Unconventional wisdom about P/E ratios

Email to a Friend

Stocks with low price-to-earning (P/E) ratios are good buys, right? Not so, according to Ken Fisher, author of “The Only Three Questions That Count: Investing by Knowing What Others Don’t.”

“[T]here is no correlation between market P/E level and subsequent return … For every example you find of a high or low P/E year doing well you will find the same number of examples from that level of it doing badly,” Fisher says. For example, the P/E of the market on Jan. 1, 2001, was the sixth-highest of the 134 years from 1871 to 2005, he says. The S&P 500 was down 12 percent that year. It was the fourth-highest on Jan. 1, 2000, and the market was down 9.1 percent then, he says. But the market’s P/E was higher still in 2003, the third-highest year, and the S&P was up 26.7 percent, and the second-highest year was 1999, and the S&P was up 21.0 percent, he notes.

Studies “proving” the correlation “use a bizarre P/E definition and apply an inflation adjustment using an arbitrary and little-used inflation indicator,” Fisher contends. In his own study, he plotted years based on P/E and found that P/Es are not predictive. Not only that, P/Es can be high, because earnings have dipped, which could mean the stock is a good buy. He says investors should change their point of reference by flipping the P/E to an E/P to get the earnings yield, quoted as a percentage. “[T]he earnings yield easily compares to the bond yield and the relative cost for firms to borrow expansion capital or transact mergers or buybacks,” Fisher says.