Abstract
Prior studies suggest that auditor's industry specialization are associated with highly earnings quality because the increase of client-specific knowledge. However, the effect of audit tenure on earnings quality is mixed. In this study, we examine whether auditor's industry specialization affect the association between auditor tenure and earnings quality. Our empirical findings suggest that auditor's industry specialization improve earnings quality than non-specialists audit firms. Additionally, results support the assumption that medium auditor tenure (auditors in their fourth fifth and sixth year with a client) is the best period of audit mission. By comparing the relative effects of audit industry specialization on the association between audit tenure and earnings quality, we suggest that audit-firm specialization plays a key role in determining the quality of earnings.
Keywords: Auditor tenure, Auditor industry specialization, earnings quality, value relevance, Tunisian listed companies
1. Introduction
Modern companies are characterized by separations of the functions of ownership and control. In this case, the theory of agency develops and supposes that separation of ownership and control creates certain problems bound essentially to the manager opportunism. This problem shows itself by the manager maximize their own personal wealth and not act in the shareholders' interest. It is for it there that shareholders use financial statement to monitor management's performance. However, management can manipulate financial statement. Auditing is one of the tools of control available to shareholders to ensure the sincerity and the regularity of the financial and accounting information contained in the financial statement. As underlines it DeAngelo (1981), external auditors, presumptively independent and competent, parties charged with verifying the fairness of the information provided to shareholders by managers. This argument is based on the assumption that the likelihood that all fraud and misstatements will be detected depends on the audit quality - the higher audit quality, the greater the chances of detecting fraud and misstatements.
The quality of an audit depends on one hand, the auditor is an industry specialist or not, and the other hand, the audit tenure. Prior studies in the accounting and auditing literature suggest that auditor's industry specialist is associated with higher earnings quality (Carcello and Nagy, 2002; Balsam and al., 2003; Krishnan, 2003; Velury, 2003; Jenkins and al., 2006; Gul and al., (2009). However, empirical studies which analyzed the association between audit tenure and earnings quality report mixed results (Shockley, 1981; Moscove, 1998; Imhoff, 2003; and Dopuch and al., 2001; Chung and Kallapur, 2003). Myers and al. (2003) define auditor tenure as the number of years an auditor is retained by the firm. Two different explanations can be provided for this relationship (Kaplan and Mauldin, 2008). The first is based on the argument that short-tenured auditors guaranteed the independence of the auditors and enhance the quality of the audit and financial reporting. The second is based on the argument that short-tenured auditors are associated with a lower knowledge industry-specific and reduce the quality of the audit and financial reporting.
In this study, we analyze the effect of audit quality on financial reporting quality. We first examine the relation between auditors industry specialization, auditor tenure and earnings quality. Second, we examine whether the auditors industry specialization can be affect the relation between auditor tenure and earnings quality. This study extends the stream of research that examines an additional dimension of earnings quality- value relevance.
The remainder of the study is organized as follows. We present first of all the institutional development of the Tunisian context. We describe previous studies on audit quality and earning quality in the third section. We detail our research design in the fourth section. We discuss our results in the fifth section, and we conclude in the final section.
2. Institutional Development
Tunisia belongs to the group of code-law regime countries; specifically Tunisia has a legal system of French origin. Code-law regime countries provide weaker legal protection of investors than common-law countries. However, in 1997 and with the publication of the new system accountants of companies, the accounting it Tunisia was modified. Indeed the new accounting law finds its origins of a combination between the French law and the English law. This new law provides highly legal protection of investors.
The auditing and accounting profession, within the Tunisia business setting have been developed into a crucial issue during this years, especially after the Batam stock market crash in 2002. The crash supported the notion that the audit, external governance mechanism, was rather inefficient to detect and report the fraudulent financial statements.
The first effort on develop the audit profession in Tunisia was made in 2002 with the introduction of the International Audit Standard (ISA) published by the International Federation of Accountants (IFAC). The second effort was made in 2005 with the introducing a specific legislative (the new law relating to the financial security) in the aftermath of accounting scandals involving fraudulent financial statement and illegal acts in Batam. This new law is framework in order to secure the proper functioning of the Tunisian stock market, through firms' efficient management. That law contains detailed instructions about the audit function and the form of a firm's corporate governance and essentially, the firm's audit committees organization and the participation of shareholders and her representatives in the decision making process.
