In a study covering US stocks from 1973 to 1996, Dreman found that stocks with EPS surprises in the lowest P/E, P/B and P/CF quintiles vastly outperformed those with EPS surprises in the highest respective quintiles.
But earnings surprises can come in two flavours: positive and negative. Therefore, Dreman subsequently studied the effects of both of these types of surprises on the most popular (high price-to-X) and the most out-of-favour (low price-to-X) stocks. The results showed that negative earnings had a huge effect on the high-flying stocks (these stocks lost 4.3% of their value in the quarter following the disappointment, and 8.9% of their value in the year that followed), but little effect on the out-of-favour stocks (these stocks lost only .7% of their value in the quarter following the disappointment, and remarkably only .1% of their value in the year that followed, meaning they actually increased in value in the nine months after the first quarter following the poor results).
On positive earnings surprises, however, the out-of-favour stocks reacted very well, showing 3.6% gains in the first quarter following the good news, and 8.1% gains in the year following the good news. The high-flyers (those in the highest P/E quintile, for example) showed only small gains (1.7% in the quarter following, and 1.2% in the year following).
Dreman argues that the data clearly shows that stocks trading at premiums to earnings or book value have high expectations built into them. When earnings are positive, these stocks gain little, as strong news was already priced in. When they are negative, however, they take big haircuts. Meanwhile, stocks trading at low multiples of earnings or book values have the worst already priced into the stock. As such, disappointing data does not affect the price much, but positive data can generate strong returns! This research suggests the type of companies investors should be buying...
In an all-encompassing study, Dreman studied the returns of a 25-year strategy (ending just before the book’s publication) involving annual switching into the quintile of the market’s lowest priced stocks. The study found that low P/E stocks returned an astonishing 19% per year (low P/B: 18.8%, low P/CF: 18%, high-yield: 16.1%) compared to the market’s return of 14.9%.
Seeing as how 1970 to 1996 was a fairly bullish period for the market, Dreman also studied price performance during bear markets, and once again cheap stocks outperformed. This time, the high-yield dividend stocks took the top honours (negative returns of 3.8% per year, versus the market’s return of negative 7.5%), but stocks with low P/E, low P/B and low P/CF ratios also outperformed the general market.
Despite all the evidence, why are contrarians and value investors a small minority of market participants? Dreman argues the problem is psychological. Investors get caught up in new ideas, and despite their better judgement (including all the evidence cited above), they can’t bring themselves to invest in companies that the market has beaten down. This makes them go for IPOs of glitzy companies like Planet Hollywood and SpyGlassat P/E ratios of 100+ times earnings instead of boring companies that have been around a while that trade with small P/E or P/B ratios.