nasser et al - 2006 - auditor‐client relationship- the case of audit tenure and auditor switching in malaysia

Managerial Auditing Journal
Auditor-client relationship: the case of audit tenure and auditor switching in Malaysia
Abu Thahir Abdul Nasser Emelin Abdul Wahid Sharifah Nazatul Faiza Syed Mustapha Nazri Mohammad
Hudaib
Article information:
To cite this document:
Abu Thahir Abdul Nasser Emelin Abdul Wahid Sharifah Nazatul Faiza Syed Mustapha Nazri Mohammad
Hudaib, (2006),"Auditor-client relationship: the case of audit tenure and auditor switching in Malaysia",
Managerial Auditing Journal, Vol. 21 Iss 7 pp. 724 - 737
Permanent link to this document:
/>Downloaded on: 22 December 2014, At: 05:19 (PT)
References: this document contains references to 28 other documents.
To copy this document:
The fulltext of this document has been downloaded 3369 times since 2006*
Users who downloaded this article also downloaded:
Winifred D. Scott, Willie E. Gist, (2013),"Forced auditor change, industry specialization and audit fees",
Managerial Auditing Journal, Vol. 28 Iss 8 pp. 708-734 />Rani Hoitash, Ariel Markelevich, Charles A. Barragato, (2007),"Auditor fees and audit quality", Managerial
Auditing Journal, Vol. 22 Iss 8 pp. 761-786 />Andrew B. Jackson, Michael Moldrich, Peter Roebuck, (2008),"Mandatory audit firm
rotation and audit quality", Managerial Auditing Journal, Vol. 23 Iss 5 pp. 420-437 http://
dx.doi.org/10.1108/02686900810875271
Access to this document was granted through an Emerald subscription provided by 123705 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive

preservation.
*Related content and download information correct at time of download.
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
Auditor-client relationship: the
case of audit tenure and auditor
switching in Malaysia
Abu Thahir Abdul Nasser and Emelin Abdul Wahid
Faculty of Accountancy, University Technology Mara, Johor, Malaysia
Sharifah Nazatul Faiza Syed Mustapha Nazri
Faculty of Accountancy, University Technology Mara, Selangor, Malaysia, and
Mohammad Hudaib
Department of Accounting and Finance, Bradford University School of
Management, Bradford, UK
Abstract
Purpose – The main purpose of this paper is to examine one aspect of auditor-client relationship,
namely audit tenure and switching behaviour, and factors affecting it. Lengthy audit tenure with the
same client has been cited as one of the threats to auditor independence. Given the importance of this
issue, this research attempts to shed some light on the effect of audit tenure and switching behaviour
on auditor independence in Malaysia.
Design/methodology/approach – This study evaluates the effects of various independent
variables on switching behaviour and audit tenure using logistic regression analysis.
Findings – An examination of 297 companies listed on the Kuala Lumpur Stock Exchange over a
period of 11 years reveals audit firm switching to be significantly associated with distressed large
clients and that the length and direction of switch depend upon the type of audit firm.
Research limitations/implications – A number of important variables such as corporate
governance characteristics, audit and non-audit fees and types of audit opinion that could enhance our
understanding of audit tenure and auditor switching in Malaysia, were not incorporated into our
regression models. Hence, future studies may consider such variables.
Originality/value – This study is the first conducted using Malaysian data where audit tenure and
switching are used as dependent variables. The results have important implications on the auditing

profession and regulators in Malaysia.
Keywords Auditors, Auditing, Customer relations, Malaysia
Paper type Research paper
1. Introduction
Auditor independence is the cornerstone of the auditing profession. In general, auditor
independence can be of two forms: “independence in fact” and “independence in
appearance”. The former requires auditors to form and express an opinion in the audit
report as a disinterested and expert observer, uninfluenced by personal bias, while the
latter expects auditors to avoid situations that might cause others to conclude that they
are not maintaining an unbiased objective attitude of mind (Porter et al., 2003).
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0268-6902.htm
The authors acknowledge the research funding from IRDC of University Teknologi Mara,
Malaysia in conducting this research.
MAJ
21,7
724
Managerial Auditing Journal
Vol. 21 No. 7, 2006
pp. 724-737
q Emerald Group Publishing Limited
0268-6902
DOI 10.1108/02686900610680512
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
Flint (1988) argued that independence will be lost if the auditor is involved in a
personal relationship with the client, as this may influence their mental attitude and
opinion. One of such threats is lengthy tenure. He contends that lengthy tenure in office
may cause the auditors to develop “over-cosy relationships” as well as strong loyalty or
emotional relationships with their clients, which could reach a stage where auditor
independence is threatened. Lengthy tenure also results in “over familiarity” and

