Knowledge that Transforms

This paper investigates whether client importance affects auditor independence within the local offices of audit firms. Client importance is measured as the proportion of audit fees, non-audit service fees, or total fees that a distressed, public client contributes to the total public-client revenue earned by the individual audit offices. Auditor independence is measured as the auditor's propensity to issue a going-concern opinion. The paper focuses on changes in the relation between fee ratios and auditor reporting decisions from the pre-SOX (2001) to post-SOX (2003) period. In the pre-SOX period, I do not find statistically significant association between any of the fee ratios and the auditor's propensity to issue a going concern opinion. However, in the post-SOX period, I find evidence that higher audit fee and total fee ratios are positively associated with the auditor's propensity to issue a going concern opinion. That is, post-SOX, relatively more important clients are more likely to receive a going concern opinion. These results allay concerns that auditor independence is compromised for significant clients.

Prior research reports seemingly conflicting evidence and interpretations concerning the relation between dispersion in analysts' earnings forecasts and stock returns. Diether et al. (2002) and Johnson (2004) find a negative relation between levels of dispersion in analysts' forecasts and future stock returns. Yet, changes in forecast dispersion are negatively associated with contemporaneous stock returns (L'Her and Suret 1996). We demonstrate that levels and changes in dispersion reflect different theoretical constructs. Changes in dispersion primarily reflect changes in information asymmetry whereas levels of dispersion primarily reflect levels of uncertainty. Further, the uncertainty component of dispersion levels reflects idiosyncratic risk that is negatively associated with future stock returns. These findings provide support for Johnson's (2004) explanation that dispersion levels reflect idiosyncratic uncertainty that increases the option value of the firm and generally refute Diether et al.'s (2002) explanation that dispersion levels reflect information asymmetry. In addition, we reconcile L'Her and Suret's (1996) findings with the findings of Johnson (2004). We find that the negative association between changes in dispersion and contemporaneous stock returns is not due to increased uncertainty but rather increased information asymmetry.

Impact of the type of audit team discussions on auditors' generation of material frauds