jackson et al - 2008 - mandatory audit firm rotation and audit quality in australia [mafr]

Design/methodology/approach – Using two measures of audit quality, being the propensity to issue a going-concern report and the level of discretionary accruals, the paper examines the sw

Mandatory audit firm rotation

and audit quality

Andrew B Jackson School of Accounting, The University of New South Wales, Sydney, Australia

Michael Moldrich KPMG, Sydney, Australia, and Peter Roebuck School of Accounting, The University of New South Wales, Sydney, Australia

Abstract Purpose – The purpose of this paper is to investigate the effect that a regime of mandatory audit firm rotation would have on audit quality.

Design/methodology/approach – Using two measures of audit quality, being the propensity to issue a going-concern report and the level of discretionary accruals, the paper examines the switching patterns of clients in their current voluntary switching capacity, and the levels of audit quality Findings – The main finding is that audit quality increases with audit firm tenure, when proxied by the propensity to issue a going-concern opinion, and is unaffected when proxied by the level of discretionary expenses Given the additional costs associated with switching auditors, it is concluded that there are minimal, if any, benefits of mandatory audit firm rotation.

Research limitations/implications – A limitation of this study is that only actual audit quality is examined While the results suggest that actual audit quality is associated with the length of audit tenure, the perception of audit quality is not addressed, which may increase with audit firm rotation Originality/value – The results go against the move towards mandatory audit firm rotation, and suggest that other initiatives may need to be considered to address concerns about auditor independence and audit quality.

Keywords Auditing, Auditors, Laws and legislation, Australia Paper type Research paper

Introduction Considerable research has examined the effect that the length of the auditor-client tenure has had on audit quality, but commonly fails to consider the financial characteristics of clients in the years preceding a switch This study examines the financial attributes of clients in the years preceding and succeeding an audit firm switch, as well as voluntary auditor switching behaviour to determine whether such auditor movements would be favourable under a scheme of mandatory rotation The ability of the top-tier audit firms and industry-specialists to provide consistently high levels of audit quality is also studied

www.emeraldinsight.com/0268-6902.htm

The views expressed in this paper are not necessarily the views of KPMG

Financial assistance for this study was kindly provided by the Capital Markets and Co-operative Research Centre and the Australian School of Business, UNSW The authors would also like to acknowledge the helpful comments of participants at the 2005 Faculty of Commerce and Economics UNSW National Honours Colloquium, Liz Carson, Jeff Coulton, Asher Curtis, Rob Czernkowski, Andrew Ferguson, Gerry Gallery, Caitlin Ruddock, Roger Simnett, Stephen Taylor, and two anonymous reviewers All errors remain the authors’ responsibility

MAJ

23,5

420

Managerial Auditing Journal

Vol 23 No 5, 2008

pp 420-437

q Emerald Group Publishing Limited

0268-6902

Following recent legislative changes imposed on the auditing profession through

the introduction of the Corporate Law Economic Reform Program (Audit Reform &

Corporate Disclosure) Act 1999 (CLERP 9), there is a need to determine whether the

current regulations are enough, or whether further regulatory changes, such as a

system of mandatory audit firm rotation, are desirable There has also been a call for

further research on this topic by both the international standard setters and academics

(Government Accountability Office – GAO, 2003; Nagy, 2005) This study is further

motivated by a lack of research examining characteristics of the client in the years

prior to the audit firm switch, which may have an impact on subsequent audit quality

The aim of this study is to understand whether the client’s financial characteristics

under the predecessor auditor have an impact on audit quality under the incumbent

auditor An additional aim is to determine whether current voluntary switching

patterns would be conducive to a system of forced rotation, and whether the top-tier

audit firms and industry specialists are able to provide higher levels of audit quality

over a period of time

We find that audit quality increases with audit firm tenure, when proxied by the

propensity to issue a going-concern opinion, and is unaffected when proxied by the

level of discretionary accruals (DA) Given our results, we conclude that mandatory

audit firm rotation will not improve audit quality Considering the costs involved

during the early stages of an auditor-client relationship, requiring firms to rotate their

auditors will place unnecessary costs on both the auditor and the client with minimal

benefits In order to address concerns that have arisen surrounding auditor

independence issues, we conclude that other initiatives are more likely to have a

greater impact than mandatory audit firm rotation

Background and empirical predictions

Audit firm rotation

A system of mandatory audit firm rotation would require companies to rotate their

independent auditor periodically Currently, listed companies in Italy and Brazil are

required to rotate their independent auditor every nine and five years, respectively In

