Different Makes a Difference^®

P/E Valuations are Neither Rich Nor Predictive

Investors bemoan the high price of stocks.  With the U.S. S&P 500 up almost 200% from its March 2009 low, and the Canadian S&P/TSX Index double its financial crisis trough, there is no doubt stocks have come a long way. It’s also true that valuation multiples have risen considerably. Yet, with price-earnings (P/E) valuation multiples only a tad above their long-run average, it seems a little silly to characterize the market as expensive.

It’s human nature to anchor on the recent past, and therefore understandable that some are getting vertigo considering how much stocks have risen. Still, all the talk about above-average P/E multiples is largely a waste of time because the P/E of the market is a lousy predictor of future returns. Ken Fisher, Forbes columnist, money manager and one of America’s 400 richest people, wrote about this fact extensively in his bestselling book The Only Three Questions That Still Count.

When Mr. Fisher looked for the correlation between the P/E of the U.S. stock market at the start of each year from 1872 to 2010 and actual subsequent 10-year returns, he got an R-squared of 0.25. Translated to layman’s language, that means the P/E only potentially explains 25% of 10-year returns. That’s pretty random. Other variables explain the other 75% of price returns. 

The P/E is even more useless as a forecasting tool over shorter time horizons. When Mr. Fisher compared P/E ratios and subsequent one-year returns over the 138 year period he couldn’t find a relationship. The R-squared was 0.01. 

Clearly, history tells us that we should ignore those who point to the market’s P/E ratio as a reason to be fearful. 

From a portfolio management perspective, it’s the valuation of individual stocks that really matters and not the price of the overall market. In fact, there are times when the benchmark averages look expensive on a P/E valuation basis and yet plenty of statistically cheap stocks exist. The turn of the century is a case in point. At the end of 1999, the major North American equity benchmarks had P/E multiples that were more than double the long-run average because technology firms and big, high quality company stocks were ultra-expensive. Nonetheless, it was a great time to buy stocks given it was a two-tiered market with scores of neglected and cheap stocks as well. We had our best year ever in 2000, with the Odlum Brown Model Portfolio gaining more than 40%.

Price is but one variable in the investment equation and it’s not always best to buy the statistically cheapest stocks. Sometimes it’s better to pay more for higher quality businesses if they can be bought at reasonable prices, as is the case today.

Unlike the situation in 2000, individual stock P/E ratios are tightly clustered around the market multiple, with little difference between the valuation of average businesses and great ones. While those investors who look for deep value are frustrated because there are fewer stocks that trade at a steep discount to the average stock, we remain excited because the premium for quality businesses is unusually small. 

In 2000 we did well because we were able to buy average businesses at great prices. Today, we are upbeat because we can buy great businesses at average prices.