31-10-2016, 02:14 PM
Audit firm rotation, audit fees and audit quality: The experience of Italian public companies
1462476714-AuditfirmrotationauditfeesandauditqualityTheexperienceofItalianpubliccompanies.pdf (Size: 721.59 KB / Downloads: 35)
a b s t r a c t
This paper examines some ofthe costs and benefits associated with audit firm rotation using
data from Italy, where mandatory audit firm rotation has been in place since 1975. Previous
studies in this area did not find consistent evidence of an association between audit quality
and voluntary or mandatory audit firm rotation. A recent paper, examining Italian public
companies audited by a Big 4 audit firm, uses proprietary data and finds no statistically
significant association between audit firm rotation and audit quality. In this study, we handcollect
publicly available data for a larger sample of Italian public companies audited by a
Big 4 and non-Big 4 audit firm (1583 firm-year observations) over a longer time horizon
(1998–2011). We find that audit quality, proxied by two different measures of earnings
management, improves following audit firm rotation for companies audited by a non-Big 4
audit firm. Additionally, we examine whether higher audit fees are associated with audit
firm rotation. Our results indicate that following audit firm rotation, the total amount of
fees paid to the auditor was lower for companies audited by a Big 4 and unchanged for
companies audited by a non-Big 4 audit firm. The results ofthis study should be of interestto
European and U.S. legislators who are currently, or have recently, considered implementing
mandatory audit firm rotation in order to improve financial reporting quality
. Introduction
Agency theory indicates thatthe separation of management(agent)from ownership (stakeholder)leads to a moral hazard
problem because the agent (management) may pursue his own self-interest at the expense of the principal (stakeholder)
(Jensen & Meckling, 1976). The moral hazard problem is amplified by information asymmetry between the two parties:
managers, who run the company, know more about the company and its future prospects than do shareholders. One way
to reduce the consequences and the costs associated with moral hazard is to hire an external third party – an independent
public accounting firm – to audit the books, records, and financial statements of a company, thereby reducing information
asymmetry between the company’s agents and its principals.
Audit quality is a function of the auditors’ education, training, and knowledge of professional standards, as well as their
independence and objectivity,their knowledge ofthe client’s business operations and industry, and the auditteam’s working relationship with the client company’s management. There are two primary schools of thought regarding long audit firm
tenure. One school believes that audit firms with relatively longer tenure have greater knowledge of the company’s business
and industry, thereby providing a higher quality and more efficient audit (Geiger & Raghunandan, 2002; Johnson, Khurana,
& Reynolds, 2002; Myers, Myers, & Omer, 2003; Carcello & Nagy, 2004). The other school believes that audit firms with
relatively longer tenure provide an increased likelihood of familiarity (or even friendships) forming between the audit staff
members and client staff members, an increased likelihood of a stale audit program, and a decreased likelihood that the
auditor will make decisions contrary to the prior year decisions, thereby providing a lower quality and less efficient audit
(Defond & Subramanyam, 1998; Arel, Brody, & Pany, 2005; Gates, Lowe, & Reckers, 2007; Dao, Mishra, & Raghunandan,
2008; Daniels & Booker, 2009). Interestingly, this latter school of thought is driven primarily by perceptions not empirical
evidence.
In an effort to strengthen auditor independence, many countries have legislated limitations on the auditor–client relationship
including: mandatory audit partner rotation, hiring and firing of the audit firm by the audit committee rather
than management, internal reviews of audit engagements, and external peer or regulated reviews of audit engagements.
Additionally, some countries, including the United States (U.S.) post-Sarbanes–Oxley Act of 2002 (SOX), limit the types of
services a public accounting firm can provide to its audit clients1, and the type of employment an auditor can take with a
client company2.
