Knowledge that Transforms

This paper provides a general version of the accounting-based valuation model that equates the market value of a firm's equity to book value plus the present value of expected abnormal earnings. Prior theoretical work (e.g., Ohlson 1995; Feltham and Ohlson 1995, 1996) assumes investors are risk neutral and interest rates are nonstochastic and flat. Our more general analysis rests on only two assumptions: no arbitrage in financial markets and clean surplus accounting. These assumptions imply a risk-adjusted formula for the present value of expected abnormal earnings. The risk adjustments consist of certainty-equivalent reductions of expected abnormal earnings. A key issue deals with the capital charge component of abnormal earnings. It is measured by applying the (uncertain) riskless spot interest rate to start-of-period book value. Risks do not affect the rate used in the capital charge, and accounting policies do not affect the formula's constructs. An application of the general formula shows how the classic risk-adjusted expected cash flows model derives as a special case.

This study examines whether the market values financial statement data differently if it is disclosed instead of recognized in the body of the financial statements. We identify a sample of 229 SFAS No. 106 adopters who disclose an estimate of their anticipated liability for retiree benefits other than pensions (PRB) in their financial reports prior to the year of recognition. We then test whether the disclosed estimate of the PRB liability is valued differently by the market than is the subsequently recognized PRB liability. We provide modest and model-sensitive evidence that the recognized PRB liability receives more weight than the disclosed liability in market value association tests.

We investigate if earnings forecast accuracy matters to security analysts by examining its association with analyst turnover. Controlling for firm- and time-period effects, forecast horizon and industry forecasting experience, we find that an analyst is more likely to turn over if his forecast accuracy is lower than his peers. We find no association between an analyst's probability of turnover and his absolute forecast accuracy. We also investigate another observable measure of the analyst's performance, the profitability of his stock recommendations. There is no statistical relation between the absolute or relative profitability of an analyst's stock recommendations and his probability of turnover. We interpret our findings as indicating that forecast accuracy is important to analysts.

Two competing theories of initial engagement audit pricing are examined empirically. DeAngelo's (1981a) model predicts initial engagement discounts in all settings, while Dye's (1991) model specifically predicts discounting will not occur in settings where audit fees are publicly disclosed. Unlike the United States and most countries, audit fees are publicly disclosed in Australia. Our study examines initial engagement pricing in Australia during a time period when comparable U.S. studies report discounts of 25 percent (Ettredge and Greenberg 1990; Simon and Francis 1988). The Australian evidence finds initial engagement discounting only for upgrades from non-Big 8 to Big 8 auditors. Discounting for upgrades to Big 8 auditors is consistent with economic theories of discount pricing by sellers of higher-priced, higher-quality experience goods as an inducement to purchase when uncertainty about product quality is resolved through buying (experiencing) the goods. The evidence in our study is generally consistent with Dye's (1991) conclusion that public disclosure of audit fees precludes initial engagement discounting and the potential independence problems arising from such discounting.

This study examines whether personally performing other audit tasks can bias supervising seniors' going-concern judgments. During an audit, the senior performs some audit tasks him/herself and delegates other tasks to staff members. When personally performing an audit task, the senior would focus on the evidence related to that task. We predict that such evidence will have greater influence on the senior's subsequent going-concern judgment. The results of our experiment are consistent with our predictions. When provided with an identical set of information, seniors who performed another audit task for which the underlying facts of the case reflected positively (negatively) on the company's viability, subsequently made going-concern judgments that were relatively more positive (negative). Our results also demonstrate that the well-documented tendency of auditors to attend more to negative information does not always dominate auditors' information processing. Subjects who performed the task for which the underlying facts reflected positively on the company's viability directed their attention to such positive information and, consequently, both their memory and judgments were more positive than those of subjects in the other conditions. Recent findings indicating that biases in seniors' going-concern judgments may not be fully offset in the review process are discussed along with other potential implications of our results.

This paper examines investor welfare under two different liability regimes for holding auditors liable for investor losses, the due care and the strict liability regimes. In both regimes, the investor pays the expected legal liability cost to the auditor, and a portion of any subsequent damages awarded by the court is retained by the lawyer as a contingent fee, which is called the recovery friction. This study finds that the presence of the recovery friction leads to second-best efforts by the auditor and the manager. Investor welfare in the due care regime is higher than in the strict liability regime because the expected litigation cost for the investor is lower. Investor welfare is higher in the due care regime than in the strict liability regime even when audit effort in the due care regime is lower than audit effort in the strict liability regime.

Auditors encounter both relevant and irrelevant information during the performance of audit tasks. Prior studies have shown that the presence of irrelevant information weakens the impact of relevant information on auditors' judgments. Such studies, however, have not considered whether experience moderates the diluting effect of irrelevant information on auditors' judgments. This study reports the results of an experiment in which the effect of irrelevant information on the going-concern judgments of less-experienced auditors—audit seniors—is compared to the effect of irrelevant information on the going-concern judgments of more-experienced auditors—audit managers and partners. The experiment reaffirms that irrelevant information does have a diluting effect on the judgments of audit seniors but provides new evidence that irrelevant information does not have a diluting effect on the judgments of audit managers and partners.

This paper studies the impact of liability rules and damage awards on audit effort and the value of an audit (the net benefits to society of an audit) when an auditor and an investor may settle before proceeding to trial. It is demonstrated that audit effort increases with size of the damage award, but may decrease with the rigor of the auditing standards. For a given level of damage award, allowing pre-trial settlements may reduce the value of the audit despite the reduction in the deadweight legal costs. On the other hand, if the damage award is optimally chosen, then allowing settlements increases social welfare. With an appropriately set damage award, strict liability standards result in the first-best outcome, while the first-best result cannot be obtained with vague negligence rules.

Studies in the capital market context indicate that earnings changes and earnings levels considered jointly provide a more comprehensive representation of unexpected earnings than either earnings changes or earnings levels considered alone. Recent studies of executive compensation demonstrate that executive compensation revisions are greater when earnings innovations are permanent, than when innovations are transitory. Together, these literatures imply that both earnings changes and earnings levels explain revisions to CEO compensation. Specifically, formal analysis implies that weights on earnings changes vary directly with the persistence of earnings innovations and that weights on earnings levels vary directly with persistence for low persistence observations and inversely with persistence for high persistence observations. Evidence for compensation to 712 executives of U.S. corporations is consistent with these expectations. Such results suggest that earnings levels, earnings changes, and earnings persistence need to be considered when investigating relations between accounting earnings and executive compensation.