감사인 교체연도의 이익조정 수준과 전임 감사인의 계속감사기간 사이의 관계

The accounting scandals of Enron occurred in year 2001 greatly influence world economy. One of the reason for the audit failure of Arthur Andersen to fail to find the cooked book of Enron is regarded as the extended auditor tenure of Arthur Andersen (22 years in a row) to audit Enron consecutively. Thus. after the Enron incident, auditor tenure has been received grater attention. Regulators called for the research on the effect of the auditor tenure on audit quality as well as financial reporting quality. They worry the potential audit quality impairment due to prolonged auditor tenure which could result in close relationship between auditor and client. In contrast, academic literature on auditor tenure has documented that longer auditor tenure is associated with lower levels of earnings management and more propensity to issue going-concern audit opinion. This finding is in sharp contrast with the popular beliefs of regulators but is consistent with the claim that auditors can gain firm-specific expertise over time. which enhances audit quality. This paper extends prior research by relating issues in auditor tenure and those in auditor change. This paper examines the association between the magnitude of earnings management by the client firms at the first year with a successor auditor and the tenure of the predecessor auditor in case of the auditor change. If successor auditor evaluates audit risk of the client high as the tenure of predecessor auditor increases. and if the successor auditor wants to avoid significant decline in audit quality. the successor auditor will increase audit effort. This will lead to lower level of earnings management. which is similar to the level of the earnings management before the auditor change. As a result. there would be significant relation between the tenure of predecessor auditor and the level of earnings management at the first year of auditor change. The sample is 384 firm-years which have changed auditor during 2002-2006. We chose the year of 2002 for the start of the sample period in order to avoid the potential effect of the mandatory auditor retention and rotation policies which became full effects starting from 2002. There is a great variety in he auditor tenure that tenure of prior auditor ranges from 1 year 05.89%) to more than 13 years (5.72%). The average of the tenure is about 4.79 years and he median of the tenure is 3 years. We use discretionary accruals estimated from modified Jones model as a proxy for audit quality. Our model examines the relation between absolute values of discretionary accruals at the first year with new auditor and prior auditor's tenure after controlling for other factors expected to affect the accruals. For our main analyses. we use the absolute value of discretionary accruals. which is the consistent method that used in prior studies. However. we also use he discretionary accruals after separating the full sample into sub-samples depending on the sign of the accruals in the sensitivity analyses. The empirical results support our hypothesis that the magnitude of earnings management at the first year with a successor auditor is negatively associated with the length of prior auditor's tenure. Specifically. the absolute value of discretionary accruals decreases with the tenure of prior auditor at a diminishing rate. Further test on signed (positive and negative) accruals reveals that new auditor constrains income-increasing accruals. Additional analysis suggests that the association is mainly driven by auditor change between a Big 4 and a non-Big 4 auditors, either the change from a Big 4 to a non-Big 4 auditor or the change from a non-Big 4 to a Big 4 auditor. The change from a Big 4 auditor to another Big 4 auditor or that from a non-Big 4 auditor to another non-Big 4 auditor is not significant in the relation. Compared to that of last year with prior auditor. the magnitude of earnings management at the first year with successor auditor has increased. but not significantly. Finally, we examine the association between tenure with prior auditor and audit fees charged by new auditor. We find a positive relation between length of tenure and audit fees for sub-samples which have tenure more han 3 years. This results indirectly supper the argument that new auditor considers extended tenure with prior audit as additional risk factors. Our results have several policy implications for 6-year mandatory auditor change rules. Korea's regulation requires mandatory auditor rotation every 6 years to prevent any detrimental effect on auditor independence from extended auditor-client relationship, However, some are raising concerns that this mandatory rotation will decrease the audit quality in the early years of an engagement since a new auditor lacks sufficient knowledge regarding firm-specific risks, This paper's findings suggest that, contrary to this concern, the high audit quality resulting from auditor-client relation over long time does not drop sharply at the first year with new auditors because a new auditor views long tenure with the prior auditor as a source of increased audit risk and increases the audit efforts. In addition, the findings also suggest that the prolonged auditor tenure could lower the level of earnings management even after the auditor change. Thus, it implies that the longer auditor tenure could contribute audit quality. whereas the frequent auditor change actually impairs the audit quality.