By JACK HOUGH
Two companies report earnings that surpass Wall Street estimates by 10 percent. Shares of the first company jump, but those of the second are unchanged. Why?
Nowadays, so many companies regularly beat their earnings forecasts that surprises aren't always, well, surprising. Some of these companies set the bar low; others perform the accounting equivalent of searching the couch cushions for loose change just before the end of the quarter.
That's important, because studies have long shown that stock investors should favor companies that beat earnings forecasts, all else being equal. Fortunately, new evidence suggests three ways investors can filter out such unimpressive forecast beaters and focus on genuine outperformers -- shares of which tend to do much better than the rest.
The same Wall Street analysts who give buy, hold and sell advice on stocks make estimates of quarterly earnings, and pollsters gather those estimates to form a single consensus for each stock that analysts follow. Back in 1968, accounting professors Ray Ball and Philip Brown famously showed that shares of companies that surpass consensus forecasts tend to perform better than the broad stock market; the phenomenon, which has been backed up by more recent studies, is known as post-earnings-announcement drift, or PEAD.
The most interesting part of PEAD is the D. In a perfectly efficient stock market, investors would fully price in good news as soon as it's received. Shares of forecast beaters would tend to rise right away and then perform in line with the market afterward. But that's not what happens. Shares indeed often leap on good earnings news, but afterward, they tend to gradually drift higher for months. Researchers suspect the reason is rooted in investor psychology -- that stock buyers are simply slow to adjust their biases in reaction to fresh evidence. Whatever the reason, the gradual nature of PEAD suggests that even slow-poke traders can take advantage.
The parade of positive "surprises" continues. Among S&P 500 members, more companies have beaten forecasts than have missed them for 11 straight quarters, with upside surprises often making up more than two-thirds of total surprises. In other words, what started as a clue to good stock performance became a rigged system, where companies and analysts chronically underestimate earnings, or else companies that are running just short of forecasts slash discretionary spending on things like research to make up the difference.
If rigged seems too strong a word, consider two pieces of evidence. A 2009 study published in the Journal of Finance showed that companies that beat earnings estimates by mere pennies per share while cutting spending on things like research and advertising went on to perform worse than those that slightly missed forecasts but didn't cut such spending. And managers at the former group of companies were more likely to sell shares, suggesting they knew their short-sighted actions would hurt long-term performance.
A 2010 study from Stanford documented something even more telling: a scarcity of the number 4 in earnings reports. See, companies always round their earnings per share to the nearest penny. If the math comes out to $1.495 per share, they report $1.50. If it's $1.494 a share, it becomes $1.49. It turns out 4s are rare in that third decimal place -- and they become more rare as firms gain more analyst coverage, according to the study.
What harm could a tenth of a penny do? Plenty. The study found that firms exhibiting higher levels of quadrophobia (fear of 4s) were more likely to restate earnings later and to be the subject of class-action lawsuits and regulatory action related to their accounting.
Clearly, investors buying stocks based on earnings surprises should be choosy. The first of three ways to do that is obvious: Look for firms that beat forecasts by a lot rather than a little. Their results are likely driven by strong operations rather than accounting tweaks.
The second way is to look for companies that beat revenue forecasts at the same time they beat earnings forecasts. Companies that top earnings estimates alone might have done so through cost-cutting, which is fine, but opportunities for cost-cutting can run out quickly. Companies that outperform on revenue, too, are likely benefiting from growth.
A 2006 study published in Financial Analysts Journal found that at any given point, the top 20 percent of companies ranked by recent upside earnings surprises alone went on to beat the market by three percentage points, on average, over the subsequent six months. The top 20 percent by upside revenue surprises beat the market by 2.6 percentage points. However, the top 20 percent ranked by a combination of earnings and revenue surprises beat the market by 5.3 percentage points over the following six months.
The third way to get better results from a search for upside earnings surprises comes from a study slated for publication this year in the same academic journal: Watch investor reaction to the earnings reports. "If investors are truly surprised by an upside earnings announcement, the stock price should jump in the short term," explains author Haigang Zhou, a professor at Cleveland State University. Zhou and a colleague looked at returns from 1971 to 2009 and found that firms with upside earnings surprises followed by unusually high price gains over the next three days went on to beat the market by 5.7 percentage points over the subsequent 60 trading days.
The table on page 37 lists a handful of stocks that turned up in a screen for all of these things: big, upside earnings surprises; revenue surprises too; and sudden price gains in the days following the earnings reports.