External Auditors provide two roles to the capital market participants: an information role on the financial situation of the company and an insurance role on the quality of financial statement published by company. Prior studies suggest that auditor's industry specializations are associated with highly audit quality and earnings quality. However the effect of audit tenure on earnings quality is mixed. In this study, we examine whether auditor's industry specialization affects the association between auditor tenure and earnings quality.
3. Literature review and hypothesis development
Due to the increase in size and complexity of business companies, external auditors are seen as an external control mechanism to provide greater assurance on the information quality provided by organizations to aid decision making (Armstrong, 1987; Chandler and Edwards, 1996), to reduce the possibility of innocent mistakes and deliberate misstatements such as fraud and earnings management (Watts and Zimmerman, 1986; Chandler, Edwards and Anderson, 1993; Joshi and al., 2008).
The empirical association between audit quality and earnings quality it not new and has been studied extensively in the accounting and auditing literature (Becker and al., 1998; Johnson and al., 2002; Krishnan, 2003; Balsam and al., 2003; Myers and al., 2003; Ghosh and Moon, 2005).
This relation is based on the argument that high audit quality, as a result of more effective control, are more likely to detect management manipulation practices and misrepresentations by management than low audit quality. We examine the effect of auditor specialization and audit tenure on financial reporting quality by focusing on the Tunisia local government audit market, which in recent years has become characterized by the appearance of auditor specialization. Furthermore, we examine whether industry specialization of auditors affect the association between auditor tenure and earnings quality.
3.1. Auditor industry specialization effect
Auditor industry specialists can be affecting the audit quality. Indeed, industry-specialist audit firms possess more experience and more knowledge with industry-specific than non-specialist audit firms. Such experience and industry-specific knowledge arguably helps industry-specialist auditors identify many problems in their audit mission. Audit firms are also likely to make investments in staff training, physical facilities, organizational control systems and technologies in the industries in which they have extensive experience (Gramling and al., 1999). Similarly, Beasley and Petroni (2001) argue that superior industry knowledge and better audit technologies help industry specialists perform audits mission. Price waterhouse Coopers (2002) note that the quality of an audit depends on the ''knowledge of the company being audited and the industry-specific in which it operates.'' This assumption is based on the argument that auditor with industry expertise detect more irregularities than auditors without industry expertise.
Consistent with the notion that industry-specialist auditor's increase audit quality relative to non-specialist auditors, empirical research has demonstrated that industry-specialist audit firms perform superior quality financial reporting (Balsam and al., 2003; Krishnan, 2003 Dunn and Mayhew, 2004; Gul and al., 2009). Balsam and al. (2003) note that clients of industry specialist audit firms have larger earnings response coefficients as a proxy for earnings quality. Dunn and Mayhew (2004) provide evidence of a positive link between audit firm industry specialization and client disclosure quality proxy by analysts' evaluations in AIMR reports. Furthermore, Carcello and Nagy (2004a) found a negative association between audit firm industry specialization and client fraudulent financial reporting identified in SEC AAERs. Our hypothesis is stated as follows:
H1: Ceteris paribus, there is a positive relation between earnings quality and audit firm specialization.
3.1. Audit tenure effect
Previous audit tenure studies have predicted that audit tenure affect the financial reporting quality (Shockley, 1981; Moscove, 1998; Imhoff, 2003; and Dopuch and al., 2001; Chung and Kallapur, 2003; Gates et al., 2007; Kaplan and Mauldin, 2008). Studies in this area note early tenure audit reduce earnings quality. These results due to a lack of client-specific knowledge and/or a lack of independence due to the external auditor's incentive to maintain new client relationships (Shockley, 1981; Meyer and al., 2007; Fairchild, 2008; Davis and al., 2009). However, other studies provide evidence of a positive link between audit tenure and earning quality (Libby and Frederick, 1990; Myers et al., 2003; Carcello and Nagy, 2004; Iyer and Rama, 2004; Mansi and al., 2004; Ghosh and Moon, 2005; Jackson and al., 2008). These findings are consistent with the theory that length of auditor-client relationship to achieve a specific knowledge of the characteristics of companies, product differentiation and provide higher quality audits.
