Executive Summary 

Financial analysts play an important role in firm valuation by providing benchmarks to evaluate performance. There is now abundant evidence in accounting and finance indicating that firms pay close attention to earnings forecasts issued by analysts and respond to the magnitude of the forecast errors at the earnings announcement date. Furthermore, recent studies document a market premium (penalty) to the act of meeting or exceeding (failing to meet) analysts’ earnings forecasts even after controlling for the magnitude of the forecast error. This evidence suggests that the market views the act of meeting (or missing) earnings forecasts as a signal regarding future profitability.

Clearly, the market has access to other signals at the earnings announcement date, including additional signals provided by analysts. For example, earnings announcement reports often discuss whether firms also meet (or fail to meet) revenue expectations and why. This raises some interesting research questions. Does the market attach a premium to the act of meeting revenue expectations, just as it responds to the act of meeting earnings forecasts? Is the documented market premium to the act of meeting earnings expectations a function of whether or not revenue expectations are also met? Does the sign of revenue forecast errors influence the magnitude of earnings response coefficients? What are some of the determinants of the market response to the joint acts of meeting revenue and earnings expectations? Answering these questions constitutes a natural next step in this literature to advance our understanding of how the market processes multiple accounting signals, and is the purpose of this paper.

It is important to note that investors have access to many performance metrics beyond just earnings and revenues. However, we limit our analysis to these two performance metrics for two reasons. First, our research question focuses on the market response to the meeting or beating of analysts’ expectations and, thus, it is necessary that forecasts be available for whatever performance metric we examine. Next to earnings, revenues forecasts are perhaps the most widely followed performance metric by analysts. It takes only a cursory examination of a random sample of earnings announcements to appreciate the importance that the market appears to place on revenues and earnings signals. Second, prior research on revenue forecasts has documented a significant association between returns and revenue forecast errors after controlling for earnings forecast errors.

Within traditional valuation models, earnings are explicitly included either as a primary variable (for example, residual income model) or as a proxy for cash flows. The inclusion of revenues in a valuation model is essentially equivalent to partitioning earnings into their revenue and expense components. Accordingly, prior studies that have focused on slope coefficients on earnings and revenue forecast errors have examined whether the market responds differently to revenue and expense surprises. Our study differs from this research in that our primary focus is on the joint signaling implications of the acts of meeting earnings and revenue forecasts, rather than on the association between returns and the magnitude of revenue (and expense) surprises.

Anecdotal evidence is readily available in the financial press indicating that the market’s interpretation of earnings news depends on revenue performance. For example, after the markets closed on October 8, 2005, Apple Computer, Inc. announced fourth-quarter earnings of $0.52 per share, which was substantially above the consensus earnings estimate of $0.37 per share issued by Thomson Financial. However, in after-hours trading, Apple’s stock price decreased nearly 11 percent. The reason stated by the financial press was that reported revenues of $3.68 billion failed to impress analysts, being well below forecasted revenues of $3.74 billion. Thus, while Apple’s earnings performance exceeded expectations, its revenue performance did not, and the market reacted negatively. While recent research has investigated the effect of revenue performance on earnings response coefficients, our study is the first to examine the effect of meeting an alternative accounting benchmark (that is, revenue forecasts) on the observed market premium/penalty to meeting/missing earnings forecasts.

Consistent with our hypotheses, we document a significant increase (reduction) in the market premium to meeting earnings forecasts when the revenue forecast is also met (not met). Similarly, the market penalty to missing earnings forecasts is significantly attenuated (accentuated) when the revenue forecast is met (not met). Interestingly, we also find that the association between the magnitude of the revenue forecast error and abnormal returns surrounding the earnings announcement, documented in prior studies, vanishes when we explicitly take into account the market effects of meeting/beating earnings and revenue forecasts. This result suggests that meeting revenue forecasts has greater valuation implications than the actual magnitude of the revenue forecast error. We find some evidence that the earnings response coefficient (ERC) is significantly different depending on whether or not both revenue and earnings forecasts are met; however, this result is not robust in a specification that allows for nonlinearity in the relationship between abnormal returns and forecast errors.

We extend our analysis by investigating conditions under which revenue forecasts are potentially more informative to market participants. We find that the confirmatory effect of meeting revenue forecasts on the premium/penalty to meeting/missing earnings forecasts is more pronounced in certain industries and during specific time periods. We also find that firm growth has a significant effect on the market premium to meeting earnings and revenue forecasts and the ERC.

We next examine how analysts revise their earnings forecasts for the next quarter conditioned on whether the earnings and revenue forecasts were met. Consistent with our market tests, we find that analysts’ responses are more (less) pronounced when the earnings and revenue signals are of the same (opposite) sign. These results have implications for the market’s (and the analysts’) perception of the persistence (or quality) of earnings when revenue forecast errors are confirmatory in nature.

In summary, our study documents several findings that are new to the literature. First, we find that the market attaches an equity premium to the act of meeting revenue forecasts that is separate and distinct from the equity premium attached to meeting earnings forecasts. Second, we show that the market response to meeting earnings forecasts is not always positive. Specifically, the mean stock price return to meeting earnings forecasts is not significantly different from zero when revenue forecasts are not met. Third, our results offer new insight regarding the value-relevance of revenue forecast errors. Specifically, prior studies provide evidence that revenue forecast errors bear a significant association with announcement period returns. In contrast, after we control for the valuation effects of meeting earnings and revenue forecasts, we do not find this association, which suggests that the magnitude of the revenue forecast error is not as value-relevant as the mere act of meeting the revenue forecast.

All Images & Content Copyright 2006 - Canadian Academic Accounting Association
Powered by WebPal CMS, Palomino Inc.