior 10 years. It traces its roots to a 1988 paper by Shiller and Harvard professor John Y. Campbell titled “Stock Prices, Earnings and Expected Dividends.” A key presupposition of the Shiller P/E is that profits are mean-reverting and that by eliminating earnings peaks and troughs, investors are left with a more stable valuation measure. For this reason, Asness considers the Shiller P/E to be a superior, if imperfect, alternative to focusing on one year of trailing earnings.

Asness notes that the Shiller P/E is currently higher than 80% of the time since 1926. He examined S&P 500 Index real returns over rolling decades since 1926 and sorted the data into deciles by starting Shiller P/E ratios. Asness found that at current valuation levels, the average expected real stock market return over the next decade is a sobering 0.9% per year. The worst-case scenario is an annualized real return of −4.4% while the best case is a real return of 8.3%. While it has correctly signaled most major market tops over the past century, Asness concedes that the Shiller P/E has also been shown to be wrong for extended periods of time, which is why he considers his analysis most useful not as a trading tool but for setting investor expectations about future stock market returns.

Critics of the Shiller P/E, including Wharton professor Jeremy Siegel, often point to the two severe earnings recessions of the last decade as an aberration that unduly distorts the ratio. Asness responds by noting that 10-year real earnings used in the Shiller P/E calculation are actually above their long-term trend. Further, he argues that the huge collapse in earnings was due, at least in part, to earnings being “borrowed from the future.”

Siegel, author of Stocks for the Long Run, prefers to focus on prospective earnings or trailing earnings, and, perhaps not surprisingly, is widely recognized as being a long-term bull on equities. In a recent article titled “The Case for Dow 17,000,” he revisits an earlier prediction that US stocks could rise over 30% by the end of 2013. And despite a sluggish domestic economy, the approaching fiscal cliff, and recession in Europe, he remains optimistic with this call. Siegel notes that, historically, extended periods of market weakness are followed by market strength. He adds:

“Specifically, when stock returns for the preceding five years fall in the bottom 25% of all five-year returns, as did the returns from 2007 through 2011, the annualized return on stocks in the following two-year period is 20% — more than double the market’s long-term return of 9.9% annualized. Moreover, the market achieves annual gains of 10% or more about 70% of the time.”

Siegel, who will be speaking at the upcoming CFA Institute Equity Research and Valuation Conference, thinks the current market P/E multiple on both trailing one-year and projected 2013 earnings positions US stocks as attractive not only on an absolute basis but also relative to fixed-income alternatives. Of course, experience tells us that prospective earnings are often a moving target. And of late, estimates have been moving south.

Asness’ work confirms that the Shiller P/E is not infallible — no valuation measure is — but it also makes clear that to ignore it when it is at extreme levels is usually a mistake.


Please note that the content of this site should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute.

Photo credit: ©iStockphoto.com/bunhill

About the Author(s)
David Larrabee, CFA

David Larrabee, CFA, was director of member and corporate products at CFA Institute and served as the subject matter expert in portfolio management and equity investments. Previously, he spent two decades in the asset management industry as a portfolio manager and analyst. He holds a BA in economics from Colgate University and an MBA in finance from Fordham University. Topical Expertise: Equity Investments · Portfolio Management

3 thoughts on “Asness: Shiller P/E is Signaling Risk of Another Lost Decade for US Stocks”

  1. Gregg Barrett says:

    Thanks for the note David.

    Recommendation for sell out attendance – The CFA Institute Equity Research and Valuation Conference should have Bill Gross speaking alongside Siegel.

    Re Asness and the Shiller yardstick – the work by Mark Spitznagel at Universa should also be considered. You will be able to find it on the Universa website, with a more comprehensive outline forthcoming in his book “The Dao of Capital” to be released on March 2013.

    Gregg Barrett

  2. john bailey says:

    You can get better results if you use trailing peak earnings rather than a long moving average. This smooths out the trough problems.

  3. Pfizer says:

    Investors examine an average of earnings over not less than five years peak earnings Stocks.

Leave a Reply

Your email address will not be published. Required fields are marked *



By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.

Close