For example, Myers and al. (2003), Ghosh and Moon (2005) provide evidence that companies with longer auditor tenure are associated with higher earnings response coefficients. This interpretation is consistent with investors perceived earnings quality is better for firms with longer auditor tenure than the earnings quality of firms with shorter auditor tenure. Similarly, Chung and Kallapur (2003) found that the audit tenure was inversely related to abnormal accruals. Carcello and Nagy (2004b) evaluated the link between audit tenure client financial fraud disclosed in SEC Accounting and Auditing Enforcement Releases (AAERs). They provide evidence that the likelihood of financial fraud is greater in the initial three years of audit tenure. Alternatively, they do not find that length of auditor-client relationship is associated with increased likelihood of fraud.
Brandon and Mueller (2008) propose an empirical model which verifies the hypothesis of the impact of the length of auditor-client relationship on the audit quality perception. The contribution of these two authors is particularly interesting. Indeed, the results show that audit tenure can be analyzed according to two shutters. On one hand, the auditor wins of competence with the length of auditor-client relationship while accumulating a better experience of the audited company. On the other hand, the independence of the auditor decreases with the length of auditor-client relationship (reduce the probability to reveal irregularity and fraud contained in the financial reporting).
Deix and Giroux (1992) note that the length of auditor-client relationship reduces audit quality. Also, Meyers and al. (2007) and Moore and al. (2004) and Reynolds and al. (2004) found that audit tenure create certain attachment between both parties and modify the audit opinion. On the basis of the preceding discussion, we state the following hypothesis:
H2: Ceteris paribus, there is a negative relation between earnings quality and the length of auditor-client relationship.
3.3. Auditor industry specialization affects the association between auditor tenure and earnings quality
Previous audit tenure studies (Geiger and Raghunandan, 2002; Johnson et al., 2002; Myers et al., 2003; Mansi et al., 2004; Carcello and Nagy, 2004b; Ghosh and Moon, 2005) have predicted that early tenure audit reduce earning quality. Extend the literature, other studies provide evidence that audit quality (auditors' industry specialization or auditor Big) has an impact on the association between auditor tenure and earnings quality. For example, Stanley and DeZoort (2007) found a negative relation between the length of the auditor-client relationship and the likelihood of restatement. For short tenure engagements, they note that auditor industry specialization is negatively related to the likelihood of restatement. Similarly, Gul and al. (2009) integrate these two streams of research on auditor tenure and auditor industry specialization and examine whether auditors' industry specialization has an effect on the association between auditor tenure and earnings quality. They found that the association between shorter auditor tenure and lower earnings quality is weaker for firms audited by industry specialists compared to non-specialists.
We extend the literature involving the length of the auditor-client relationship to examine whether auditors' industry specialization has an effect on the association between auditor tenure and earnings quality. In this study, we supposed that the length of the auditor-client relationship reduces earnings quality. Then, we put the hypothesis that auditors' industry specialization improves earnings quality. If auditors' industry specialization improves earnings quality thus we wait in the fact that short audit tenure performs earnings quality. Also, we suppose that the length of the auditor industry specialization -client relationship improves the earnings quality.
Our third hypothesis predicts that the association between auditor-client relationship and earnings quality will differ for specialist and non-specialist auditors. We state the following hypothesis:
H3: Ceteris paribus, the association between tenure and earnings quality differs for specialist and non-specialist auditors
4. Research Design
4.1. Sample selection
The data for this paper come from the non financial companies stated in Tunisian Stock Market (BVMT). The data cover the period 2000-2005. All variables needed to compute stock return measures, auditors' quality were collected for a sample of 20 listed non-financial companies. The data on the auditors' industry specialization is collected by the means of a research questionnaire send with the auditors. The other variables correspond to data collected with the financial statement. The firms placed on the stock market during their first year are not included in the sample. This choice is explained by the argument that the stock market prices of the companies of their first year of initial public offering are not yet stable.