Australia, there is currently no legislative requirement for reporting entities to rotate

their independent audit firm, although periodic rotation of lead audit partners is now

obligatory under s324DA of the Corporations Act 2001 (Cth)

Advocates of audit firm rotation believe a scheme of compulsory rotation would

prevent auditors from becoming too aligned with managers, impacting on their

independence A client may be a significant source of revenue for an auditor, and the

auditor may be reluctant to jeopardise this revenue stream (Hoyle, 1978) Firm rotation

may also help to prevent large-scale corporate collapses Morgan Stanley estimates the

market capitalisation loss of the collapses of WorldCom, Tyco, Quest, Enron and

Computer Associates to be $US460 billion Firm rotation can help restore confidence in

the regulatory system, which was found to be the case in Italy (Healey and Kim, 2003)

Further, if a client seeks a new auditor, auditors will compete with other audit firms to

win the tender and differentiate themselves in terms of service, improving audit quality

(Hoyle, 1978) Despite the increased start-up costs which are involved with introducing

a new auditor, supporters of audit firm rotation propose that the costs of corporate

collapses, which may not have occurred had audit quality been higher, outweigh the

increase in audit costs involved when introducing a new auditor From this

Audit firm rotation and audit quality

421

perspective, a new auditor brings in more objectivity as they are not familiar with the client, potentially improving the quality of the audit

On the contrary, mandating firm rotation would lead to a loss of client knowledge when the auditor is forced to resign Auditors experience a significant learning curve with new clients (Knapp, 1991), and much of the knowledge acquired during an audit is client-specific (Kinney and McDaniel, 1996) Audit failures are generally higher in the first years of the auditor-client relationship as the new auditor becomes familiar with the client’s operations (Arel et al., 2005) Audit costs would also rise due to the additional work needed by the new audit firm The GAO (2003) estimated that companies would incur additional auditor selection costs equal to 17 per cent of their first year audit fees (GAO, 2003) Opportunity costs would also arise because of a mismatch between the client’s needs and those which the auditor can offer (Arrun˜ada and Paz-Ares, 1997) Auditor resignations provide valuable signals to the market (Wells and Loudder, 1997) Under mandatory rotation, if a client is experiencing conflicts with its auditor over accounting treatments and the auditor is forced to rotate, the market misses out on valuable signals that would have taken place under voluntary rotation The largest accounting firms may also increase their market share under mandatory rotation, as has been the case in Italy (Buck and Michaels, 2005), leading to a less-competitive environment

Even if research has generally not found significant positive effects of firm rotation

on audit quality, rotation may nevertheless be effective in increasing perceived audit quality Audit quality comprises actual and perceived quality (Taylor, 2005) Actual quality is the degree to which the risk of reporting a material error in the financial accounts is reduced, while perceived quality is how effective users of financial statements believe the auditor is at reducing material misstatements Higher perceived audit quality may then help promote investment in audited clients

Related literature and empirical predictions

It has been argued that companies switch auditors to avoid receiving qualified audit reports This argument assumes that managers dislike qualified reports and that manager’s influence the appointment decision

The first assumption is relatively uncontroversial A qualified report may signal to investors that managers are poor stewards of the company’s affairs, or that managers have attempted to present an over-favourable view of the company’s performance and/or position In addition, qualified reports cause share prices to fall, reducing managerial utility if managers own shares or if their compensation is directly related to market value (Firth, 1978; Banks and Kinney, 1982; Fleak and Wilson, 1994; Chen and Church, 1996; Jones, 1996)

The second assumption is more controversial By law in Australia, auditors are appointed by shareholders However, managers may exert considerable influence over auditor appointments Incumbent auditors could be dismissed by managers without consulting shareholders, with the shareholders merely voting on whether to accept their recommendation regarding the appointment of a new auditor or the re-appointment of the incumbent Secondly, managers have some influence over the appointment of a new auditor or the re-appointment of the incumbent auditor Managers set meeting agendas when auditor appointments are proposed, and they

