The Role of Earnings Guidance in Resolving Sentiment-driven Overvaluation

The following post comes to us from Nicholas Seybert of the Department of Accounting & Information Assurance University of Maryland, and Holly Yang of the Accounting Department at the University of Pennsylvania.

In our paper, The Party’s Over: The Role of Earnings Guidance in Resolving Sentiment-driven Overvaluation, which was recently made publicly available on SSRN, we show that an important link between investor sentiment and firm overvaluation is optimistic earnings expectations, and that management earnings guidance aids in resolving sentiment-driven overvaluation. Understanding the underlying process linking investor sentiment to overvaluation provides insight into investor psychology and difficult-to-predict bull and bear markets. Currently, there are multiple possible explanations for why uncertain stocks are overvalued during high sentiment periods. For example, investors may exhibit different preferences, such as reduced risk aversion, during high sentiment periods, which would lead them to overpay for stocks with high valuation uncertainty. Under this scenario, in subsequent months, a general shift in investing trends or psychology would lead to the gradual decline in prices. Alternatively, investors may engage in a more detailed thought process that involves unrealistic expectations of future firm earnings, where there is more potential to overestimate future earnings for uncertain firms. Under this scenario, in subsequent months, revisions in earnings guidance or other earnings news released by the overvalued firms should lead to predictable price declines. We focus on the second explanation, examining management earnings guidance to test the extent to which the correction of earnings expectations mitigates the overvaluation problem.

In theory, if sentiment reflects investors’ irrational expectations about future earnings, then firm-specific earnings information should correct for investors’ expectation errors and reduce overvaluation. Consistent with this conjecture, growth stocks and stocks subject to high differences of opinion tend to earn significantly lower returns around earnings announcements. However, prior investor sentiment research finds only weak evidence that the average earnings announcement return is inversely related to sentiment for the subset of firms that are overvalued (undervalued) when sentiment is high (low). In addition, there is no evidence that earnings announcement effects are any greater than would be expected during a randomly chosen three-day window. We focus on returns around earnings guidance windows because these are important events which provide information to investors about firms’ expected performance, prior to the actual earnings announcement. It is well established that guidance events influence both stock prices and analysts’ opinions. Moreover, several recent accounting studies find that guidance events are more important than earnings announcements in conveying information to equity markets.

We use the First Call Company Issued Guidelines database to delve deeper into the process through which sentiment-driven overvaluation is resolved. Whereas previous studies (Baker and Wurgler (2006), Lemmon and Portniaguina (2006), Baker and Wurgler (2007), Hribar and McInnis (2011)) examine the returns over the subsequent month or quarter depending upon prior investor sentiment, we examine the pattern of returns within those subsequent months. Specifically, we test whether the lower subsequent returns are concentrated around earnings guidance issued by managers of overvalued firms. Like previous studies of sentiment, we find that small firms and young firms, as well as those with high return volatility, high intangible assets, and low dividend payments, have lower monthly returns when prior investor sentiment is high. However, these returns are mostly contained within the three-day window around the issuance of management earnings forecasts. Roughly three-fourths of the predictable negative returns occur in this three day window, while the remaining fourth occurs over the other eighteen trading days in the month. Additionally, returns over entire months during which guidance was not issued are roughly half the magnitude of three-day guidance window returns, again suggesting that guidance plays a strong role in price correction. We also find that the market is more sensitive to earnings surprises (particularly negative surprises) following high sentiment periods. However, results from additional tests suggest that managers’ attempts to preempt the earnings announcement are not always successful, as we still find some evidence that returns around the three-day earnings announcement window are significantly negative for small and loss firms, and firms with higher volatility following high sentiment periods. Moreover, consistent with Baker and Wurgler (2006), we do not find any systematic evidence that sentiment-driven overvaluation resolves around the three-day earnings announcement window for firms that did not issue guidance. This pattern of results suggests that managers may intentionally issue guidance in an attempt to preempt negative earnings surprises, which could explain why sentiment-driven overvaluation does not resolve around earnings announcements. Consistent with prior work showing that transient institutional ownership induces earnings-related myopia (Bushee (1998)), we find that firms with higher transient ownership are less likely to guide (and their guidance is less likely to contain a negative surprise) following high sentiment periods.

Taken together, our results suggest that earnings expectations play a strong role in sentiment-driven overvaluation. In addition, investors do not appear to ignore bad news when firms are overvalued, but rather exhibit a stronger reaction to bad news forecasts following high sentiment periods. Our findings are important for understanding the nature of investor sentiment and earnings expectations, as well as the effects of management earnings guidance on the market. While prior research on investor sentiment investigates how various agents (managers, investors and analysts) are affected by investor sentiment, none examines the factors that resolve sentiment-driven overvaluation. Our results indicate that investor sentiment does not lead to overvaluation through general beliefs or preferences such as shifts in risk aversion, but rather through firm-specific earnings expectations, a finding which may be relevant to the formulation of hypotheses and research designs in future studies of investor sentiment and market bubbles. Our paper also expands upon prior research investigating the interplay of accounting information and investor sentiment. Whereas these studies conclude that investor sentiment leads managers (analysts) to intentionally (unintentionally) bias their disclosures and forecasts, our study suggests that management earnings guidance plays a significant role in correcting optimistic market expectations, and thus that firm-specific disclosures also have beneficial effects. Finally, Berkman et al. (2009) find that earnings announcements reduce overvaluation in stocks subject to high short sale constraints and high differences of opinion, or in other words in stocks for which valuation difficulty is identifiably high. Our findings suggest that earnings guidance events play an important role in mitigating overvaluation driven by a different source (investor sentiment), suggesting that earnings expectations likely play a broad role in overvaluation and that the disclosure of earnings-related information plays a broad role in mitigating this problem.

The full paper is available for download here.

Both comments and trackbacks are currently closed.