March 27, 2008

As if the recent widespread decline in stocks hasn't been bad enough for investors, a new study has now uncovered a worrisome hazard even among the market's best-performing companies, the top one percent in terms of stock gains over the previous year.

The hazard consists in buying shares of those top winners just before they announce their earnings.

The study, drawing on almost 35 years' worth of data, from 1971 to 2005, finds smaller investors especially prone to do this. And it also reveals the tactic not only unlikely to yield quick gains but likely to produce losses instead.

The new research, by David Aboody and Brett Trueman of UCLA Anderson School of Management and Reuven Lehavy of the University of Michigan, finds that, while shares of the market's top-performing companies rise quickly on average in the five trading days before their earnings announcement, they fall just as precipitously in the five following days.

According to the study, "The top percentile of stocks in terms of the past 12-month price performance (sometimes referred to as past winners) experiences a significant average market-adjusted return of 1.58% during the week prior to their earnings announcement (the pre-announcement period) and a significant average market-adjusted return of minus 1.86% in the week after (the post announcement period). By way of contrast, the average pre-announcement market-adjusted return for our entire sample of stocks is a meager .3%, while the average post-announcement market-adjusted return is a negligible .1%."

Adds Prof. Trueman: "While it may seem sensible or even clever to anticipate quick gains from latching onto earnings announcements of the market's biggest past winners, our findings suggest it generally doesn't work. In one analysis we did, we found that investors who bought shares of top-performers on the final day before their earnings announcements were behind a week later by a market-adjusted amount of close to 2% on average and that they were behind by about 1.5% on average a month later. And this is with a very tempting group of stocks, a group that had been rising very impressively."

In trying to account for this surprising pattern, the authors consider whether it reflects a tendency of stock analysts to revise their earnings forecasts for past winners upward in the week before earnings announcements and downward in the following week. But their extensive analysis finds "no evidence to support this potential explanation. Less than 2% of our sample observations are characterized by both an upward revision in analysts' forecasts in the week prior to earnings announcements and a negative earnings surprise or downward forecast revision during the week thereafter. Moreover, dropping these few observations from our sample does not significantly affect the magnitude of the pre- and post-announcement returns."

Having determined that a disproportionate number of pre-announcement trades of past winners' shares are by small buyers (trades under $50,000) and medium buyers (trades between $50,000 and $100,000), the authors conclude that it is the relative lack of sophistication among these buyers that accounts for steep pre-announcement rises and post-announcement falls. As the professors put it, "Smaller investors, faced with limited time and resources, are more likely to invest in stocks that draw their attention. Among stocks capturing these investors' attention are arguably those that have increased sharply in price. Moreover, their attention is likely to be heightened just before earnings releases due to media focus on the upcoming announcements. Price pressure from these investors could partially explain the positive pre-announcement returns. A lessening of that pressure subsequent to the earnings announcements could, in part, explain the post-announcement return reversal."

While viewing their study as conveying a cautionary message, the authors do provide evidence that the pattern of sharp rises and falls they uncover could have been used profitably by sophisticated investors. To test this, they focused on a subset of stocks that announced earnings after normal trading hours. The optimal strategy, they write, would have involved "purchasing our subset of stocks five days before their earnings announcements, closing the positions at the open on [the following day] then initiating short positions which are closed at the end of day 5." This would have generated market-adjusted return over the 10-day period, they note, that would have exceeded the imposition of both the bid-ask spread and brokerage commissions."

The study's initial data set consisted of  293,630 quarterly earnings announcements from January 1, 1971 through September 30, 2005. The professors computed each stock's raw returns for the 12-month period ending on the last day of the previous quarter, ranked the stocks in ascending order according to their returns, and partitioned the firms into deciles.They then partitioned the top decile into 10 percentiles according to prior 12-month return. The top percentile's return, they found, was almost twice that of the corresponding return for the next-highest percentile; the average five-day pre-announcement market-adjusted rise for this top percentile was more than 60 percent higher than that of the top decile as a whole, and its five-day post-announcement market-adjusted decline was more than twice that of the top decile's. Given the dramatic difference in returns, the professors focused particular attention on the top percentile.

In conclusion, they observe that the pattern of buying encountered in the study is "reminiscent of the adage 'buy on the rumor, sell on the fact.' There is a difference here, though, in that the 'rumor' is simply that there is an upcoming earnings announcement, not that the news will necessarily be better than expected." In that sense, the professors add, the pattern is similar to the price rise in initial public offerings that commonly occurs "in advance of the ending of the quiet period, with the 'rumor' being only that the lead banker's analyst will shortly be issuing a research report, not that the content of the report will be any more positive than expected."

Media Relations