The Information Content of Annual Reports: A Price and Trading Response Analysis.

Abstract This research examines and interprets the security market response around annual report release dates. Annual reports contain data beyond that reported in preliminary earnings announcements. While prior research has concentrated on the price response to annual report releases and the issue of information content with mixed results, this paper extends earlier work by employing both price and trading measures. The trading measures, which include number of transactions, size-stratified transactions, and mean transaction size, enable the analysis to address not only information content issues, but also social welfare issues in the sense of Lev and Ohlson (1982) and to coarsely identify what type of investor responds to the annual report by trading. inferences as to the relative usefulness of annual reports to different investor groups are also made based on these trading measures. A number of studies address, often indirectly, annual report informativeness by examining the price response accompanying the report's release. Of these, studies by Wilson (1987) and Lobo and Song (1989) suggest a price response to the earlier of the annual report or 10-K. However, other studies, including Foster et al. (1986), Mynatt (1988), and Bernard and Stober (1989), fail to detect such a price response. The informativeness of annual reports from a price-based perspective remains an open question and from a trading-based perspective it is an unaddressed question. In this study, price response to the annual report is measured using both absolute value and squared unexpected returns. A market model approach is used to estimate unexpected returns. A simulation analysis is included to substantiate the ability of these two metrics to detect a price response. Unexpected volume and unexpected number of transactions, both in general and for size-stratified trading, are estimated via market model regressions. Unexpected mean transaction sizes are estimated from a temporal drift regression. No evidence of a price response and little evidence of a volume of shares response at annual report dates is found. Number of transactions, however, increases significantly around annual reports, peaking four to five trading days after the annual report release date. This trading response suggests that investors find annual reports informative. The simultaneous absence of an economically significant price response is interpreted as indicating that annual reports provide social welfare value in the sense proposed by Lev and Ohlson (1982). Most notably, the analysis of trading response stratified by transaction size suggests that the trading response occurs mostly within the smallest size strata (i. e. , 100 and 200 share transactions, transactions of less than 30, 000 in value). Consistent with this, mean transaction sizes during the annual report period are less than expected. This is contrary to mean trans- action size behavior at earnings announcement dates (Cready 1988) and suggests that annual reports are of greater value to less wealthy individual investors relative to wealtheir individual and institutional investors. Such a finding is consistent with Hakansson (1977) in that it suggests that "small" investors rely on the public information system (i. e. , the annual report) while "large" investors rely more on predisclosure information in making investment decisions.