consequently, the quality and competence of auditors’ work may decline when they start
to make unjustified assumptions instead of objective evaluation of current evidence.
The GPES 1.201 (para 2.5) of the ICAEW (2001) recognised that this problem may
be perceived as a threat to auditor independence and recommends that auditors avoid
situations that may lead them to become over-influenced or to be too trusting of the
client’s directors and key personnel which could consequently lead to audit staff being
too sympathetic to the client interest. Since, the relationship between the auditor and
his client can be close in one way or another regardless of the number of years in office,
there is little the profession can do with regard to this matter (Dunn, 1996). In other
words, the profession does not object to auditors serving their clients for a long period
of time, but the objection seems to be over the worry that long period of service may
lead to “cosy relationships” that may threaten auditor independence.
Hence, to maintain public confidence in the audit function and to protect auditors’
objectivity, the profession through a series of clauses proscribes auditors from having
personal relationships with their clients that may give rise to a potential conflict of
interests. One of the proposals is to have mandatory auditor rotation (AICPA, 1978a, b)
as it may increase auditors’ ability in protecting the public via the increase in alertness
to any possible improprieties, increase in quality service and prevent closer
relationship with client (Mautz, 1974; Winters, 1976; Hoyle, 1978; Brody and Moscove,
1998). However, some are against the idea because they believe that the costs outweigh
the benefits. Frequent rotation and switching would result in increased audit fees as
the benefits to be gained from subsequent lower cost after the initial years of any audit
would not be fully realised. Another disadvantage is that the knowledge gained over
time in improving quality of audit work would be wasted with the appointment of a
new auditor (AICPA, 1992). Nevertheless, in 1994, the professional bodies in the UK
introduced a seven year rotation policy for the audit engagement partners auditing
public listed companies (Porter et al., 2003). A similar rotation policy was also
introduced for audit engagement partners involved in auditing SEC registered
companies (Wood and Sommer, 1985; Brody and Moscove, 1998).
In Malaysia, the issue of audit tenure and interval rotation of audit firms or audit

partners were not explicitly addressed in any of the relevant Malaysian official
documents such as the Companies Act 1965, the Security Commission regulations,
approved auditing standards, etc. Lack of official pronouncements on this issue could
be due to the rejection of such rotation idea by the business community. Jaffar and
Alias (2002) found only 35 per cent of the audit firms’ partners and only 32.4 per cent of
the chief finance officers surveyed favoured audit firm rotation every three years of
engagement. However, in light of the Enron case, the Chairman of the Malaysian
Accounting Standard Board announced the intention of the board to make it
mandatory to rotate the audit firm once every five years (The Edge, 2002). While some
countries are either considering or have already imposed the five-year restriction to
rotate the audit firms, the length of audit tenure and the possible effect of switching on
Auditor-client
relationship
725
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
auditor independence in Malaysia is still unclear. Hence, the results would highlight if
audit tenure and switching should be of concerned before any rotation length is
imposed. This is our main contribution to the literature.
Our analysis involves an examination of 297 Malaysian listed companies between
1990 and 2000 using logistic regression. Results indicate the probability of switching
the audit firm is greatest for distressed large client and that the direction of switch and
length of tenure are dependent on type of audit firm. This implies that auditors in such
environment fear losing their tenure and being switched, hence their independence and
objectivity may be impaired. As such, we assert that rotation policy should be partially
imposed on distressed clients as a first step forward in protecting auditor independence
in the country. The paper proceeds as follows. The next section discusses the literature
review and the development of the hypotheses. Sections 3 and 4 present the research
method and results, respectively. Section 5 presents the summary and conclusions.
2. Literature review and development of hypo theses
It has been suggested in the literature that larger audit firms (Big 4) are usually

perceived as more capable of maintaining an adequate degree of independence than
their smaller counterparts because they usually provide a range of services to a large
number of clients, hence reducing their dependence on certain clients (Dopuch, 1984;
Wilson and Grimlund, 1990). In addition, larger audit firms are generally perceived as
the provider of high audit quality and enjoy a high reputation in the business
environment and as such, would strive to maintain their independence to keep up their
image (DeAngelo, 1981; Dopuch, 1984; Wilson and Grimlund, 1990). Furthermore,
larger audit firms are also perceived to be more independent than their smaller
counterparts in withstanding management’s pressure in the event of disputes as they
normally have more clients and can afford to give up some of their more “difficult”
clients (Chow and Rice, 1982).
In Malaysia, high dependence on a few clients has been found to affect perception of
independence (Teoh and Lim, 1996). This is not surprising as the market for audit
services for public companies in Malaysia is dominated by the international Big 4
(previously the Big 6) audit firms. In fact, Che-Ahmad and Derashid (1996) report that
the Big 6 (and their affiliates) audited 75.9 per cent of the Bursa Malaysia (Main Board)
listed companies in 1991. Based on the above arguments, we expect the length of tenure
by Big 4 audit firms to be significantly longer as their clients would be less likely to
switch them compared to their smaller counterparts. Hence, the H1 is stated as follows:
H1. The probability of switching big audit firms (Big 4) is significantly lower than
the probability of switching smaller audit firms, ceteris paribus.
Besides the possible effect of the type of audit firms on the length of tenure, the choice
of audit firm can be related to the size of the auditee and the type of services needed.
It has been argued that larger auditees, due to the complexity of their operations and
the increase in the separation between management and ownership, demand highly
independent audit firm to reduce agency costs (Watts and Zimmerman, 1986) and
auditors’ self-interest threat (Hudaib and Cooke, 2005). Furthermore, as the size of the
companies increases, it is likely that the number of agency conflicts also increases and
this might increase the demand for quality-differentiated auditors (Palmrose, 1984),
i.e. Big 4 audit firms.