SOX, Section 203, requires rotation of the partner on an audit engagement of a public company every 5 years, but does
not, currently, require audit firm rotation (Bradshaw & Sloan, 2002). In 2003, the General Accounting Office (GAO) released
the results of a study on the potential effects of mandatory audit firm rotation. The GAO concluded that “mandatory audit
firm rotation may not be the most efficient way to strengthen auditor independence and improve audit quality considering
the additional financial costs and the loss of institutional knowledge of the company’s previous audit firm of record, as well
as the current reforms being implemented.” Public Accounting Firms: Required Study on the Potential Effects of Mandatory
Audit Firm Rotation 2003, p.1). The GAO also suggested a “wait and see” attitude until the other reforms put in place by SOX
were in effect for several years, thereby leaving open the possibility that audit firm rotation would be considered again in
the future (Government Accounting Office (GAO), 2003) Public Accounting Firms: Required Study on the Potential Effects of
Mandatory Audit Firm Rotation 2003).
In August, 2011, the Public Company Accounting Oversight Board (PCAOB) issued a Concept Release (no. 39) on auditor
independence, objectivity, and professional skepticism, including consideration of mandatory audit firm rotation. The comment
period originally expired in December, 2011, but was extended to April of 2012 in order to solicit more feedback. In
total, the PCAOB received 659 comment letters related to this concept release; most letters vehemently opposed mandatory
audit firm rotation. During July, 2013, the Financial Services Committee of the U.S. House of Representatives took the decision
out of the hands of the PCAOB by overwhelmingly passing a bill amending the Sarbanes–Oxley Act of 2002 to prohibit
the PCAOB from requiring public companies to use specific audit firms or requiring public companies to change audit firms
on a rotating basis; the bill also directs the GAO to revisit their 2003 study mentioned above. The bill next will be taken up
by the Senate Committee on Banking, Housing, and Urban Affairs. Interestingly, in April of 2013, the European Parliament’s
Legal Affairs Committee took related action by approving a draft law that would require public entities to rotate audit firms
every 14 years (with a possible extension to 25 years if certain safeguards are in place).
This paper adds to the existing literature regarding mandatory audit firm rotation and also informs both the decision
taken by authorities in the U.S. to end discussion of mandatory audit firm rotation and the seemingly opposite decision taken
by the authorities in the European Parliament. We examine some of the costs and benefits associated with mandatory audit
firm rotation using data from a country, Italy, where mandatory audit firm rotation has been in place since 1975. Italy is
one of the very few countries in the world to mandate audit firm rotation and is, therefore, a unique setting to examine this
topic. Specifically, we test whether there is a change in audit quality associated with both mandatory and voluntary audit
firm rotation. We also test whether there is a change in total audit fees paid to the auditor when there is a mandatory or
voluntary audit firm rotation.
A recent study examining Italian public companies audited by a Big 4 audit firm with private data provided by the Big
4 audit firms (Cameran, Francis, Marra, & Pettinicchio, 2015) found no statistically significant association between audit
firm rotation and audit quality. We first replicate the results of Cameran et al. (2015) using publicly available data and then
extend our sample to include Italian public companies audited by a non-Big 4 audit firm and to examine a longer time period.
Extensive empirical research has shown that earnings quality is different for companies audited by one of the Big 4 audit
firms vs. companies audited by non-Big 4 audit firms. This body of research has examined both U.S. companies (DeAngelo,
1981; Khurana & Raman, 2004) and companies from around the world (Francis & Wang, 2008).
Overall, our results indicate that for companies audited by non-Big 4 audit firms, audit firm rotation is associated with
an increase in audit quality without the added cost of an increase in audit fees. By contrast, for companies audited by Big
4 audit firms, audit firm rotation is not associated with an increase in audit quality but is associated with a decrease in
audit fees; these latter results confirm the findings in previous literature. This study makes several contributions to the
literature. First, it replicates and extends a recent study (Cameran et al., 2015) using publicly available data that include not only public Italian firms audited by a Big 4 audit firm but also companies audited by a non-Big 4 audit firm. Second, our
study expands the existing literature to examine voluntary and mandatory audit firm rotation for both companies with Big
4 and non-Big 4 audit firms. The results of this study indicate that, for non-Big 4 audit firms audit quality improves following
audit firm rotation. Third, this study provides evidence that following audit firm rotation companies with Big 4 audit firms
experience lower audit fees, while companies with non-Big 4 audit firms do not experience a change in audit fees. These
results should be interesting to policy setters and regulators in Italy, policy setters and regulators in countries considering
implementing mandatory audit firm rotation (the European Parliament and others), the U.S. House of Representatives, the
GAO, and academic researchers.
