The Effect of Abnormal Audit Hours on Earnings Persistence and Discretionary Accruals: Using the Rank-Specific Audit Hours Data*

Abstract

This study examines how abnormal total audit hours and rank-specific audit hours affect earnings quality, proxied by earnings persistence or discretionary accruals. The sample is collected from listed firms in Korea between 2014 and 2019. In particular, this study uses rank-specific audit hours which have been publicly disclosed under the External Audit Act Amendment in 2014 in order to estimate abnormal audit hours for five different ranks: audit quality reviewer, director, registered accountant, associate accountant, and expert in other areas such as tax, IT audit, and asset valuation. Our major findings are fourfold. Firstly, abnormal total audit hours as well as abnormal audit hours for audit quality reviewer, registered accountant and associate accountant are negatively associated with earnings persistence. Secondly, we do not observe any significant relations between discretionary accruals and abnormal audit hours, total or rank-specific. Thirdly, the results stay qualitatively similar when discretionary accruals are replaced with loss avoidance (i.e., reporting small positive earnings), while absolute discretionary accruals appear to be positively related to abnormal audit hours by two positions (audit reviewer and director) as well as abnormal total audit hours. Lastly, results based on log of audit hours (total or rank-specific) are sensitive to the inclusion of firm size as a control variable in the regression model whereas those based on abnormal audit hours are relatively stable regardless of the presence of firm size in the control. Our results suggest that audit hours may act as an indicator of audit risk rather than audit quality.