MAJ
21,7
726
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
Based on the above arguments, we expect the length of tenure of auditors of large
clients to be longer than that of auditors of smaller clients in Malaysia. In other words,
we expect the propensity to switch auditors to be lower for large clients than their
smaller counterparts. This leads us to the following hypothesis:
H2. The probability of large clients switching audit firms is significantly lower
than the probability of small clients switching audit firms, ceteris paribus.
When businesses continue to grow, the demand for the highly independent and
qualified audit firm to reduce agency costs and to provide non-audit services needed
for the expansion of the firm increases. Therefore, growing businesses are expected to
be more likely to retain their audit firms than their lower growth counterparts. Sinason
et al. (2001) examined 16,976 COMPUSTAT companies in the US over a period of
20 years and found that audit tenure is significantly affected by client’s growth rate.
Since, the literature indicates that audit tenure is affected by the client’s growth rate,
we hypothesised the length of tenure of auditors of high growth clients in Malaysia to
be longer than low growth clients. In other words, high growth clients are less likely
to switch their auditors. As such, our next hypothesis is:
H3. The probability of high growth clients switching audit firms is significantly
lower than the probability of low growth clients switching audit firms, ceteris
paribus.
The financial position of auditees may have important implications on decisions in
retaining the audit firm. Auditees who are insolvent (have high gearing ratios) and are
experiencing an unhealthy financial position are more likely to engage auditors having
high independence to boost the confidence of shareholders and creditors as well as to
reduce the risk of litigation (Francis and Wilson, 1988). In addition, financially stressed
clients are more likely to replace their audit firms compared to their healthier
counterparts (Schwartz and Menon, 1985; Hudaib and Cooke, 2005).

As such, we expect auditors of distressed clients to have shorter tenure compared to
their counterparts auditing healthier clients and will in turn be less likely to be
switched. The next hypothesis is stated as follows:
H4. The probability of distressed clients switching audit firms is significantly
higher than the probability of healthier clients switching audit firms, ceteris
paribus.
Sinason et al. (2001) found the length of audit tenure to be positively affected by the
type of audit firm. In other words, smaller audit firms experience shorter tenure
compared to their larger counterparts who often enjoy lengthy tenure. Differences in
the length of tenure between the two types of audit firms could impair independence
because in the long run, small audit firms will find it difficult to keep their existing
clients and at the same time maintain a high degree of independence and objectivity
due to increased competition and size mismatch. Ideally, the size of audit firm should
match the size of auditee. A size mismatch between large auditees audited by small
audit firms could cause termination of the audit engagement (Hudaib and Cooke, 2005),
i.e. a switch of auditor.
Since, there are four possibilities of switching auditors, viz. switch from non-Big 4 to
non-Big 4, from non-Big 4 to Big 4, from Big 4 to non-Big 4 and from Big 4 to Big 4,
Auditor-client
relationship
727
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
we expect the lengths of tenure for the four types of switches to be significantly
different. Specifically, we hypothesise that:
H5. The length of audit tenure for non-Big 4 replaced by Big 4 (
m
2
) is significantly
shorter than the length of audit tenure for non-Big 4 replaced by another
non-Big 4 (

m
1
). Hence,
m
1
–
m
2
–
m
3
–
m
4
.
where,
m
1
, switched from non-Big 4 to non-Big 4;
m
2
, switched from non-Big 4 to Big 4;
m
3
, switched from Big 4 to non-Big 4;
m
4
, switched from Big 4 to Big 4.
3. Research method
3.1 Data sources

There were 810 listed companies on both the main and second board of the Kuala
Lumpur Stock Exchange (KLSE) or Bursa Malaysia as at 10 June 2002. The sample of
this study consists of 297 randomly selected companies listed on both boards. Based on
the sample error formula (Burns and Bush, 2003), a sample size of 297 is deemed
appropriate. The relevant data had been collected from KLSE Research Institute
Database for a period of 11 years from 1990 to 2000. Besides the relevant financial
statements, the relevant audit reports of the sample firms available at KLSE were also
used in this research.
The dependent variables are auditor switching (SWITCH
t
) and the four directions of
switch (
m
xt
) which are noted from the audit reports during the study period. The
independent variables consist of book value of equity (BVE) and market value of
equity (MVE), client size (CLI.SIZE), changes in total assets (Ln DTA), changes in sales
(Ln DS), type of audit firm (AUDIT), changes in income from continuing operations in
the two years preceding the audit change (Ln DR
t2 2
), financial distress of the client
company (Ln DZ ),[1] interactive effects of length of tenure held by Big 4 and non-Big 4
before being switched (TENU
*
AUDIT
xt
) and the interactive effects of length of tenure
before the first switch (TENU
*
B1.SWI).

