Do Stock Prices Fully Reflect the Implications of Current Earnings for Future Earnings for Ar1 Firms?

Abstract

The extant literature has concluded that stock prices do not fully reflect the implications of current earnings for future earnings. This evidence is based on random samples of firms, whose quarterly earnings are assumed to be generated by the same, relatively complex, Brown-Rozeff (1979, BR hereafter) process. It is conceivable that model complexity is the source of the market's failure to fully reflect the implications of current earnings for future earnings. The purpose of this study is to determine whether stock prices fully reflect the implications of current earnings for future earnings for a subset of firms whose quarterly earnings generating process is much simpler than the BR model. For the entire sample of AR1 firms, we find that stock prices do not fully reflect the implications of current quarterly earnings for future quarterly earnings. They erroneously act as if the error of the AR1 model at lag four has negative valuation implications. When we segment the sample by several proxy variables for firms' information environments (i.e., firm size, number of institutional shareholders, and the existence of security analyst following), we find that the failure of stock prices to fully reflect the implications of current quarterly earnings for future quarterly earnings pertains only to firms with less predisclosure information. When we examine the relation between current earnings surprise and contemporaneous and future CARs, we obtain additional insights into what stock prices do not understand about firms with less predisclosure information. For firms with less predisclosure information, the failure of stock prices to fully reflect the implications of current quarterly earnings for future quarterly earnings pertains to large positive surprises, but not to large negative ones.

JEL Classification: G14, M41