4.2. Model specification
To test the hypotheses relative to the relation between audit quality and earnings quality, we opt for techniques econometric of estimation on panel data. We use returns-earnings regressions to measure earnings quality. We use the following regression:
We measure earnings quality (EQit) by return-earnings relation. Our measure of earnings quality is based on regressions of market-adjusted returns on annual earnings levels and changes. The return-earnings relation is measured in the period following the announcement of annual reports. We prefer a window which includes the period of the date financial statement publication and the fiscal years end because we verify how the market assesses the quality of the audited annual report.
The earnings quality is estimated by earnings value relevance that is by the association between the stock-exchange returns and the annual earnings. The earnings value relevance is equal in adjusted R-square with the following regression:
Rit: are market-adjusted returns computed from the period of the date financial statement publication and the fiscal years end
EA and ∆EA represent levels of and changes in annual earnings, both deflated by the market value of equity at the end of fiscal year t -1.
SPEC is a dummy variable coded 1 if the firm is audited by a specialist and 0 if not. We portfolio share approach and define auditor industry specialization if the auditor maintains at least a 10 percent market share for the industry. Neal and Riley (2004) propose a new proxy for auditor industry specialization. This proxy is based on the total sales audited by an audit firm within an industry. However, we can not use this proxy in the Tunisia context because the audit fees are set by law. So the audit fees are not published by the companies and we can not have information on the audit fees.
Audit tenure is represented according to three variables. TENURE is the audit tenure, measured by the number of consecutive years when the auditor proceeded to the audit of the company
SHORT is equal to 1 if TENURE ≤ 3, and 0 otherwise. MEDUIM is equal to 1 if 4 ≤ TENURE ≤ 6, and 0 otherwise. LONG is equal to 1 if TENURE ≥ 7, and 0 otherwise
SIZE is the natural log of the market value of equity in the period of the date financial statement publication and the fiscal years end.
DEBT is the natural log of the ratio of short and long term debt to total equity for fiscal year t.
The indication i represents the company whereas the indication t indicates the study period considered (on 2000 in 2005). Thus the indication t represents the year t and the indication t-1 represents the year t-1. β is the coefficients relative to the variables of the study. Finally, εit: a standard residual term is.
In order to test our third hypothesis (3a) and to control whether auditors' industry specialization has an effect on the association between auditor tenure and earning quality, we divide the total sample into auditor industry specialists and auditor non- specialists sub-samples and conducts tests separately for each of the sub-samples. We define the following regression:
5. Empirical Results
5.1. Descriptive statistics and univariate tests
Table 1 presents descriptive statistics for earnings quality, auditors' industry specialization, auditor tenure and other variables used in the study. Approximately 38% of sample companies are audited by an industry specialist. Univariate tests show that tenure and auditors' industry specialization are positively associated with earnings quality. Also, result show that auditors in their primary, second and third years with a client (SHORT) have significant negatively effect on earnings quality. However, auditors with tenure of fourth, fifth and sixth year (MEDIUM) have significantly positive impact on earnings quality. Furthermore, auditors with seven or more years (LONG) have no significant effect on earnings quality. Finally, the univariate test report evidence that size and debt are not statistically significant related with earnings quality.
Table 1: Descriptive statistics and univariate tests
valeur
overall groupes
Intra-groupe
Inter-groupe
EQ
Freq
percent
Freq
percent
percent
Variable
SPEC
0
1
67
42
61.47
38.53
15
9
75.00
45.00
84.81
80.77
3.04***
(0.000)
SHORT
0
1
67
42
61.47
38.53
19
17
95.00
85.00
65.05
44.21
-2.60***
(0.009)
MEDUIM
0
1
71
38
65.14
34.86
20
15
100.00
75.00
65.14
46.34
2.89***
(0.004)
LONG
0
1
84
25
77.06
22.94
20
13
100.00
65.00
77.06
35.21
0.21
(0.832)
Mean
St.Deviation
Minimum
Maximum
EQ
TENURE
Intra-groupe
Inter-groupe
4.495413
1.132198
2.150222
1
9
2.04**
(0.041)
SIZE
Intra-groupe
Inter-groupe
17.89204
.9410656
.4527693
16.48924
21.90919
0.46
(0.646)
DEBT
Intra-groupe
Inter-groupe
7.279099
28.21619
65.75156
0
1.090521
-0.16 (0.871)
The statistics reproduced in the table are the average (Mean), the standard deviation (St Deviation), the minimum (Minimum) and the maximum (Maximum), frequency (Freq) and percentage (Percent).
