Is Shiller's CAPE as Scary as it Seems?

In the 1990’s, economists Robert Shiller and John Campbell created a valuation metric called the “cyclically adjusted price-earnings” ratio, or CAPE. A Wall Street Journal article from earlier this month examines whether this metric might be sending a false signal that the market is overheated.

The CAPE ratio values shares based on 10 years rather than one year of earnings which, the article explains, “smooths out periods like just prior to the housing bust, when unusually strong earnings made stocks look reasonably priced, and post-recession recoveries, when weak earnings make stocks look expensive.” The CAPE is now at 27, about where it was before the financial crisis in 2007 and well above its 50-year average of 20. As per data compiled by Mr. Shiller, the article cites, it has been higher only “during the 2000 bubble and bust, and just prior to the 1929 crash.”

However, the WSJ article suggests the possibility that Shiller’s data might be causing more concern than necessary. The argument is centered on the integrity of the earnings figures used in calculating the metric. In short, Shiller’s version uses S&P 500 earnings under generally accepted accounting principles (GAAP), intended to avoid some of the manipulations companies can use to polish their numbers. The issue, according to Wharton professor Jeremy Siegel, is that accounting-rule changes have pushed recent earnings lower, which has made the CAPE artificially high.

The Wall Street Journal constructed an alternative CAPE that uses the same methodology as Mr. Shiller but which uses quarterly earnings data compiled by the U.S. Commerce Department and stock prices based on Federal Reserve data (rather than the S&P). These figures are much more consistent because they are not prone to shifts in accounting rules and standards. Using this alternative metric, the article says, stocks “look much cheaper than Mr. Shiller’s data suggests.”