Which metric has better predictive power when it comes to stock returns.

V alue investors like to sort stocks by trailing P/E ratios and focus only on the lowest P/E quartile as they believe that stocks that have low valuations in relation to trailing earnings, on average, outperform high P/E stocks. Other investors, however, like to use forward P/E ratios to identify outperforming stocks going forward. But which metric has a better predictive power as far as stock returns are concerned, trailing or forward P/E ratios? Do low P/E quartile stocks yield higher or lower return vis-a-vis high P/E quartile stocks when P/E is defined based on trailing or forward earnings?  There is no doubt that low P/E stocks (however defined) beat high P/E stocks.  Evidence in this regard is overwhelming. Low P/E stocks outperform high P/E stocks by between 6% and 13% depending on the country.  This outperformance, on average, holds steady over different time periods, as well as during recoveries/recessions and bear/bull markets. Nevertheless, there is no much evidence as to which stocks do better, those that have low forward P/E or low trailing P/E.

My research shows that while forward P/E ratios are a good predictor of future returns for NYSE stocks, they are not as good predictor of future returns for AMEX and NASDAQ stocks.

I find that low forward P/E quartile stocks experience the same return as high forward P/E quartile stocks for AMEX stocks, and beat the high P/E quartile stocks by 2.40% for NASDAQ and 11.64% for NYSE stocks.

The evidence using trailing P/E ratios has as follows. Low trailing P/E quartile stocks beat the high P/E quartile stocks by 6.24% for AMEX, 11.40% for NASDAQ and 9.00% for NYSE stocks.

Why do forward P/E ratios have inferior forecasting ability compared to trailing P/E ratios?  The answer is simple. Trailing earnings are based on realized earnings, while forward earnings are based on earnings forecasted by analysts. Analysts tend to be overoptimistic when forecasting earnings. This biases forward P/E ratios down giving the impression that a stock commands a low P/E, when in fact it may not be as earnings are eventually revised downwards and what appeared at first to be a value stock may, in fact, turn out to be a high P/E stock.

For example, on average within a calendar year, analysts overestimate actual earnings by about 2.5%.  But the overestimation is about 8% at the start of the forecasting period. Accuracy improves as analysts approach the end of the year which they are forecasting.

What is more interesting, however, is that analysts are not overoptimistic across all companies covered.  They tend to be overoptimistic only for stocks for which there is high uncertainty about the future. For stocks with low uncertainty, analysts tend to be pretty accurate. In fact, I find that for the lowest uncertainty stocks analysts, on average, exhibit no upward earnings forecast bias within a calendar year. But when it comes to the highest uncertainty group of stocks, on average, analysts tend to overestimate actual earnings by 21%. These are big errors, hence the inaccuracy of forward P/E ratios, particularly for this group of stocks.

In other words, forward P/E ratios can be a good predictor of future returns for low uncertainty stocks but are quite inaccurate for high uncertainty stocks.  This may explain why forward P/E ratios work well in predicting future returns for NYSE stocks, but they do not work as well (compared to trailing P/E ratios) for AMEX and NASDAQ stocks, namely, the riskier group of stocks. As a result, if investors want to stop worrying about the uncertainty underlying a company’s future, they should consistently use trailing P/E ratios to screen stocks. In fact, this is the approach that the father of value investing Benjamin Graham followed when he wanted to identify stocks that were likely to outperform. Investors should follow his lead, especially when it comes to riskier markets.