SPEC= 1 if the firm is audited by a specialist, 0 otherwise. TENURE= the numbers of consecutive years that the auditor proceeded to the audit of the firm. SHORT is equal to 1 if TENURE ≤ 3, and 0 otherwise. MEDUIM is equal to 1 if 4 ≤ TENURE ≤ 6, and 0 otherwise. LONG is equal to 1 if TENURE ≥ 7, and 0 otherwise. SIZE= Ln market value of equity. DEBT= Ln of the ratio of short and long term debt to total equity for fiscal year t.
EQ = adjusted R-square for the following model :
5.2. Tests on Panel Data
The panel data necessite certain tests, has knowledge: test of presence of individual effect, Hausman test, Heteroscedasticity test and Correlation test. The results of these tests are reported in table 3, 4 and 5. Table 3 shows the presence of individual effect. The hausman test is a test of specification which makes it possible to determine if the coefficients of the fixed and random estimates are statistically different. The probability of the test is 86% implies that the individual effect is random effect. The heteroscedasticity justified our choice of random effect model. Indeed, test indicates significant problems with heteroscedasticity.
Table 3: Test of panel data
Test of presence of individual effect
80.58 ***
(0,000)
Hausman test
2.51
(0.8677)
Heteroscedasticity test
5.14
(0.0234)
Spearman correlation coefficients are reported in table 4 and table 5.Table 4 shows that no problem of co-linearity can be detected between all variable retained in the same model. Table 5 indicate a co-linearity problem exist with SHORT, MEDUIM and LONG variable. To resolve this problem, these variables are introduced one by one into the model.
Tableau 4: Correlation matrix of Spearman
SPEC
SHORT
MEDUIM
LONG
SHORT*SPEC
TAILL
SHORT
-0.1620
0.0923
MEDUIM
0.0933
0.3347
-0.5792
0.0000
LONG
0.0613
0.5266
-0.4319
0.0000
-0.3991
0.0000
SHORT*SPEC
0.4285
0.0000
0.4285
0.0000
-0.2808
0.0031
-0.2094
0.0288
SIZE
-0.0971
0.3154
-0.0102
0.9163
-0.1970
0.0400
0.2171
0.0234
-0.1499
0.1198
DEBT
0.0665
0.4920
-0.1749
0.0688
0.0694
0.4730
0.1009
0.2965
-0.0113
0.9069
0.4147
0.0000
Tableau 5: VIFs of the independent variables
SPEC
SHORT
MEDUIM
LONG
SHORT
SIZE
DEBT
VIFs
1.55
7.26
7.19
5.75
1.85
1.38
1.29
5.3. Empirical Results
Table 7 reports regression results for H1 and H2. Return-earning relation is our proxy for earnings quality. Table 7 report results using indicator variables for specialization along with categorical variables for audit tenure (SHORT, MEDUIM and LONG). Furthermore, we use a random effect model.
The coefficient on specialization is positive and statistically significant (two-tailed p-value < 0.01). Consistent with H1, earnings quality is, on average, higher for firms audited by auditors' industry specialization. These results are consistent with auditor industry specialization provide higher audit quality and increase earnings quality. These results are consisting with industry-specialist audit firms possess more experience and more knowledge with industry-specific than non-specialist audit firms and perform audit mission.
In the same way, auditors in their primary, second and third years with a client (SHORT) have significantly lower earnings quality than MEDIUM auditors in their fourth fifth and sixth year with a client. We find that auditors with tenure of seven or more years (LONG) have a negative and not significant impact on earnings quality. Result provides that medium auditor tenure is the best period of audit mission. Indeed this period is characterized by a better audit quality which affects positively the earnings quality of the audited companies. For three first one years of the audit mission of audit (first audit mandate) it turns out that lack of knowledge of auditors of the characteristics of audited companies reduced the audit quality and the earnings quality of client.
