The Problem with the Shiller CAPE

Shiller’s “CAPE” Ratio Has Been in “Overvalued” in 24 of the Last 25 Years

There has been no dearth of articles in the business press about how overvalued the market is today. Many of these articles cite the Shiller Cyclically Adjusted P/E ratio (CAPE) as evidence of the overvaluation.  This article looks a little deeper:

Since the historic mean of Shiller’s CAPE is 16.58, last week’s CAPE ratio of 27.52 is a whopping 66% above the mean. However, we’re still far below the all-time high CAPE ratio of 47.2, set intraday on Friday, March 24, 2000 – a peak day for the S&P 500, right before the dot-com bubble burst in April.

The second-highest cyclical P/E peak came in September 1929 at 32.6. The market fell nearly 90% from there, so today’s 27.52 reading seems risky. But times have changed. Since the Berlin Wall fell in 1989, earnings have sustained a higher average plateau. As a result, CAPE has traded above “average” for 97% of the last 25 years, including three long, strong bull markets in 1990-2000, 2003-07, and 2009-15.

Last July, economist and Reuters columnist Anatole Kaletsky dissected CAPE in The New York Times:

“Since 1989, the S&P 500 has multiplied eightfold, while total returns, including dividends, have increased the value of an average equity investment 12-fold. But investors who followed Mr. Shiller’s methodology would have missed out on almost all these gains, since the Shiller P/E showed the stock market to be overvalued 97% of the time during this 25 years. And even during the two brief periods when the Shiller P/E was below its long-term average, in early 1990 and from November 2008 to April 2009, it never sent a clear buy signal. Instead, Mr. Shiller’s approach suggested that the valuations in 1990 and 2009 were only just below fair value, implying there was very limited upside at the beginning of bull markets in which prices multiplied fivefold from 1990 to 2000 and threefold from 2009 to 2014 (so far).

“The Shiller P/E’s predictive performance would have been just as bad in earlier decades. During the equity bull market of the 1950s and 1960s, for example, the Shiller P/E would have shown Wall Street to be overvalued for 96% of the 19-year period from early 1955 to late 1973. Only in January 1974 did the Shiller P/E move below what was then its long-run average, implying that it might finally be a ‘safe’ time to buy stocks. Straight after the Shiller P/E sent this first ‘buy signal’ in almost two decades, Wall Street crashed by 40% in 12 months.”

In addition, Deutsche Bank’s Stuart Kirk wrote, “The trouble is that the last two times the CAPE crossed 25 (in 1996 and 2003) the bull market was just getting started and ran for another four years.” In the first example, the S&P 500 went from 614 in December 1995 to 1,485 in August 2000 for a 141% rally. In the second example, the S&P 500 went from 1,038 in September 2003 to 1,520 in 2007 for a 46% gain.