3.2 Data analysis
Logistic regression was adopted to assess the relationships since the dependent variables
are dichotomous. The model parameters are estimated using the maximum-likelihood
method whereby the coefficients that make the observed results most “likely” are selected
on the basis of an iterative algorithm. Furthermore, the maximum-likelihood method has
also the advantage of asymptotic normality (Hudaib and Cooke, 2005). Model 1 has the
dependent variable as switching/non-switching (SWITCH
t
) and Model 2, direction of
switching/otherwise (
m
xt
). Model 1 provides answers to the research hypotheses
associated with audit tenure and switching while Model 2 investigates the impact of
independent variables on the directions of auditor switching.
3.3 Model specification
3.3.1 Model 1. We use the following logistic regression model to test for the
relationships between auditor switching (SWITCH
t
) and type of audit firm (AUDIT),
client size (CLI.SIZE), client growth (Ln DS), client financial risk (Ln Z), the interactive
effects of length of tenure of remaining in the office before being switched
(TENU
*
AUDIT
xt
), the change in operating income (Ln DR
t2 2
), change in market value
of equity (Ln DMVE) as well as change in total assets (Ln DTA):

MAJ
21,7
728
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
SWITCH
t
¼
a
0
þ
b
1
AUDIT þ
b
2
CLI:SIZE þ
b
3
Ln DS þ
b
4
Ln Zþ
X
2
i¼1
d
ixt
TENU
*
AUDIT

xt
þ
b
7
Ln DR
t22
þ
b
8
Ln DMVE þ
b
9
Ln DTA þ1
SWITCH
t
is a binary variable indicating whether or not the audit firm was switched or
not switched. The other independent variables are as summarised in Table I.
3.3.2 Model 2. To further explain the impact of independent variables on the
switched directions of audit firms (
m
xt
), the following logistic regression model was
adopted to test the association between auditor switching from one type of audit firm
to another and the nine independent variables, viz. type of audit firm (AUDIT), client
size (CLI.SIZE), client growth (Ln DS), financial risk (Ln Z), change in operating income
(Ln DR
t2 2
), change in market value of equity (Ln DMVE), the change in total assets
(Ln DTA), the interactive effects of length of tenure before the first switching
(TENU

*
B1.SWI) and the interactive effects of length of tenure of a large audit firm
(Big 4) remaining in the office before being switched (TENU
*
AUDIT
B4
):
a
0
Intercept
SWITCH A dummy variable, 1 if the audit firm is switched at time
t
, and 0 otherwise
m
xt
Four propositions of switch directions
1.
m
1
¼ from non-Big 4 to non-Big 4
2.
m
2
¼ from non-Big 4 to Big 4
3.
m
3
¼ from Big 4 to non-Big 4
4.
m

4
¼ from Big 4 to Big 4
AUDIT A dummy variable, 1 if the firm is a Big4 audit firm, and 0 otherwise
CLI.SIZE A dummy variable, 1 if the client’s TA is large (ln TA . the mean), and
0 otherwise
Ln DS The natural logarithm of squared changes in sales scaled by ln TA
½lnððDS=ln TAÞ
2
Þ;
Ln Z The natural logarithm of company’s financial risk: cash flow from operations
over long-term debt [ln(Z)
2
]
SQ ln BVE The square root of natural logarithm of BVE ½
ffiffiffi
2
lnðBEÞ
2
p
;
ln TA The natural logarithm of the client’s assets [ln TA]
Ln DTA The natural logarithm of squared changes in of total assets ½lnðDTAÞ
2
;
ln MVE MVE [ln MVE]
Ln DMVE The natural logarithm of squared changes in MVE ½lnðDMVEÞ
2
;
LnDR
t2 2

The natural logarithm of squared changes in operating income two years
(year
t2 2
) before switching in year
t2 0
½lnðDR
t22
Þ
2
;
TOT.TENU Total length of tenure (years) for the period
TENU
*
B1.SWI The interactive effects of length of tenure before the first switch, company j is
coded 1 up to 11 depending of the length of tenure before the first switch in
period
tx
TENU
*
AUDIT
xt
Two propositions of the interactive effects of tenure length before switching
1. TENU.AUDIT
B4
¼ number of years the incumbent Big4 held office before
being replaced divided by TOT.TENU
2. TENU.AUDIT
nB4
¼ number of years the incumbent non-Big4 held office
before being replaced divided by TOT.TENU

1 Error term
Table I.
Variables in the logistic
regression models
Auditor-client
relationship
729
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
m
t
¼
a
0
þ
b
1
AUDIT þ
b
2
CLI:SIZE þ
b
3
LnDS þ
b
4
Ln Z þ
b
5
Ln DR
t22

þ
b
6
Ln DMVE
b
7
Ln DTA þ
b
8
TENU
*
B1:SWI þ
b
9
TENU
*
AUDIT
B4
þ 1
The model is run four times to cover all switch directions taken by clients to appoint
their new audit firms after terminating the services of their incumbent audit firms.
4. Empirical results
4.1 Descriptive statistics
Table II (Panel A) presents descriptive statistics for the experimental, control and
dependent variables. The mean size of clients is RM 1.5 billion. The negative signs in
Minimum Maximum Sum Mean SD
Panel A (Continuous)
TA 35,251 51,453,100,000 1546788049.45 4855264576.1
Ln DTA 2 15.2018 1.1077 2 5.106514 2.6
BE 2 1,882,020,548 14,518,600,000 449149349.93 1182242201.7

