잠정이익과 보고이익의 보수주의 및 감사효과

Preliminary earnings are the earnings disclosed by managers with board of directors’ approval before audit, and audited earnings are that after the audit. A CPA is supposed to correct and to recalculate the earnings again if accounting treatment of preliminary earnings has any disobedience in GAAP. As the CPA corrects the preliminary earnings, there is difference between preliminary earnings and audited earnings. The difference between them is that solely from CPA's audit without any involvement of other information. The CPA's audit effects may have three cases: 1) the case that audited earnings are less than preliminary earnings, 2) the case that preliminary earnings and audited earnings have almost the same, 3) the case that audited earnings are larger than preliminary earnings. In this study, I define the difference between preliminary earnings and audited earnings as audit effect, and inspect the difference of audited effects according to the types of the differences with conservative model. I also indirectly investigate audit quality showing that such an audit effect is reflected in the difference between Big4 and non-Big4. In this study, I analyze the audit effects investigating how the conservatisms in three groups are different from one another according to the types of difference between preliminary earnings and audited earnings from 2000 to 2008. In the case that audited earnings are less than preliminary earnings, audit effect may be meaningful only if conservatism of audited earnings is less than that of preliminary earnings. For the verification, I do the regression analyses introducing grouping-divided dummies into Basu model(1997), and I do F-test for verifying the difference of conservatism between preliminary earnings and audited earnings by groups. The results show that the conservatism of audited earnings is higher than the other groups preliminary earnings with statistical significance and that auditing firm in Korea does conservative audit. In the case of the difference between preliminary earnings and audited earnings, non-conservatism of preliminary earnings is higher than that of other groups when preliminary earnings are greater than audited earnings after the audit and non-conservatism of audited earnings is not significant. This means that auditors significantly devaluate the preliminary earnings when the manager’s preliminary earnings are highly reported. On the contrary, the conservatism of preliminary earnings is higher than that of other groups when the preliminary earnings are almost same with audited earnings, but conservatism is not observed in the audited earnings after the audit. It can be understood as auditing is done for the adjustment of conservatism which goes against GAAP when the preliminary earnings are almost same with audited earnings. Compared with other groups, I cannot find any conservatism in preliminary earnings and audited earnings in the case that preliminary earnings are smaller than audited earnings. This means that accounting information is not distorted even if auditor makes audited earnings greater through the adjustment of preliminary earnings. The results of additional test to verify the different audit effects between Big4 and non-Big4 show that Big4 makes lower non-conservatism of preliminary earnings when preliminary earnings are greater than audited earnings, that Big4 makes the more conservative audit than non-Big4 even in the case that they are almost the same with each other, and that there is no difference between Big4 and non-Big4 when preliminary earnings are smaller than audited earnings. From these, I summarize that Big4 prefers the more conservative accounting treatment compared with non-Big4. Therefore we can say that Big4’s audit quality measured with conservatism is higher than that of non-Big4. Measuring the audit effects through conservative difference between preliminary earnings and audited earnings, this study indirectly proves that auditor effectively adjusts managerial discretion with accounting audit even if managers make conservative or non-conservative reports on preliminary earnings. This study makes points from the fact that I measure audit effects with conservatism between preliminary earnings and audited earnings, and that I indirectly prove audit quality improvement through audit effects.