본 연구는 감사인의 입장에서 피감사기업에 대한 경제적 의존도 즉, 고객중요성이 감사인의 외관상 독립성에 미치는 영향을 분석하였다. 또한, 고객중요성과 감사인의 독립성 간의 관계가 감사인의 종류에 따라 차별적으로 나타나는지를 살펴보았다. 감사인의 외관상 독립성은 합리적인 제3자가 감사인의 독립성이 훼손된 것으로 판단하게 만드는 중요한 사실이나 상황을 회피하는 것으로 정의된다. 이에 본 연구는 감사인의 외관상 독립성을 이익반응계수로 측정하였다. 이익반응계수는 투자자들의 회계이익에 대한 평가를 나타내므로 감사인 독립성에 대한 자본시장의 인식 즉, 감사인의 외관상 독립성을 측정하기에 적합하다. 또한, 본 연구에서 감사인의 경제적 의존도는 회계법인의 총수입에서 특정 피감사기업의 감사보수가 차지하는 비중으로 측정하였다.
분석 결과 첫째, 고객중요성이 증가할수록 감사인의 외관상 독립성이 훼손되는 것으로 나타났다. 둘째, BIG4 감사인으로부터 감사를 받은 표본과 Non-BIG4 감사인으로부터 감사를 받은 표본 간 피감사기업의 특성에 유의한 차이가 존재하므로 성향매칭점수기법(PSM) 분석을 수행하여 피감사기업의 특징을 통제하는 경우, BIG4와 Non-BIG4 간에 고객중요성에 따른 외관상 독립성 차이는 없는 것으로 나타났다. 다만, 전체 표본을 대상으로 수행한 분석에서는 BIG4와 Non-BIG4 간 고객중요성과 외관상 독립성의 관련성 차이는 유의적으로 나타난 바, 이는 감사인 선택의 내생성을 충분히 통제하지 못함에 따라 나타난 피감사기업들간의 특성차이로 이해된다.
This study examines the association between the economic dependence of auditors on clients (client importance) and auditor independence-in-appearance using earnings response coefficients (ERCs) as a proxy for investor perceptions of earnings, and therefore audit quality. More importantly, this study investigates whether the auditor type affects the relation between client importance and audit quality.
Auditors must be independent in fact and in appearance. Previous research which examines the effect of client importance on auditor independence-in-appearance uses ERCs as a proxy for independence-in-appearance while the prior studies which examine the effect of fee dependence on auditor independence-in-fact uses audit opinion and earnings management measure such as discretionary accruals as a proxy for independence-in-fact. Following prior literature, we use ERCs as a measure of independence-in-appearance. We compute ERCs from regressions of 15-month market-adjusted stock returns ending three months after fiscal year-end on annual earnings (levels and changes). Also, this study measures the economic fee dependence by the ratio of audit fee (or total fee including non-audit fee) paid by each client over total revenue of the auditors, which means the significance of a client relative to the auditor’s portfolio of clients. Fee data is from the annual report of audit firms to measure the client importance more correctly. International Federation of Accountants (IFAC) has revised code of ethics to reinforce auditor independence requirement since June, 2005 and revised again in July, 2009 to limit the scope of non-audit services that could be provided by auditors. Korean Institute of Certified Public accountant (KICPA) also enacted code of conduct in July, 2006 and revision is in process to reflect the revision of IFAC’s restriction on non-audit services provided by auditors in 2009. The amendment includes constraint on total fees from each client, requiring pre-issuance review or post-issuance review by the regulators or other auditors in case that total fees from the client are greater than 15% of total revenues of audit firms. This revision posits that higher client importance impairs the auditor independence.
While prior studies investigated whether client importance impair auditor independence, however, they did not provide consistent results. Especially, in Korea, there is few evidence that fee dependence affects independence-in-appearance. Using 7,283 Korean data between 2005 and 2010, the results suggest that economic fee dependence is on average negatively associated with ERCs, demonstrating that investors are concerned about audit firms’ fee dependence on clients while economic fee dependence is not significantly associated with earnings management, independence-in-fact, in prior studies. On the other hand, the significant negative relationship between client importance and ERC holds only for Non-BIG4, consistent with the view that BIG4 auditors, because of reputation and litigation costs, have stronger incentives to maintain higher audit quality, however, Non-BIG4 auditors depend more on small number of clients compared with BIG4 auditors. The results are robust when we use fee dependence measure based on total fees, instead of audit fees.
However, we employ propensity score matching (PSM) models in an attempt to control for differences in client characteristics between BIG4 and Non-BIG4 auditor groups. The results show that client importance is negatively associated with independence-in-apperance for Non-BIG4 auditors, however, the effect of client importance on independence-in-apperance is not significant and the relationship between client importance and independence-in-appearance of BIG4 auditors are insignificantly different form that of Non-BIG4 auditors. We use 4,922 company- years for propensity-score matching using 3% caliper matching method and the result holds when we use nearest matching. The findings from the propensity-score matched sample suggest that the BIG4 effect on the relationship between client importance and auditor independence is from reflection of their respective clients' characteristics. In other words, BIG4 does not affect the association between client importance and auditor independence. In addition, when we examine the effect of client importance on independence-in-fact using the same sample confirms that fee dependence does not affect independence-in-fact, but affects investor's perceptions of auditor independence.
Our study contributes to the literature on auditor independence by examining whether the client importance affects auditor independence-in-appearance in Korea. Although prior studies provide some evidence on independence-in-appearance in US, they do not examine the effect of client importance on independence-in-appearance for Non-BIG4 clients. By examining fee dependence on auditor independence using propensity score matching, our paper provides additional market-based empirical evidence that difference in market perception might be attributable to client characteristics. We expect that the results would be helpful for policy makers and regulators in that auditor independence requirement is on debate since 2000's to improve accounting transparency.