It's largely accepted on Wall Street that projected earnings are of the utmost importance in basing an investment decision. Companies with the best earnings outlook are usually afforded higher price to earnings (P/E) ratios. With limitations in human information processing, the inherent human psychological biases and the poor performance track record of investment professionals, Dreman asks the pertinent question of whether buying companies shunned by the crowds would produce superior investment results. The companies that the market expects will have the best futures are generally ones with high price to earnings ratios, so Dreman looked at the question of whether low P/E stocks would outperform high P/E stocks.
Dreman references studies conducted by Graham and Dodd, Francis Nicholson, Paul Miller Jr., William Breen, Sanjoy Basu and his own research conducted in various time periods between 1937 and 1980 and with American stock data. The data was broken up into different "buckets" (quintiles and/or deciles) of stocks ranging from low P/E to high P/E.
All of these studies consistently show that low P/E stocks perform significantly better than the high P/E stocks. Even in periods such as 1970 to 1973 when growth stocks were all the rage in the markets, the lowest P/E decile of stocks had superior returns to the highest P/E decile. Except at the peak of growth stock market mania, the second lowest P/E decile also performs better than the second highest P/E decile.
Efficient market theorists would claim that the low P/E stocks have more risk as measured by price volatility and should therefore be expected to have higher returns on average. Their stance is that the extra risk is rewarded with higher returns. However, studies conducted by Dreman and others have shown that the lowest decile of P/E stocks have even less volatility than the highest decile of P/E stocks.
Dreman asks the question of why more people are not buying the low P/E stocks in light of the research evidence? The research supports buying low P/E stocks but the idea is almost ludicrous on Wall Street. Perhaps the following excerpt taken from a prominent money manager best captures the reasons of why more experts are not condoning the low P/E approach:
"Do you know what's down there? How can any prudent man look at companies like these with such unthinkably poor visibility? The high P/E's, on the other hand, present the best visibility money can buy. They are the real blue bloods of the investment scene. How, then, can one recommend such a reversal of course?"
Dreman summarizes the problem as people not wanting to let go of predicting which companies will perform well based on extensive earnings forecasts. Dreman touches back on cognitive biases, where people believe that as P/E ratios move further out from those of average companies that the ratios are increasingly justified. The reverse is true with out of favor stocks. Dreman believes that people are too confident with their evaluations and make systematic errors that lead to the poor results for high P/E stocks and superior results for low P/E stocks.