Mandatory Audit Firm Rotation and Audit Quality

Abstract

In a setting where mandatory audit firm rotation has been effective for more than 20 years (i.e. Italy), we analyse changes in audit quality during the auditor engagement period. In our research setting, auditors are appointed for a three-year period and their term can be renewed twice up to a maximum of nine years. Since the auditor has incentives to be re-appointed at the end of the first and the second three-year periods, we expect audit quality to be lower in the first two three-year periods compared to the third (i.e. the last) term. Assuming that a better audit quality is associated with a higher level of accounting conservatism, and using abnormal working capital accruals as a proxy for the latter, we find that the auditor becomes more conservative in the last three-year period, i.e. the one preceding the mandatory rotation. These results are confirmed using Basu's [1997. The conservatism principle and the asymmetric timeliness of earnings. Journal of Accounting and Economics, 24(1), 3–37] timely loss recognition model. In an additional analysis, we use earnings response coefficients as a proxy for investor perception of audit quality, and we observe results consistent with an increase in audit quality perception in the last engagement period.

Acknowledgements

We would like to thank the Associate Editor, Ann Vanstraelen, and two anonymous reviewers for their guidance and constructive comments. We also thank Bocconi University for financial support.

Notes

¹The most recent draft agreement between Parliament and Council states that ‘auditor can inspect a company's books for a maximum 10 years, which may be increased to 10 additional years if new tenders are carried out, and by up to 14 additional years in the case of joint audits’. The Parliament vote on the issue is expected in April 2014.

²In a first stage, only the largest listed companies were obliged to comply with it (Consob, 1992).

³According to the national regulation, audit firms who audit listed companies ‘may provide services limited to the accounting organization of the firms, as well as auditing services’ (Cameran, 2007, p. 155).

⁴Moreover, Cameran (2007) reports that auditing services account for about 90% of revenues of Big audit firms in Italy. Considering the fact that more than 90% of Italian listed companies are audited by Big audit firms (Cameran, 2005), we can assert that financial reporting represents the primary concern of auditors in charge of auditing Italian listed companies.

⁵‘For example […] an audit firm may be sanctioned by the regulators to suspend or cease practice; it may be required by a government agency (e.g. the SASAC) to rotate off the client; or it may have self-liquidated. […]’ (Firth et al., 2012, p. 118).

⁶In an attempt to overcome this limitation, some papers have tried to model a MAR setting on a theoretical basis, but again the conclusions are conflicting (Arruñada & Paz-Ares, 1997; Elitzur & Falk, 1996; Gietzman & Sen, 2002).

⁷This is consistent with what was reported by PCAOB (2011, p. 17): ‘an auditor that knows its work will be scrutinized at some point by a competitor may have an increased incentive to ensure that the audit is done correctly’.

⁸Note that there are no ‘early adopters’ of IFRS in our sample.

⁹Calepino dell'azionista is a yearly publication by one of the major Italian financial institutions (Mediobanca) that contains financial information – including main financial statement data - on all companies quoted on the Italian stock exchanges.

¹⁰AIDA is the Italian version of Amadeus provided by Bureau Van Dijk, which contained, at the time of our data collection, comprehensive information for more than 500,000 Italian companies, included listed ones.

¹¹For example, this is true in cases of companies that acquire other firms. In such a case, the accrual data related to the year of acquisition are excluded because of the lack of comparable data from the previous year.

¹²We excluded those companies that changed the audit firms at the end of the first or the second three-year engagement period.

¹³Note that over 94% of our initial sample is audited by Big-N audit firms.

¹⁴Although small in size – our sample represents fairly well the underlying population of non-financial listed companies. Indeed, with few exceptions (e.g. construction, electronics, utilities), the industry distribution for the latter is fairly similar to the one in our sample, that is: food and beverage = 4.0%; automotive = 5.9%; chemical = 12.4%; construction = 10.6%; electronics = 13.9%; machinery = 5.2%; textile = 12.2%; media = 5.3%; utilities = 8.8%; transportation-tourism = 9.0%; new economy = 8.5%; and miscellaneous = 4.0%.

¹⁵These characteristics of the Italian setting are consistent with those reported in more recent studies on Italian listed companies, e.g. Prencipe and Bar-Yosef (2011).

¹⁶We use for robustness also the Tamhane's T2 test. The results are qualitatively similar to those reported for the Games–Howell test.

¹⁷However, we also performed our analysis on a sample of raw accruals observations (1067) made up of the sum of the positive accruals (565) and negative accruals (502) observations. The results (untabulated) are qualitatively similar to those presented in Table 5.

¹⁸Indeed, a coefficient of −0.012 indicates that when the variable PERIOD_3 is equal to 1, the dependent variable (i.e. AWCA deflated by sales) decreases by 1.2% compared to PERIOD_1, taken as a benchmark.

¹⁹Note that the median (mean) ROS in the sample is about 4.3% (3.5%); therefore, a decrease of 1.2% has a significant effect on it.