SQ ln BVE 5.22 6.84 6.1446 0.2
MVE 350 17,432,850 404402.64 1614674.7
Ln DMVE 2 9.6318 12.9503 2 0.753392 2.1
Ln DS 2 13.6249 9.5961 2 3.629986 3.3
Ln DR
t2 2
2 18.4207 10.1909 2 1.388829 3.2
Risk (Z) 2 0.0300 1835.8700 24.510732 172.1
Ln Z 2 20.8286 15.0305 2 0.525179 4.2
TENU
*
B1.SWI 1 11 6.10 3.5
TENU
*
AUDIT
B4
0.00 1.00 0.5576 0.46640
TENU
*
AUDIT
nB4
0.00 1.00 0.4424 0.46640
Panel B (Dichotomous)
CLI.SIZE 0 1 129 0.43 0.497
AUDIT 0 1 175 0.59 0.493
SWITCH 0 1 87 0.29 0.457
TOT.TENU 1 11 1,987 6.69 3.619
m
xt
0 4 198 0.67 1.191

m
1
0 1 23 0.08 0.268
m
2
0 1 36 0.12 0.327
m
3
0 1 9 0.03 0.172
m
4
0 1 19 0.06 0.245
Notes: n – 297; TA – total assets; ln TA – the natural logarithm of total assets; Ln DTA – the
natural logarithm of squared changes in total assets; SQ ln BVE – the square root of natural logarithm
of BVE; Ln DMVE – the natural logarithm of squared changes in MVE; LnDS – the natural logarithm
of squared changes in sales scaled by ln TA; Ln DR
t2 2
– the natural logarithm of squared changes in
operating income; Ln Z – the natural logarithm of company’s financial risk; TENU
*
B1.SWI – the
interactive effects of length of tenure before the first switching; TENU
*
AUDIT
B4
– the interactive
effect of years in office by Big4 before switching over total tenure period; TENU
*
AUDIT
nB4

– the
interactive effects of years in office by non-Big4 before switching over total tenure period; CLI.SIZE – a
dummy variable, 1 if the client’s TA is large, and 0 otherwise; AUDIT – a dummy variable, 1 if the
firm is a Big4 audit firm, and 0 otherwise; SWITCH – a dummy variable, 1 if the audit firm is
switched, and 0 otherwise; TOT.TENU – total length of tenure (years) for the period;
m
1
– switched
from non-Big4 to non-Big4;
m
2
– switched from non-Big4 to Big4;
m
3
– switched from Big4 to
non-Big4;
m
4
– switched from Big4 to Big4
Table II.
Descriptive statistics
MAJ
21,7
730
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
the mean values indicate poor financial performance during the period of analysis.
Table II (Panel B) reports the descriptive statistics of the dichotomous variables. It can
be seen from the table that the sample consisted of 43 per cent (129) large companies,
60 per cent (175) Big 4 and 29 per cent (87) cases of audit termination (switching). It also
shows that switching from small to large audit firm (i.e. from non-Big 4 to Big 4) is the

most common type of auditor change (36 cases), followed by switching from small
audit firm to small audit firm (23 cases) and switching from large audit firm to large
audit firm (19 cases). Not surprising, the least common type of auditor switch is from
large audit firm to small audit firm (9 cases).
Table III presents the correlation matrix for the dependent and continuous independent
variables. It indicates no multicollinearity problem, as the correlations are relatively
low except the correlation between two independent variables: type of audit firm (AUDIT)
and (TENU
*
AUDIT
xt
– the tenure length of Big 4 and non-Big 4). Therefore, AUDIT and
TENU
*
AUDIT
nB4
are dropped when we run our regression models.
4.2 Results of the logistic regressions
4.2.1 Model 1. Model 1 tests which independent variables (viz. client size, client growth,
client financial risk, the length tenure in the office before being switched, the changes
in operating income, market value of equity as well as the changes in total assets)
explain the behaviour of ending audit firm tenure in office (i.e. audit switching). The
results are shown in Table IV. Client size, client financial risk, the changes in total
assets and the interactive effects of the length of tenure in office before switching large
audit firm (TENU
*
AUDIT
B4
), were found to be significantly associated with audit
switching while client growth, changes in operating income and market value of equity

were found not to be significant. The results indicate that the main factors for
switching audit firm are the increase in total assets and the financial risk of the
company. The highest log odds of CLI.SIZE [Exp(B) ¼ 1.76] indicates that client size is
the most important factor followed by financial risk in explaining switching and as
such H2 and H4 are accepted. However, the growth of company’s business, the
increase in operating income and increase in market value do not affect the length of
tenure of the audit firm, thus allowing us to reject H3. The fact that the length of tenure
of large audit firms (Big 4) is negatively associated with switching suggests that larger
audit firms experience fewer instances of being replaced compared to their smaller
counterparts who face shorter tenure as they tend to be replaced more often and as
such, H1 is accepted.
4.2.2 Model 2. Table V shows the results of testing Model 2, i.e. the impact of
various independent variables on the four switching directions (
m
xt
). As indicated by
the results, the only factor that significantly impacts switching behaviour is the
interactive effects of length of tenure before the first switch (TENU
*
B1.SWI), followed
by changes in total assets of the client (Ln DTA).
Table VI (Panel A) provides details on the pattern of switching behaviour. As can be
seen, switching for similar sized audit firms, i.e. switching from a small audit firm to
another or from a large audit firm to another, often occurs after lengthy tenure.
However, switching to dissimilar sized audit firms, i.e. switching from a small to a
large audit firm or from a large to a small audit firm, occurs in a relatively shorter
period. This suggests that the propensity to switch to a large audit firm is greater for
clients experiencing an increase in total assets.
Auditor-client
relationship