We organize the rest of this paper as follows. Section 2 provides a description of the institutional background, literature
review and hypotheses development. Section 3 describes the sample selection procedures and data collection. Section 4
describes the research design, our measure of audit quality and our test models. Section 5 reports our results. Section 6
describes our sensitivity tests and Section 7 concludes the paper and identifies the limitations of the study.
2. Institutional background, literature review, and hypotheses development
2.1. Legislation regarding mandatory audit firm rotation in Italy
Italy first legislated mandatory audit firm rotation in 1975 and has since made five significant modifications to the
regulations, presumably in an effort to make the legislation as efficient and effective as possible (Table 1).
The first regulation on mandatory audit firm rotation3 provided for an audit firm tenure of three years, renewable, if
desired, for two additional three year terms. Before the end of the three-year appointment, voluntary audit firm changes
are possible under certain conditions. After nine consecutive years, a new audit firm must be appointed and the original firm
must wait a minimum of five years (cooling-off period) before the original firm could be reappointed. This first regulation
was in force until 1997.
After 1997, the original decree was partially modified to no longer explicitly identify the length of the cooling-off
period4. Due to the vagueness in the new legislation, audit firms commonly interpreted that the original audit firm could be
reappointed after only one three year cooling-off term with another audit firm.
The third change in legislation occurred in 20055, in response to the well-known financial scandals of Cirio and Parmalat.
This law modified the Law of Finance (Testo Unico della Finanza, 1998) extending the audit firm term from three years to six
years, decreasing the number of audit firm term renewals from three to two, and explicitly introducing a three year coolingoff
period. Under this new law, each audit firm’s existing term was extended from three to six years, and the maximum tenure was extended from nine to twelve years. This law also regulated audit partner rotation, requiring that the partner in
charge of the audit be replaced by another partner after six consecutive years. A three year cooling-off period is provided
during which the same partner may not be responsible for the audit of a previous auditee and its associates, even if he/she
works for another audit firm.
After just one year, in 2006, the Law of Finance was modified again to extend audit firm tenure from six years to nine
years, renewable after a three year cooling-off period6. A provisional rule was also introduced providing that all unexpired
audit firm terms as of the effective date of the Legislative Decree, with a total audit firm tenure of less than nine years,
could be extended to nine years at the next shareholders’ meeting. No changes were made to the provisions regarding audit
partner rotation.
In January 2010, the fifth legislative change became effective extending the audit partner tenure from six to seven years7.
The audit firm term and cooling-off period were not changed.
Overall these legislative changes, intended to improve audit quality, have increased the audit tenure compared to the
first regulation of 1975, implicitly indicating that the Italian authorities believed that relatively longer tenured auditors were
associated with better audit quality.
2.2. Auditor tenure literature
We detail below some of the previous literature related to audit firm rotation and auditor tenure, please see Stefaniak,
Robertson, and Houston (2009) for a more thorough review of this literature. Several U.S. studies examine the relation
between audit quality and auditor tenure and (voluntary) auditor change. In general, these studies do not support the claims
that long-tenured auditors are associated with lower audit quality, indicating that mandatory audit firm rotation may not
improve audit quality as intended. In these studies, researchers have used several proxies for audit quality: audit opinions
(Geiger & Raghunandan, 2002), discretionary accruals (Johnson et al., 2002; Myers et al., 2003), total accruals (Myers et al.,
2003), persistence of accruals (Johnson et al., 2002), and alleged fraudulent financial reporting (Carcello & Nagy, 2004).