MAJ

23,5

422

typically have the proxy votes of a large number of shareholders Thus, managers have

considerable influence over auditor hiring and firing

If managers dislike qualified reports and have some influence over auditor

appointment, they may try to use auditor switching to avoid receiving qualified

reports Teoh (1992) usefully identifies two ways in which this could occur First, a

manager may actively use the auditor switch decision to avoid receiving a qualified

report If a new auditor is less likely to give a qualified report compared to the

incumbent auditor, the manager may choose to switch Similarly, if a new auditor is

more likely to give a qualified report compared to the incumbent, the manager may

choose not to switch Second, if auditors earn client-specific rents, a manager may

obtain a more favourable report from an incumbent auditor by threatening to switch to

a new auditor However, post-Enron, HIH and the CLERP Act 2004, the top 300 listed

companies are required to have an audit committee, which limits the ability for

managers to exercise such influence

Geiger and Raghunandan (2002) find that auditors are more likely to issue an

unqualified audit opinion prior to a client filing for bankruptcy in the early years of the

auditor-client relationship Myers et al (2003) on the other hand find that as auditor

tenure increases, auditors place greater constraints on the degree of discretionary and

current accruals allowed by management These results suggest that mandatory audit

firm rotation may have adverse effects on audit quality, as audit quality is lower in the

earlier years of the auditor-client relationship

A number of other reasons have been proffered as to why clients decide to change

auditors Chow and Rice (1982) find that the incidence of a qualified report was a

significant reason for clients to switch auditors Schwartz and Menon (1985) find that

failing clients had a greater propensity to switch auditors, while Williams (1988) argues

that financially distressed clients have greater incentives to change auditors than healthy

clients in order for managers to portray a good image of the company Hence, there is some

evidence that a large proportion of switching clients may be financially distressed

However, a change in auditor is not always initiated by the client, but rather may be

initiated by the audit firm Prior research indicates that audit firms by virtue of their own

internal quality procedures have a tendency to shed risky clients An audit firm with a

risky client may decide to drop the client from its portfolio in order to reduce their

engagement risk ( Johnstone and Bedard, 2004) Auditors generally resign from clients

with high-financial distress and in receipt of a modified (particularly going-concern)

opinion (Krishnan and Krishnan, 1997) Shu (2000) also finds that an auditor resignation

is positively related to client legal exposure The preceding arguments suggest that as a

result of risky clients deciding to change auditors, or auditors deciding to resign from

risky clients, a large proportion of clients that do switch will be financially unsound

A large body of research underscores the higher levels of audit quality that the

top-tier audit firms can provide to their clients DeAngelo (1981) argues that audit firms

with more clients have greater incentives to supply higher quality audits Teoh and

Wong (1993) find that clients of Big N audit firms generally have higher earnings

response coefficients to audited earnings announcements, while Francis et al (1999)

find that Big N audit firms were able to reduce the level of DA reported by their clients,

indicating a more effective assertion of independence The prior literature suggests

that firms with higher levels of DA are able to manage earnings which lead to lower

audit quality

Audit firm rotation and audit quality

423

Some audit firms differentiate themselves from their competitors by specialising in auditing clients in particular industries Craswell et al (1995) find that industry-specialists command a fee premium of around 16 per cent over non-industry specialists, indicating that clients are willing to pay more for the services of such an auditor Schauer (2002) finds that clients of industry specialists had lower bid-ask spreads, signifying lower levels of information asymmetry associated, and Krishnan (2003) observes that clients of industry-specialists reported lower levels of accruals From the prior research, it appears that audit quality as measured by the propensity

to issue a going-concern opinion (GCO) and the level of DA is impacted by a change in auditor Therefore, we propose the following hypothesis, stated in the null:

Sample selection and research design Sample selection

This study examines auditor switches occurring between 1995 and 2003 for Australian listed entities It was decided to study the period from 1995, as data were readily available for the year, and follows many of the Big N audit firm mergers The period 1995-1999 also offers a comparable time period with 2000-2003 The period from 2000 has been marked by a number of considerable changes and events The Ramsay Report, issued in 2001, highlighted a number of threats to auditor independence, including the provision of non-audit services along with recommendations for audit partner rotation, which is included in CLERP 9 In addition, international high-profile corporate collapses, including Enron, WorldCom, and HIH, brought auditor-client relationships under even greater examination

