Observers have been citing the high level of Shiller’s P/E (above 26) as reason to predict a drop in U.S. equities. I highlight below some of the details of its calculation that impact how it should and should not be used.

First off, Shiller’s cyclically-adjusted P/E (“CAPE”) is often misused as a predictive factor of near-term (say, 6-to-12 month) equity index returns. In his book Irrational Exuberance, Prof. Shiller did not suggest that it be used this way. He linked CAPE to subsequent 20-year average equity market return, not to near-term return. Starting today, 20-year average return will be known only in 20 years. Thus, this measure can be used ex-post (to explain the past), and perhaps ex-ante – but only over very long horizons.

Sources: Robert Shiller via irrationalexhiberance.com, Standard & Poor’s, and MCM’s calculations

Another of it’s weaknesses, in my view, is its severe lag compared with the most recent annual earnings. At this time, 2005 earnings have the same weight in 10-Year CAPE as TTM earnings. Using a 10-year average tends to understate the “E,” and overstate the P/E ratio in periods of fast real earnings growth – and S&P earnings grew quite rapidly since 1990, albeit interrupted by two deep downturns (see chart).

Our tactical research has instead shown that using “good old” actual TTM earnings, with no adjustments, works best as predictor of near-term equity returns for tactical management. While this isn’t perfect, of course – it’s subject to cyclical ups and downs – it avoids the shortfalls of the CAPE described above. Measured this way, the S&P 500 currently trades at 17.5 times its TTM earnings.

Model Capital Management LLC is a tactical investment manager. Please visit the following web pages for more information on Model Capital’s approach to tactical investment management and our tactical asset allocation models/strategies.