Several recent archival studies also failto support mandatory audit firm rotation. First, Jenkins and Velury (2008) examine
the relation between audit firm tenure and conservatism for U.S. publicly listed firms. Using different measures of conservatism,
these authors find a positive association between conservatism and the length of audit firm tenure. Interestingly,
they find an increase in conservatism between short8 and medium9 tenure auditor–client relationships, and this higher
level of conservatism does not deteriorate for long10 tenure relationships. These results indicate that audit firm rotation
may have an adverse effect on the conservatism of reported earnings due to a short tenure (lower conservatism) condition
being frequently repeated.
In an international study,Jackson,Moldrich, andRoebuck (2008) examine the relation between auditfirmtenure and audit
quality for Australian companies. These authors find that audit quality, proxied by the likelihood of issuing a going concern
opinion, increases as tenure increases. However, when audit quality is proxied by discretionary accruals, no difference
is noted when tenure increases. These results would seem to indicate that mandatory audit firm rotation would not be
associated with improved audit quality in Australia.
Several experimental studies have examined the perception of longer-tenured auditors vs. new auditors. The results of
these studies are mixed. The results in Gates et al. (2007) indicate that MBA students demonstrate more confidence in a
company’s financial statements after audit firm rotation. Other studies find that both the capital and debt markets value
longer-term auditors more than new auditors (Mansi, Maxwell, & Miller, 2004; Ghosh & Moon, 2005). More recently, Kaplan
and Mauldin (2008) perform two experiments designed to test whether non-professional investors in the U.S. hold different
independence-related perceptions for audit partner rotation vs. audit firm rotation. This study proxies independence-related
perceptions with how much of an income decreasing audit adjustment management would be willing to record. Kaplan and
Mauldin find no statistically significant difference in independence-related perceptions between the two rotation conditions,
indicating that, for non-professional investors, audit firm rotation does not seem to be associated with a higher perception of
independence than audit partner rotation. The mixed results of these studies seem to indicate the need for further empirical
testing in this area.
2.3. Mandatory audit firm rotation literature
Mandatory audit firm rotation, over time, may actually preclude selection of the most qualified audit firm. On the other
hand, successor auditors can offer a fresh perspective to the audit of a company. Audit firm rotation offers two advantages
over partner rotation: first, a new partner from a new audit firm may be more willing to contradict judgments made by the
predecessor partner; second, in a mandatory audit firm rotation environment, each partner is aware that his/her judgments will be reviewed by another audit firm in a predetermined period of time. Presumably, either of these circumstances could
lead to improved audit quality.
Several countries currently have mandatory audit firm rotation regulation. Italy has required audit firm rotation since
1975, Brazil since 1999, and Singapore has required audit firm rotation for local banks since 2002. Numerous other countries
including Austria, Canada, Greece, Spain, Slovakia, and Turkey previously required mandatory audit firm rotation and have
since eliminated it due, in part, to increased audit costs (Raiborn, Schorg, & Massoud, 2006; Johnson, 2007). Peter Wyman,
the (then) head of professional affairs for PricewaterhouseCoopers, stated in a 2005 article “There is clear evidence from Italy
and the US that audit firm rotation increases costs to business, creates problems with audit quality in the period immediately
after the change of audit firms, and leads to further consolidation of audit work amongst the largest audit firms” (Wyman,
2005). At the time of Wyman’s article, Italy was, and still is, the only of the EU Member States requiring mandatory audit
firm rotation. Additionally, in Italy, the Bocconi University Report indicated that, while audit firm rotation is associated with
reduced audit quality, it seems to improve public confidence in corporations (Arel et al., 2005).
Limited empirical research has been performed to date in actual mandatory audit firm rotation regimes; one such paper
is Ruiz-Barbadillo, Gómez-Aguilar, and Carrera (2009) which examines audit firm rotation in Spain11. These authors find
that auditors were less likely to issue going concern opinions to financially stressed clients during the mandatory audit firm
rotation regime than in the six years following the mandatory rotation regime. These results would seem to indicate that
mandatory audit firm rotation was not associated with improved audit quality in Spain.
