Fortune Teller

Q&A with IFA: Prediction in Valuations

Fortune Teller

Question: Do valuation measures such as price-to-earnings or price-to-dividend ratios help predict future returns? Are they telling us anything now?

Note: The question above was originally addressed in DFA’s Fama/French Forum on December 11th, 2008, in the midst of the Global Financial Crisis. We believe that the response they gave below is completely relevant to today when the S&P 500 Index is near its all-time-high.

“Yes, but not with lots of confidence. The market return tends to be lower when aggregate ratios like E/P and D/P are low, and vice versa. The economic logic is based on the same discount rate effect we use to explain the higher expected return on value stocks. The empirical evidence leans toward a positive relation between aggregate fundamental to price ratios and future market returns, but there is lots of uncertainty about the forecast.”

Note that Fama and French follow the academic convention of flipping price-to-earnings and price-to-dividend upside down. One thing accomplished by this is to make them more readily comparable to bond yields.

While the price-to-earnings ratio was at a low level compared to its historical average at the time the question was asked, today it appears to be above its historical average, and some observers such as 2013 Nobel Laureate Robert Shiller have sounded the alarm, as seen in this video interview on Yahoo’s Daily Ticker. Professor Shiller is well known for his CAPE (cyclically adjusted price-to-earnings ratio), which currently is at a level that has only been reached in 1929, 2000, and 2007, three years that immediately preceded a severe bear market. However, Shiller emphatically advises against using CAPE as a market-timing tool.

Fama and French basically state that while valuations have some degree of predictive power, returns are influenced by so many other factors that it really does not make sense to base an asset allocation on them. This idea finds support in a study done by Vanguard Research in which they evaluated the explanatory power of sixteen commonly cited factors on subsequent stock returns over an 86-year period. The highest scoring of all these factors was Shiller’s CAPE, but it only explained 43% of subsequent returns. Eight of the sixteen factors had the same or less explanatory power than the prior year’s rainfall.

To summarize, forecasting returns or tactically allocating assets based on any market or economic indicator is a road we have decided not to travel on here at Index Fund Advisors, and we have no regrets.

1Davis, Joseph, Charles Thomas, and Roger Aliaga-Díaz, 2012. Forecasting Stock Returns: What Signals Matter, and What Do They Say Now? Valley Forge, Pa.: The Vanguard Group.