Big 5가 대부분의 시장을 차지하던 회계감사업계가 Arthur Andersen의 몰락 이후 Big 4로 재편된 이후, 규제기관에서는 4개 뿐인 대형 회계법인들이 감사시장을 과점하면서 큰 영향력을 발휘한다고 우려하고 있다. 과점시장에서는 감사인들 간 경쟁의 부재로 피감기업들이 동일한 감사인을 계속해서 고용할 수 밖에 없어서 계속감사기간이 길어지며, 그 결과로 감사인의 독립성이 저해된다는 것이다. 또한, 이러한 문제점을 막기 위해서는 감사인 강제교체제도를 도입해서 감사인을 자주 교체해야 한다는 주장도 제시되고 있다. 이러한 주장의 진위를 확인하기 위해 본 연구에서는 1994년부터 2012년까지 국내시장의 자료를 이용하여 산업별 감사인 집중도가 해당 산업에서의 감사인 평균 교체빈도 및 평균 계속감사기간으로 측정한 감사시장의 경쟁정도와 관련되어 있는지를 검증하였다. 실증분석 결과 규제기관의 예측과 일치하게 산업별 감사인 집중도가 높을수록 감사인 교체가 덜 빈번하게 일어나고 감사인의 계속감사기간이 길어지는 현상이 발견되었다. 이런 발견은 감사인 강제교체제도의 도입을 둘러싼 논란에 여러 흥미있는 시사점을 제공한다.
As a result of the Enron-related accounting scandal occurred in year 2001, Arthur Andersen collapsed. After the demise of Arthur Andersen, the audit market became dominated by the remaining 4 large auditors – Deloitte, Ernst & Young, KPMG, and PricewaterhouseCoopers and thus audit market concentration increased dramatically. Among the several critics on Arthur Andersen for the audit failure, some criticize that Arthur Andersen audited Enron for many years in a row, and thus, formed cozy relationship with Enron and lost auditor independence. In addition, after the passage of Sarbanes-Oxley Act in 2002, the audit fees skyrocketed and firms started to complain about the high audit fees. Subsequently, regulators started to ponder whether the concentrated audit market structure, specifically among large firms, is related to audit quality and audit fees. Specifically, General Accounting Office(GAO, 2003) expressed concerns that increased auditor concentration limited clients’ choice of auditors, reducing the frequency of auditor changes. These problems may lead to long auditor tenure and resulting poor audit quality(Public Company Accounting Oversight Board, 2011). In fact, European Commission(EC, 2010) described that some firms appoint the same auditors for decades without considering for auditor changes. Accordingly, GAO(2003) and EC(2010, 2011) proposed to introduce the mandatory auditor rotation policy to promote more frequent auditor changes. If the policy is introduced, they argue that, the dominance of a few large auditors in the audit market will diminish and audit quality will improve. Further, the regulators believe that the introduction of the mandatory auditor rotation and resulting frequent auditor changes will lower the audit fees. However, practitioners and academics voice against the proposal and claim that the changes will lead to the increase of audit costs (and thus audit fees) as well as the deterioration of audit quality due to the loss of the chance of industry-specific or firm-specific knowledge accumulation through the provision of audit service to a specific client for long periods(Government Accountability Office 2008). In summary, the regulators’ argument is that audit market concentration leads to the decrease in the auditor competition, manifested by less frequent auditor changes and longer auditor tenure (i.e., first link), and this leads to poor audit quality and higher audit fees (i.e., second link). We are not aware of any prior literature that investigates the first link – whether audit market concentration is related to the frequency of auditor switches and the length of tenure. In contrast, there are some studies that link audit market concentration directly to audit quality(e.g., Kallapur et al. 2010; Boone et al. 2012; Newton et al. 2013; Cho et al. 2014), that is, examine the second link. But these studies yield mixed results depending on the proxies of audit quality they used. Thus, it is not yet easy to draw any definitive conclusion whether the regulators’ argument is right. This study looks into the first link in the regulators’ argument empirically to solve this important question. Surprisingly, as mentioned above, the first link is ignored in prior related studies and they directly focus on the relation between audit market concentration and audit quality/fees. Thus, it is not yet clear whether the first link exists in the audit market. That is why this study focuses on the first link and seeks to provide empirical evidence on the concerns of the regulators on the potential effect of auditor concentration. In contrast to the arguments of regulators, there are actually several counter-arguments that auditor concentration is not necessarily related to the level of auditor competition(e.g., Davies and Geroski 1997; Dedman and Lennox 2009; Causholli et al. 2010). Thus, it is an interesting empirical question to check whether regulators’ concerns actually occur in Korean audit market. With a maximum of 10,448 observations collected for the period starting from 1994 to 2012, we empirically test this prediction – the relationship between auditor concentration and the frequency of auditor switches and the length of auditor tenure. For the estimation of auditor concentration, we adopt a frequently-used measure of Herfindahl index estimated on the basis of client total assets and sales for each industry and year. We then measure the frequency of auditor switches as the ratio of the number of auditor switch observations out of the total number of observations in the industry and year and the auditor tenure as the industry-year average length of the auditor tenure. In addition, we divide the time period to 3 sub-samples – the period of Big 6(1994-1998, 2001-2004), Big 5(1999-2000), and Big 4(2005-2012). Empirical findings are summarized as follows. First, we find that the auditor concentration is negatively associated with the industry average of the frequency of auditor switches and positively associated with the industry average of auditor tenure. These findings are consistent with the concerns of regulators. Second, these findings are robust in three sub-periods, suggesting that the findings are not restricted to certain periods only. However, we find that the relationships are relatively weak in the recent Big 4 period. The findings in this study provide important implications. Specifically, the findings suggest that regulators may introduce a mandatory auditor rotation policy in certain concentrated industries to promote more frequent auditor switches. But, given the empirical evidence on the potential negative effect of frequent auditor switches on audit quality, regulators need to pay more attention on maintaining high audit quality in the concentrated industries in the case of the introduction of the policy. In this respect, the findings of this study can be used as a valuable input to policy-makers.