In a recent paper, Cameran et al. (2015) use proprietary data provided by the Big 4 audit firms in Italy to examine the
relation between audit effort (quality) and audit fees for clients of Big 4 audit firms between 2006 and 2009. Their results
indicate that for their sample companies, audit fees in the final year of an engagement were higher than normal, which the
authors attribute to opportunistic pricing. Additionally, the new audit firm appeared to discount their audit fee, even though
more hours were incurred on the engagement, which the authors attribute to low-balling. Lastly, these authors find that
audit quality was lower in the first three years of the audit engagement, as compared to the last years of the previous audit
firm’s tenure. In this study, we use publicly available data and a longer time period to both replicate the results of Cameran
et al. (2015) for companies audited by Big 4 audit firms and to extend current research to examine companies audited by
non-Big 4 audit firms.
Overall, the current empirical studies examining audit quality in a mandatory audit firm rotation environment is both
limited and conflicting, indicating, we believe, a need for further research in this area. This paper provides both additional
empirical evidence of an association between audit quality and mandatory audit firm rotation, and evidence of a relation
between audit fees and audit firm rotation in a mandatory audit firm rotation environment. Due to conflicting results of
studies examining audit firm change and audit quality, we formulate our first hypothesis in the null form:
Hypothesis 1. There is no association between a change in audit firm and audit quality.
2.4. Audit fee literature
Previous literature indicates that higher risk clients will choose higher quality auditors (Datar, Feltham, & Hughes, 1991)
and it is reasonable that audit firms will charge higher fees to higher risk clients (Feltham, Hughes, & Simunic, 1991).
Several empirical studies support the relation between higher (lower) client riskiness, more (less) auditor effort and higher
(lower) audit fees (O’Keefe, Simunic, & Stein, 1994; Pratt & Stice, 1994; Simunic & Stein, 1996; Johnstone & Bedard, 2003). As
higher risk clients are also more likely to have higher earnings management (abnormal working capital accruals), the above
literature supports including an audit fees control variable when modeling earnings management.
Both the audit firm and the company invest significant effort and time (cost) following a change in audit firms. This
impact will be even larger for consolidated entities that require statutory audits in many countries. In a mandatory audit
firm rotation environment, these startup costs are more likely to be spread over fewer years, increasing the overall cost of
the audit function for both the audit firm and the audit client12. Extensive research has documented a relation between audit
firm change and audit fees (Ettredge & Greenberg, 1990; Pearson & Trompeter, 1994; Deis & Giroux, 1996; Simon & Francis,
1988; Cameran et al., 2015; Zain, 2013). A recent paper indicates that the relation between these two variables continues in
the post-SOX period (Huang, Raghunandan, & Rama, 2009). Due to the relation between audit fees and audit firm change,
the above literature supports including audit fees as a control variable when modeling auditor change and reporting quality
to avoid a correlated omitted variable problem.
Our second hypothesis tests whether there is an association between total fees paid to the audit firm and audit firm
rotation. Previous literature suggests that mandatory rotation might increase the auditors’ fees because of the increased
amount of time the new audit firm has to spend to audit a new company. However, the higher startup costs may be spread
over fewer years and might not impact the audit fees paid immediately after the mandatory audit firm change. Hence, we
also formulate our second hypothesis in the null form:
Hypothesis 2. There is no association between a change in audit firm and the total/abnormal audit fees paid to the audit
firm by the company.
3. Sample and data
In order to test our hypotheses, we focus on the population of non-financial companies listed on the Milan Stock
Exchange (Mercato Telematico Azionario, Borsa Italiana’s Main Market) as of December 31, 2011. We then extend our
sample backwards, to 199813.
We manually collect data about audit firm and audit partner changes from corporate annual audit reports14 and shareholders’
meeting reports available on the Borsa Italiana website or directly from the company’s website. We also manually
collect audit fee data between 1998 and 2006 from shareholders’ meeting reports and, for fiscal years ending on or after July
1, 2007, from annual reports15. Finally, we download accounting data for the period 1997–201116 from the AIDA database,
which contains financial information for both listed and unlisted companies in Italy. Our final sample for our main test
consists of 1583 company-year observations.