The initial sample consisted of 772 auditor switches for listed ASX entities between

1995 and 2003 Financial sector firms were excluded due to the inherent differences in their reporting nature After excluding clients where DA data were not available, those with insufficient data, those reporting in a foreign currency, and for clients where a matched entity could not be found, a total of 205 auditor switches resulted The total sample was used to measure changes in the level of DA with auditor tenure Both switching client firm-year observations and non-switching client firm-year observations were included, yielding a total of 1,750 firm-year observations

Each listed entity’s independent auditor was identified from Craswell’s Who Audits Australia database from 1994 to 2003 The age of clients was obtained from the ASX web site and Aspect Huntley’s DatAnalysis Financial statement balances were also obtained from Aspect Huntley’s DatAnalysis Takeover announcements were collected from Connect 4, while debt and equity issues data were collected from Thompson’s SDC Platinum New Issues Database Market capitalisation data were collected from CRIF Any missing data were hand collected from relevant sources Variables were winsorised at the fifth and 95th percentiles to remove the effect of significant outliers

Research design Audit quality has been defined in numerous ways, ranging from the relative degree to which the audit conforms to applicable auditing standards (Krishnan and Schauer, 2001; Tie, 1999; McConnell and Banks, 1998; Cook, 1987), the market-assessed probability that the financial statements contain material errors and that the auditor will discover and report them (DeAngelo, 1981), the accuracy of the information

MAJ

23,5

424

reported on by auditors (Titman and Trueman, 1986; Beatty, 1989; Krinsky and

Rotenberg, 1989), and a measure of the audit’s ability to reduce noise and bias and

improve fineness in accounting data (Wallace, 1980) Any metric used to measure audit

quality, however, are not perfect Proxies for audit quality can only be devised in most

cases using publicly available information, and not private information known to the

auditor The true audit quality is when the audit does not result in a Type I or II error –

a failing company being given an unqualified report or a non-failing company being

given a qualified report

In order to test our hypothesis that a change in audit firm has no effect on audit

quality, we estimate the following model:

ð1Þ

where AQ is audit quality, TENURE is the number of continuous years the incumbent

auditor has been with the client; FRISK is the client financial risk as measured by the

Zmijewski (1984) financial distress score[1]; SIZE is measured as the natural log of total

assets, LEV is the ratio of liabilities to assets, CLEV is the percentage change in LEV

during the year, RETURN is the percentage change in the book value of net assets over

the year, LDISTRESS is a dummy variable indicating if the client was financially

distressed[2] in the previous year, INVEST is investment securities measured by

current assets less-current debtors and inventory scaled by total assets, FEERATIO is

fee dependence as measured by non-audit fees divided by non-audit and audit fees paid

to the incumbent auditor, strong cash flows from operations (SCFO) is the net cash

flows from operations scaled by lagged total assets, PRIOR is a dummy variable

indicating if the client received a going-concern opinion in the prior year, BIG_N is a

dummy variable indicating if the firm was audited by a Big N auditor, LEADER is a

dummy variable indicating if the audit firm is a leader in the firm’s industry, and

INDUSTRY is a dummy variable indicating if a firm is in the mining sector

Two measures of audit quality are employed within this study First, audit quality

is measured as the propensity of the auditor to issue a GCO after controlling for other

factors that might affect this decision A finding that auditors have a lower (higher)

propensity to issue going-concern opinions with increased tenure would provide

convincing evidence in favour of (in opposition to) mandatory rotation, i.e if there is

lower propensity to issue GCO with increased tenure, then independence becomes

impaired As the dependent variable in this measurement of audit quality is a

dichotomous variable, being the propensity to issue a going-concern opinion, a logistic

regression is run

Second, the level of DA is used DA are accruals that do not relate to normal

operating activities, and so a higher level of these accruals may indicate that

management has been able to exert its power over the auditor by being able to report

on terms favourable to management As this second measurement of audit quality is a

continuous variable, an OLS regression is sufficient

To measure DA, a performance-matched modified-Jones (1991) DA model was used

Dechow et al (1995) find that the modified Jones model had the highest statistical

