AMERICAN ACCOUNTING ASSOCIATION
CONTACT: Ben Haimowitz (212-233-6170)
Contrary to conventional wisdom, bad-news forecasts from management are less credible than the good-news variety
If a company's stock falls more than a downbeat forecast seems to warrant, proceed with caution
By now, it is old news in the finance and accounting literature that stock markets react more strongly to companies' bad-news forecasts than to their good-news predictions. For example, one study found the average share-price drop in response to the former to be almost 10%, while the average rise in response to the latter was only about 2%. Thus has the assumption has taken root among scholars that downbeat predictions are considerably more credible to investors than the upbeat variety.
Now a new study suggests that the opposite is the case. Presented at the annual meeting this month of the American Accounting Association, this research concludes that the market's stronger reaction to bad news "likely reflects the market's discounting of a bad-news management forecasting due to its lower rather than higher credibility."
In other words, investors "perceive bad news as less credible (i.e., more optimistically biased) than good-news management forecasts and discount bad news accordingly. As the actual news could be worse than that given by the managers, this discounting means more negative reaction to bad news."
And the reason for this credibility gap, concludes the study's author, Helen Hurwitz of Columbia University, is that companies are far less likely to get sued for pulling their punches when issuing bad-news predictions than for issuing rosy forecasts that later prove too rosy. In the words of the study, while "an early disclosure of good news can expose a firm to greater litigation risk if the news is overly optimistic...it is more difficult to sue a firm for issuing too optimistic bad-news forecasts."
In short, "litigation threat reduces the optimism in good news, but has no effect on bad-news management forecasts."
Further, "this finding warns investors and policy makers to take extra caution when reacting to bad-news management forecasts, as litigation risk does not deter a firm from issuing optimistic bad-news forecasts even though it is issued to reduce the firm's legal liability."
Asked what lesson investors might take from the study, Hurwitz says, "If a downbeat annual forecast issued by a firm occasions a steep decline in its stock, and the news doesn't seem to warrant such a sharp decline, proceed with caution. The market believes that the company's outlook is worse than its announcement suggests, and, based on the findings of this study, it is likely that the market is right."
The findings derive from an analysis of management forecasts of annual earnings per share of 2,393 public companies over the 12-year period 1996 through 2007. The forecasts are defined as good news or bad news depending on whether they were higher or lower than analyst consensus forecasts over the three months prior to the company's prediction. Degree of forecast optimism is determined by comparing the company forecasts to the actual earnings results at year's end -- that is, anywhere up to 12 months later.
Hurwitz finds that management's bad-news forecasts are significantly more likely than the good-news variety to err on the upside, overestimating what earnings will be and thereby softening the impact of unwelcome tidings. The tendency is most marked for earnings estimates made in the first three quarters of a company's fiscal year, when 90% of the forecasts were issued, prompting Hurwitz to observe that "forecasts of annual earnings released in the fourth fiscal quarter are essentially forecasts of quarterly earnings, which are more accurate and usually pessimistic due to their closeness to earnings announcements."
To probe the reason for the greater optimistic bias of bad-news predictions, Hurwitz employs a model from the accounting scholarly literature that estimates companies' inherent litigation risk, a formula incorporating a whole range of factors, including market value of a firm's equity, the daily turnover of its stock, the firm's industry, and the volatility of its stock. She finds that companies at higher risk for litigation are more likely than others to issue bad-news predictions, "consistent with firms' preempting the market with bad news when litigation risk is high." But credibility is another matter. In the words of the study,"litigation risk reduces the optimism in good-news forecasts, but has no effect on bad-news management forecasts."
Thus, for the study's full sample of 2.393 companies, intrinsic litigation risk did not inhibit firms' forecast optimism to any significant degree. The risk of being sued served to constrain only good-news predictors -- that is, managements who forecast better earnings than analysts anticipated.
Hurwitz also finds forecasts to be significantly affected by RegFD, the full-disclosure rule imposed by the SEC in 2000, which "requires management forecasts to be made to the public instead of to selected groups of analysts or institutions." RegFD, Hurwitz finds, "does not change the forecasting bias in bad news, but reduces the optimistic bias in good-news management forecasts." In explanation, she offers two reasons related to litigation: "1) RegFD could increase stock price reactions to management forecasts as they are now released to the general public, increasing the possibility that firms are sued if the news in the forecasts does not materialize. 2) Unaffiliated investors are more likely to sue firms for issuing misleading forecasts compared to affiliated or favored analysts or institutions who were selected by the firms to received the forecasts prior to RegFD."
The study, entitled "Is Bad News Really More Credible than Good News? Evidence from Management Forecasts," was among hundreds of scholarly presentations this month at the annual meeting of the American Accounting Association, a worldwide organization devoted to excellence in accounting education, research, and practice. More than 3,000 scholars and practitioners attended the meeting in San Francisco, Aug. 1-4