Knowledge that Transforms

Abstract. There has been a strong growth in accounting firms' provisions of nonaudit services to their audit clients. To date, however, there have been few studies into the determinants of this joint provision of audit and nonaudit services. One reason for the paucity of research is the lack of publicly available data with which to empirically examine relationships and test theories. However, recent legislation in the United Kingdom requires publication of nonaudit fees paid to a company's auditor, and this disclosure provides the data with which to investigate the joint provision of consultancy and audit services. A model is developed that seeks to explain a company's decision to hire nonaudit services from the auditor. The model argues that companies that face potentially high agency costs purchase relatively smaller amounts of nonaudit services from their auditor. High agency‐cost companies require independent audits in order to reassure investors and creditors; the provision of joint services, which increases the economic bonding of the auditor to the client, may jeopardize independence or the appearance of independence. The model is tested using data observations from 500 companies, and the results indicate that companies that have higher agency‐cost proxies are associated with smaller purchases of nonaudit services from their auditors.

Abstract. Analysts have been found to underweight the innovation in the most recent quarterly earnings when forecasting next‐quarter earnings, and these expectations have been posited as an explanation for post‐earnings‐announcement drift. This study uses an experimental asset market to examine whether similar errors in forecasting quarterly earnings are made by student‐subjects. We examine two aspects of their behavior: (1) do subjects underestimate the autocorrelation in quarterly earnings when forming earnings expectations? and (2) are asset prices consistent with a subject's underestimation of the autocorrelation in quarterly earnings? We observe subject errors in forecasts that underweight extreme innovations in the most recent quarterly earnings by approximately 40 percent. The prices in the experimental markets also fail to reflect fully the most recent innovation in quarterly earnings. We are able to predict the sign of the incorrect pricing, from the mean initial earnings predictions of the subjects, in 74 percent of the 135 markets. These forecast errors observed in this study are consistent with forecast errors observed for analysts, and this consistency suggests that errors in analysts' forecasts may be at least partially attributable to the use of judgmental heuristics.

Abstract. This paper reports the results of an experiment examining the framing bias and a potential debiasing technique. Practicing auditors formed a judgment about a hypothetical client's inventory internal control system to determine the amount of related substantive testing. Auditors from two Big Six firms were randomly assigned to one of four treatments in a fully crossed 2 times 2 between‐subjects design. The initial description of the internal control system was identical for all treatments, as were the items of additional evidence about the system. Auditors either judged the risk of the control system or the strength of the control system. The risk‐strength frames were crossed with two levels of debiasing technique: “none” or “evidence rating.” Results indicate that without debiasing, significant framing effects were present, but that evidence rating significantly mitigated the framing effect. In this auditing context, the framing bias appears to be easily induced, but is not robust. Although the profession should be aware of this potential problem, effective remedial or proactive steps can be easily implemented and may naturally occur in current practices.

Abstract. This paper investigates how strategic trading around the time of earnings announcements affects market liquidity (e.g., bid‐ask spreads). We model an investor with private information in advance of an earnings announcement (e.g., inside information). The investor trades before and after the earnings announcement in a market populated by liquidity‐motivated traders who have some discretion over the timing of their trades. The main result of the analysis is that an earnings announcement that reduces an insider's private information may lead to a less liquid market in the postannouncement period.