Accounting Defects, Financial Statement Credibility,

We would like to acknowledge helpful discussions with Robert Bowen and David Burgstahler. Existing equity valuation models are silent on the question of how changes in financial statement credibility alter the usefulness of accounting data for valuation purposes. 1 Reported cash flow, earnings and balance sheet data are either taken at face or they are adjusted to eliminate alleged deficiencies unrelated to credibility concerns (e. contrast to the role credibility plays in principal-agent settings involving adverse selection (e.g., Bartov [1991], Sansing [1992]) and in behavioral decision theory models of cascaded inference (e.g., Beach et al. [1978], Schume and DuCharme [1971]). In both literatures, a reduction in credibility attenuates the usefulness of reported information for decision-making purposes. This paper investigates whether detectable attenuation effects are also present in the relation between accounting information and observed equity values once the credibility of the financial statements has been called into question. Specifically, we identify companies with rumored or acknowledged accounting defects, estimate parameters of a cross-sectional equity valuation model before and after the defects are made known, and then test for changes in the parameter estimates. If attenuation effects are present, the coefficient estimates for the accounting variables in the valuation model will decline in response to discovered accounting defects. On the other hand, the market may ignore the defects or assign a fixed penalty (an intercept decline) rather than placing less weight on reported accounting data. We also develop and test cross-sectional 1 Credibility refers to the quality, capability or power to elicit belief and trust. In other words, credible financial statements convey information that can be believed and trusted. This notion is closely related to the more familiar concept of reliability. For example, " (a)ccounting information is reliable to the extent that users can depend upon it to represent the economic conditions or events that it purports to represent " (Financial Accounting Standards Board [1980], para. 62). Accounting researchers have traditionally used reliability to mean measurement error, noise or precision (e.g., Barth [1991] and Choi et al. [1997]). 2 predictions about the magnitude of attenuation based on characteristics of the accounting defects themselves and on the circumstances of their discovery. Two broad types of accounting defects are examined: (1) alleged deficiencies identified by securities analysts, news reporters, or other commentators and (2) actual deficiencies acknowledged by a company spokesperson and corrected when restated interim or annual financial statements are issued. Approximately …