731
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
Variable
m
1
m
2
m
3
m
4
SQ ln BVE
SWITCH 0.45
**
0.58
**
0.27
**
0.41
**
0.14
*
m
1
1 2 0.11 2 0.05 2 0.08 0.05
m
2
1 2 0.07 2 0.10 0.02
m
3

1 2 0.05 0.05
m
4
1 0.14
*
SQ ln BVE 1
Ln DTA
Ln DMVE
Ln DS
Ln DR
t2 2
Ln Z
AUDIT
TOT.TENU
TENU
*
B1.SWI
TENU
*
AUDIT
B4
Variable Ln
D
TA Ln
D
MVE Ln
D
SLn
D
R

t2 2
Ln Z
SWITCH 0.06 2 0.03 0.08 0.10 0.08
m
1
2 0.01 0.05 0.00 0.02 0.10
m
2
0.04 2 0.09 0.12
*
2 0.00 0.05
m
3
0.09 2 0.03 0.04 0.15
**
2 0.02
m
4
0.02 0.02 2 0.04 0.06 20.03
SQ ln BVE 2 0.07 2 0.07 2 0.07 2 0.04 2 0.07
Ln DTA 1 2 0.02 0.21
**
0.03 0.02
Ln DMVE 1 2 0.04 2 0.07 20.03
Ln DS 1 0.17
**
0.12
*
Ln DR
t2 2

1 2 0.07
Ln Z 1
AUDIT
TOT.TENU
TENU
*
B1.SWI
TENU
*
AUDIT
B4
Variable AUDIT TOT.TENU TENU
*
B1.SWI TENU
*
AUDIT
B4
TENU
*
AUDIT
nB4
SWITCH 0.02 0.31
**
0.05 2 0.09 0.09
m
1
2 0.35
**
0.10 0.16
**

2 0.35
**
0.35
**
m
2
0.27
**
0.19
**
2 0.15
**
2 0.03 0.03
m
3
2 0.21
**
0.13
*
2 0.05 0.00 2 0.00
m
4
0.22
**
0.11 0.16
**
0.25
**
2 0.25
**

SQ ln BVE 0.05 0.33
**
0.29
**
0.08 2 0.08
Ln DTA 0.03 0.06 0.02 0.03 2 0.03
Ln DMVE 2 0.06 2 0.07 2 0.05 20.06 0.06
Ln DS 0.11 0.05 0.02 0.10 2 0.10
Ln DR
t2 2
0.01 0.08 0.05 0.02 2 0.02
Ln Z 2 0.06 2 0.06 2 0.06 2 0.09 0.09
AUDIT 1 0.08 0.05 0.90
**
2 0.90
**
TOT.TENU 1 0.89
**
0.09 2 0.09
TENU
*
B1.SWI 1 0.14
*
2 0.14
*
TENU
*
AUDIT
B4
1 2 1.0

**
Notes:
**
Correlation is significant at the 0.01 level two-tailed;
*
correlation is significant at the
0.05 level two-tailed
Table III.
Correlation matrix
MAJ
21,7
732
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
We further extended our analysis by using ordinary least squares regression where
TENU
*
B1.SWI is the independent variable to test the H5. The results are reported in
Table VI (Panel B). The results show that the length of tenure is longer for similar sized
switching of audit firms and shorter for dissimilar sized switching of audit firms.
Hence, based on the results, H5 is accepted.
5. Summary and conclusions
This paper examines audit tenure and switching behaviour in the Malaysian audit
environment for the period from 1990 to 2000. It provides evidence on the relationships
between switching and five related variables, viz. the interactive effects of length
tenure of Big 4 before being replaced (TENU
*
AUDIT
B4
), client size (CLI.SIZE), client
growth (Ln DS), client financial risk (Ln Z) and four switching directions (

m
xt
). The
results show that retention of audit firms depends on the size of clients based on total
assets, level of financial risk and type of audit firm but not by changes in operating
income and market value. The likelihood of non-distressed large client audited by large
audit firm to switch is significantly less compared with distressed small client audited
by small audit firm. Results also show that switching directions are mainly influenced
by the interaction between directions of switch and type of audit firm with the length of
tenure. Length of tenure before switching from a small to a large audit firm is
significantly shorter compared to the length of tenure before switching from a small to
another small audit firm. Large audit firms were found to secure longer tenure. Hence,
such differences in the lengths of tenure suggest independence impairment.
Since, financially distressed clients are more likely to switch audit firms, the smaller
auditors would be more reluctant to qualify their reports or show disagreement with
their clients for fear of being dismissed and losing a client. The implication of this
finding is that such audit-client relationship may impair auditor independence and
weaken audit quality. This in turn has important implications for policy makers in
Malaysia since auditor independence may be impaired due to the unhealthy
competition among the audit firms.
This study is subject to several limitations. First, corporate governance
characteristics that would shed more light on this topic were not included in our
analysis. Second, we have not considered the effects of audit and non-audit fees on
audit tenure and the decision to retain auditor. Third, we did not consider types of
Variables B Exp(B)
Dependent variable: switch/not switch
Independent variables
CLI.SIZE 0.566
***
1.761

