Does Cyclically-Adjusted Price-to-Earnings ratio (CAPE, or P/E10) usefully predict stock portfolio returns? In their October 2017 paper entitled “The Many Colours of CAPE”, Farouk Jivraj and Robert Shiller examine validity and usefulness of CAPE in three ways: (1) comparing predictive accuracies of CAPE at different horizons to those of seven competing valuation metrics (ratios of an income proxy or book value to price); (2) exploring alternative constructions of CAPE based on different firm earnings proxies; and, (3) assessing practical uses of CAPE for asset allocation and relative valuation (supporting rotation among asset classes, countries, sectors or individual stocks). They employ a total return CAPE, assuming reinvestment of all dividends. For forward testing, they lag earnings and related data to ensure real time availability for investment decisions. Using quarterly and annual U.S. stock market data from Shiller since the first quarter (Q1) 1871 dovetailed with end-of-quarter data since Q4 1927, and data as available for other valuation metrics, all through the Q2 2017, they find that:
- CAPE is generally a better market return predictor than competing valuation ratios. For example, CAPE out-predicts three other available ratios at horizons of 1-6 years and is competitive with other ratios at horizons of 7-10 years during Q3 1930 through Q2 2017.
- Constructing CAPE with operating earnings or as-released National Income and Production Account (NIPA) profits rather than as-reported earnings has only minor effects on CAPE in recent decades.
- Regarding usefulness of CAPE:
- CAPE is not useful for market timing because it has no stable average. In other words, whether CAPE is high or low relative to its historical average is not exploitable.
- CAPE predictions of the equity risk premium (for asset class allocation) are sensitive both to forecast horizon and to calibration lookback interval, introducing considerable uncertainty in application.
- Applying CAPE to compare country equity markets raises concerns about economic differences (such as sector concentrations and dissimilarities), accounting differences and currency exchange rates.
- During September 2002 through August 2017, a relative valuation strategy based on standardized U.S. equity sectors CAPEs, augmented by a conventional momentum filter, beats the S&P 500 Total Return Index by an average gross annual 3.4%, with gross annualized Sharpe ratio 0.70 versus 0.49 for the index.
- During September 2002 through August 2017, a relative valuation strategy based on CAPEs for certain U.S. stocks with long histories, augmented by a conventional momentum filter and adjusted for sector neutrality, beats the S&P 500 Total Return Index by an average gross annual 1.6%, with gross annualized Sharpe ratio 0.60 versus 0.52 for the index.
In summary, evidence indicates that CAPE is among the best of simple valuation ratios and may have some use for sector rotation and stock selection.
Cautions regarding findings include:
- Samples are generally short to very short in terms of number of independent 10-year measurement intervals.
- Testing multiple valuation ratios, many prediction horizons and several calibration lookback intervals introduces data snooping bias, such that best-performing scenarios overstate expectations.
- The “examples” of applying CAPE to asset class allocation and country equity market selection are discussions only. They present no strategy tests.
- The example of applying CAPE for U.S. equity sector selection involves a short test that includes a momentum filter, such that the CAPE effect is unclear. Strategy construction may impound snooping bias. Portfolio reformation is apparently monthly and results are gross of trading frictions and tax implications. Also the benchmark for the test should arguably be a portfolio of the sectors considered rather than the S&P 500 Total Return Index.
- The example of applying CAPE for U.S. stock selection involves a short test applied to a highly restricted stock sample (minimum price history of a few decades) with a momentum filter and sector neutrality, such that the CAPE effect is unclear. Strategy construction may impound snooping bias. Portfolio reformation is apparently monthly and results are gross of trading frictions and tax implications. Also the benchmark for the test should arguably be a portfolio of screened stocks and not the S&P 500 Total Return Index.
For related research see results of this search, especially “Usefulness of P/E10 as Stock Market Return Predictor” and “Alternative Tests of P/E10 Usefulness”.
