Management forecasts of earnings clearly provide valuable information about firms’ prospects. Many studies, going back at least to Foster (1973), have shown that these forecasts affect stock prices. More recent work has emphasized the conditional nature of the value of these forecasts. These studies, reviewed below, find that the impact of managerial guidance on stock market prices may be related to the complexity of the earnings process, the good versus bad news nature of the announcement, the precision of the forecast, the mode of disclosure, the quality of corporate governance, and manager credibility established by the history of past forecasts.
Recently, however, only about one-fifth of firms have actually provided management forecasts of earnings in any year (see Figure 1), and that fraction has declined over the last decade. The decision to provide earnings guidance is discretionary, and presumably managers offer guidance based on their privileged information only if they think it will be beneficial for the firm. Similarly, it would hardly be surprising if one driver of the earnings forecast is management’s assessment of the likely impact on stock market price. But if managers release a portion of their private information to the market through guidance decisions that are even in part affected by anticipated price impact, then the guidance numbers provided cannot be treated as exogenous.
Failure to account for this endogeneity will bias estimates of the price impact of guidance. While a large body of work has emerged exploring the determinants of management’s decision to guide, little attention has been paid to the impact of the potentially endogenous nature of those forecasts on the market’s reaction to them. Specifically, studies of the price response to guidance announcements have not generally addressed the problems posed by endogeneity in earnings forecasts.
As an example of potential endogeneity, suppose that managers are reluctant to announce earnings below analysts’ forecasts. They may then more consistently guide analysts’ forecasts down when analysts are too optimistic than they guide forecasts up when analysts are too pessimistic. Because managers have privileged information when they assess the relative optimism of analysts, this induces a potential feedback from stock prices to the guidance forecast. Managers may also envisage an optimal speed of “walk-down” of analysts’ expectations. In this case, too, they may tailor their guidance to gradually reduce analysts’ expectations, perhaps considering a forecast of the price response as they partially release their information to the market.