Ln DS 0.061 1.063
Ln Z 0.069
**
1.071
TENU
*
AUDIT
B4
2 0.631
***
0.532
Ln DR
t2 2
0.042 1.043
Ln DMVE 2 0.023 0.977
Ln DTA 0.084
**
1.087
Notes:
x
2
51.934 ( p ¼ 0.000); number of selected cases: 297; number of switching: 87; Nagelkerke
R
2
0.241;
***
1 per cent significance,
**
5 per cent significance,
*

10 per cent significance
Table IV.
Multivariate logistic
analysis: propensity to
switch (Model 1)
Auditor-client
relationship
733
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
Independent variables B Exp(B)
(A) Dependent variable:
m
1
(n Big 2 n Big) [n ¼ 23]
CLI.SIZE 0.227 1.255
Ln DS 0.096 1.101
Ln Z 0.091
*
1.095
Ln DR
t2 2
0.093 1.098
Ln DMVE 0.255
*
1.290
Ln DTA 0.333
***
1.394
TENU
*

B1.SWI 2 0.118
**
0.888
TENU
*
AUDIT
B4
1.144
***
3.138
TENU
*
AUDIT
nB4
––
x
2
203.75 ( p ¼ 0.000)
Number of selected cases: 297
Number of cases: 23
Nagelkerke R
2
0.723
(B) Dependent variable:
m
2
(n Big 2 Big) [n ¼ 36]
CLI.SIZE 0.390 1.476
Ln DS 0.159
**

1.173
Ln Z 0.029 1.030
Ln DR
t2 2
2 0.047 0.954
Ln DMVE 2 0.186
*
0.830
Ln DTA 0.049 1.051
TENU
*
B1.SWI 2 0.274
***
0.760
TENU
*
AUDIT
B4
––
TENU
*
AUDIT
nB4
2 0.115 0.892
x
2
180.07 ( p ¼ 0.000)
Number of selected cases: 297
Number of switching from nB5 2 B2: 36
Nagelkerke R

2
0.665
(C) Dependent variable:
m
3
(Big 2 n Big) [n ¼ 9]
CLI.SIZE 1.000 2.717
Ln DS 0.049 1.050
Ln Z 0.006 1.006
Ln DR
t2 2
0.200 1.221
Ln DMVE 2 0.138 0.871
Ln DTA 0.492
***
1.635
TENU
*
B1.SWI 2 0.338
**
0.713
TENU
*
AUDIT
B4
2 0.063 0.939
TENU
*
AUDIT
nB4

––
x
2
300.78 ( p ¼ 0.000)
Number of selected cases: 297
Number of cases: 9
Nagelkerke R
2
0.912
(D) Dependent variable:
m
4
(Big 2 Big) [n ¼ 19]
CLI.SIZE 0.478 1.613
Ln DS 0.000 1.000
Ln Z 0.004 1.004
Ln DR
t2 2
0.137
*
1.147
Ln DMVE 0.147 1.159
Ln DTA 0.384
***
1.468
TENU
*
B1.SWI 2 0.175
**
0.839

(continued)
Table V.
Multivariate logistic
analysis: propensity to
switch (Model 2)
MAJ
21,7
734
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
audit opinion and how receiving qualified opinion might affect the decision to switch or
retain the audit firm. Hence, future studies should consider such variables in the
models to enhance our understanding of the Malaysian audit environment.
Note
1. Beaver (1968) used the ratio of cash flow from operations over long-term debt to diagnose
financial risk of the company.
References
AICPA (1978a), The Commission on Auditors’ Responsibilities: Report, Conclusions and
Recommendations, American Institute of Certified Public Accountants, New York, NY.
Independent variables B Exp(B)
TENU
*
AUDIT
B4
0.731 2.077
TENU
*
AUDIT
nB4
––
x

2
217.60 ( p ¼ 0.000)
Number of selected cases: 297
Number of cases: 19
Nagelkerke R
2
0.754
Notes:
***
1 per cent significance,
**
5 per cent significance,
*
10 per cent significance
Table V.
Panel A Panel B
Switched from
Regression results
Dependent variable:
TENU
*
B1.SWI
TENU
*
B1.SWI
m
1
m
2
m

3
m
4
Estimate
10810
m
1
2.260
***
20400
m
2
2 1.397
**
33422
m
3
2 1.783
43211
m
4
0.935
5 1 2 1 2 CLI.SIZE 2.486
***
61412LnZ 0.013
7 1 4 0 1 TENU
*
AUDIT
B4
5.448

***
8 2 4 3 0 TENU
*
AUDIT
nB4
4.431
***
91202
10 1 2 0 0
11 10 0 0 9
Total 23 36 9 19
x
2
¼ 10.9
x
2
¼ 46.5
***
x
2
¼ 19.4
**
x
2
¼ 14.8 Adj. R
2
0.796
Notes: TENU
*
B1.SWI: interactive effects of length of tenure before the first switch;

m
1
: switched from
non-Big4 to non-Big4;
m
2
: switched from non-Big4 to Big4;
m
3
: switched from Big4 to non-Big4;
m
4
: switched from Big4 to Big4; TENU
*
AUDIT
B4
: interactive effects of years in office by Big4 before
being switched; TENU
*
AUDIT
nB4
: interactive effects of years in office by non-Big4 before being
switched;
***
1 per cent significance,
**
5 per cent significance,
*
10 per cent significance
Table VI.