Audit firm rotation and audit quality

425

power in detecting earnings management, while Kothari et al (2005) suggest matching for performance helps to control for changes in accruals models associated with client performance levels Performance matched firms are matched by year and industry DA are estimated by:

where TACC is total accruals (operating net profit after tax – cash flows from operations), DSales is the change in sales for the year scaled by lagged total assets, DRec is the change in accounts receivable during the year scaled by lagged total assets, PPE is the gross property, plant and equipment scaled by lagged total assets, Growth

is the ratio of next year’s sales to this years, and the residual from the regression, 1, is the measure of DA The higher the level of the residual indicates a lower level of accrual quality

The model employed in this study is based closely on that used by DeFond et al (2002) and Carey and Simnett (2006) The Zmijewski (1984) probability of bankruptcy score (FRISK) is included because clients closer to bankruptcy should have a higher chance of receiving a going-concern report SIZE is included because larger clients will have more assets to sell in the event of financial difficulty, and should be less prone to receiving a going-concern report LEV and CLEV are employed because higher levels

of debt relative to total assets indicate higher risk, and greater changes in leverage may suggest that the client is close to violating a debt covenant RETURN is included, because clients with higher levels of growth should be less likely to fail

Results Descriptive statistics Descriptive statistics are presented in Table I Panel A provides the results for the full sample, with Panels B and C providing the results for the switching and non-switching firms, respectively Means of the variables are consistent with prior research Of the total sample, 4.06 per cent of firm-year observations result in a GCO being issued, which increases to 5.37 per cent for switching firms (3.88 per cent for non-switching firms) Table II outlines the correlation coefficients between the variables, using both Pearson and Spearman rank correlations Firms that were financially distressed in the prior year have a high-negative correlation with both RETURN and SCFO

Going-concern opinion Table III presents the results of the logistic regression using the propensity to issue a going-concern audit opinion as a proxy for audit quality Panel A provides the results for the full sample, with Panels B and C providing the results for the switching and non-switching firms, respectively

Results from Table III indicate that audit-client tenure increases the likelihood of the auditor issuing a going-concern audit opinion This result is inconsistent with the results of Geiger and Raghunandan (2002) who found no statistically significant relationship, and with those of Choi and Doogar (2005) who found a significantly negative relationship between audit quality and tenure

As expected, FRISK was positively associated with the propensity to issue a going-concern opinion, indicating that firms with higher probabilities of bankruptcy are more likely to receive a going-concern opinion SIZE was found to have a

MAJ

23,5

426

Table I Sample descriptive

statistics

Audit firm rotation and audit quality

427

Table I.

MAJ

23,5

428

0.0537 (0.0245)

0.0320 (0.1806)

0.0062 (0.7947)

0.0911 (0.0001)

0.0627 (0.0087)

0.0365 (0.1267)

0.0227 (0.3435)

0.0576 (0.0160)

0.0553 (0.0207)

0.0303 (0.2052)

0.0501 (0.0360)

0.0405 (0.0905)

0.0201 (0.4015)

0.0022 (0.9260)

0.0369 (0.1229) 0.0232 (0.3323)

0.0295 (0.2172)

0.0808 (0.0007)

0.0184 (0.4415)

0.0028 (0.9085)

0.0409 (0.0874)

0.0310 (0.1950)

0.0331 (0.1665)

0.0357 (0.1349)

0.0417 (0.0814) 0.0162 (0.4996)

0.0177 (0.4583)

0.0366 (0.1255) 0.0071 (0.7652)

0.0595 (0.0127) 0.0805 (0.0007) 0.0920 (0.0001) 0.0474 (0.0475)

0.0342 (0.1529)

0.0040 (0.8665)

0.0248 (0.2991)

0.0219 (0.3608)

0.0307 (0.1992)

0.0191 (0.4251) 0.0506 (0.0344)

0.0020 (0.9329)

0.0597 (0.0125) 0.0655 (0.0061)

0.0299 (0.2112)

0.0496 (0.0381)

0.0329 (0.1689)

0.0861 (0.0003)

0.0911 (0.0001) 0.0513 (0.0320)

0.0528 (0.0270) 0.0187 (0.4339)

0.0238 (0.3189) 0.0248 (0.2991)

Table II Pearson and Spearman

correlations

Audit firm rotation and audit quality

429