Crosstabs between length
of tenure with the four
directions and regression
results
Auditor-client
relationship
735
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
AICPA (1978b), Report of Progress, American Institute of Certified Public Accountants,
New York, NY.
AICPA (1992), Statement of Position Regarding Mandatory Rotation of Audit Firms of Publicly
held Companies, American Institute of Certified Public Accountant: SEC Practice Section,
New York, NY.
Beaver, W.H. (1968), “Alternative accounting measures as predictors of failure”, The Accounting
Review, January, pp. 113-22.
Brody, R.G. and Moscove, S.A. (1998), “Mandatory auditor rotation”, The National Public
Accountant, pp. 32-5.
Burns, A.C. and Bush, R.F. (2003), Marketing Research, Prentice-Hall, Upper Saddle River, NJ.
Che-Ahmad, A. and Derashid, C. (1996), “The pricing of audit services: evidence from the KLSE
listed companies”, Journal Analysis, Vol. 4 No. 1, pp. 33-45.
Chow, C.W. and Rice, S.J. (1982), “Notes: qualified audit opinions and auditor switching”,
The Accounting Review, Vol. LVII, pp. 326-35.
DeAngelo, L. (1981), “Auditor independence, ‘low-balling’ and disclosure regulation”, Journal of
Accounting and Economics, August, pp. 113-27.
Dopuch, N. (1984), “The demand for quality-differentiated audit services in an agency-cost
setting: an empirical investigation: discussion”, in Abdel-Khalik, A. and Solomon, I. (Eds)
paper presented at Auditing Research Symposium, University of Illinois, Urbana, IL,
pp. 253-63.
Dunn, J. (1996), Auditing: Theory & Practice, Prentice-Hall, Upper Saddle River, NJ.
Flint, D. (1988), Philosophy and Principles of Auditing – An Introduction, Macmillan Education

Ltd, London.
Francis, J. and Wilson, E.R. (1988), “Auditor changes: a joint test of theories relating to agency
costs and auditor differentiation”, The Accounting Review, Vol. 63, pp. 663-82.
Hoyle, J. (1978), “Mandatory auditor rotation: the argument and alternative”, Journal of
Accountancy, pp. 69-78.
Hudaib, M. and Cooke, T.E. (2005), “The impact of managing director changes and financial
distress on audit qualification and auditor switching”, Journal of Business Finance &
Accounting, Vol. 32 Nos 9/10, pp. 1703-39.
ICAEW (2001), Members Handbook 2001, Institute of Charted Accountants in England and
Wales, London.
Jaffar, N. and Alias, N. (2002), “Audit firm rotation in Malaysia: prospects and problems”,
Finance India, Vol. XVI No. 3, pp. 933-48.
Mautz, R.K. (1974), “Rotation of auditors”, Financial Executive, July, pp. 48-56.
Palmrose, Z. (1984), “The demand for quality-differentiated audit services in an agency-cost
setting: an empirical investigation”, in Abdel-Khalik, A. and Solomon, I. (Eds) paper
presented at Auditing Research Symposium 1984, University of Illinois, Urbana, IL.
Porter, B., Simon, J. and Hatherly, D. (2003), Principles of External Auditing, Wiley, Chichester.
Schwartz, K.B. and Menon, K. (1985), “Auditor switches by failing firms”, The Accounting
Review, Vol. LX, pp. 248-61.
Sinason, D.H., Jones, J.P. and Shelton, S.W. (2001), “An investigation of auditor and client tenure”,
Mid-American Journal of Business, Vol. 16 No. 2, pp. 31-40.
Teoh, H. and Lim, C. (1996), “An empirical study of the effects of audit committees, disclosure of
nonaudit fees, and other issues on audit independence: Malaysian evidence”, Journal of
International Accounting, Auditing & Taxation, Vol. 5 No. 2, pp. 231-48.
MAJ
21,7
736
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)
The Edge (2002), “Column 4th”, The Edge, 5 November, pp. 2-3.
Watts, R.L. and Zimmerman, J.L. (1986), Positive Accounting Theory, Prentice-Hall, Inc.,

Englewood Cliffs, NJ.
Wilson, T. and Grimlund, R. (1990), “An examination of the importance of an auditor’s
reputation”, Auditing: A Journal of Practice & Theory, Vol. 9 No. 2, pp. 43-59.
Winters, A.J. (1976), “Looking at the auditor rotation issue”, Management Accounting, March,
pp. 29-30.
Wood, A.M. and Sommer, A.A. (1985), “Statement in quotes”, Journal of Accountancy, Vol. 156
No. 5, pp. 122-31.
Corresponding author
Mohammad Hudaib can be contacted at:
Auditor-client
relationship
737
To purchase reprints of this article please e-mail:
Or visit our web site for further details: www.emeraldinsight.com/reprints
Downloaded by Central Michigan University At 05:19 22 December 2014 (PT)