The control variable SIZE and DEBT are not statistically significant. This result can be explained by the sample size. But, it this is bound for the characteristics of the Tunisian financial market.
Table 7: Regression results
Model 1.1
Model 1.2
Model 1.3
Variable
PS
ßit
P>|z|
ßit
P>|z|
ßit
P>|z|
intercept
?
-.1360673
0.819
-.1406476
0.810
-.2111847
0.723
SPEC
+
.1869372***
0.009
.209704***
0.002
.2261039***
0.002
SHORT
-
-.0802524***
0.051
-
-
-
-
MEDUIM
+
-
-
.1091942***
0.007
-
-
LONG
+
-
-
-
-
-.0158128
0.747
SIZE
+
.0280315
0.394
.0236305
0.469
.0292828
0.378
DEBT
-
-.1054021
0.313
-.0930155
0.361
-.0800819
0.451
Model 1.1: Wald chi2(6) = 14.70***; within = 0.1463; between = 0.0281; overall = 0.0574
Model 1.2: Wald chi2(6) = 18.67***; within = 0.1640; between = 0.0877; overall = 0.1045
Model 1.3: Wald chi2(6) = 10.44***; within = 0.1022; between = 0.0611; overall = 0.0667
Number of observation = 109. Are reproduced in this table the coefficients of the linear estimate (ßit) and the coefficients of regression (statistical of Wald chi2 `Z') relative to each variable included in the corresponding model. *, ** and *** indicate a bilateral critical probability to the threshold of 10%, 5% and 1% respectively. The statistics of khi-deux Wald and test and the number of observation are also presented. PS corresponds to the predicted signs.
SPEC= 1 if the firm is audited by a specialist, 0 otherwise. TENURE= the numbers of consecutive years that the auditor proceeded to the audit of the firm. SHORT is equal to 1 if TENURE ≤ 3, and 0 otherwise. MEDUIM is equal to 1 if 4 ≤ TENURE ≤ 6, and 0 otherwise. LONG is equal to 1 if TENURE ≥ 7, and 0 otherwise. SIZE= Ln market value of equity. DEBT= Ln of the ratio of short and long term debt to total equity for fiscal year t.
EQ = adjusted R-square for the following model :
In order to evaluate the role of auditor industry specialists on the association between auditor tenure and earnings quality, we conduct two tests. First, we divide the total sample into auditor industry specialists and auditor non-specialists sub-samples. Then we conduct tests separately for each of the sub-samples. The result on the sub-samples, as reported in Table 7, show that the medium tenure coefficient continues to be positive and significant for the specialist sub-sample, suggesting that earnings quality increasing when auditor in their fourth fifth and sixth year with a client and companies are audited by specialists. The coefficient for the auditors in their primary, second and third years with a client have a negative and significant impact on earnings quality, suggesting that short tenure have significantly lower earnings quality than medium tenure. Consistent with our expectation, this finding suggests that auditor industry specialization affect the association between auditor tenure and earnings quality. The tenure coefficient for non-specialist sub-sample, however, has statistically not significant impact on earnings quality. This result suggests that the association between auditor tenure and earnings quality is significantly higher for firms audited by industry specialists.
Table 8: Multivariate Models Explaining Bid-Ask Spread
Panel A: Earnings quality for Companies Audited by Specialist
Model 3.1
Model 3.2
Model 3.3
Variable
PS
ßit
P>|z|
ßit
P>|z|
ßit
P>|z|
intercept
?
-1.592975
0.254
-1.925862
0.173
-1.85804
0.227
SHORT
-
-.2088725***
0.009
-
-
-
-
MEDUIM
+
-
-
.1596028**
0.021
-
-
LONG
+
-
-
-
-
.0710285
0.434
SIZE
+
.1219673
0.125
.132716
0.101
.1290914
0.144
DEBT
-
-.0709408
0.672
-.0248142
0.881
.0541238
0.753
Number of observation = 42
Model 3.1: Wald chi2 = 10.76***; within = 0.2678; between = 0.0018; overall = 0.0834
Model 3.1: Wald chi2 = 9.06***; within = 0.2379; between = 0.0052; overall = 0.0933
Model 3.1: Wald chi2 = 3.78; within = 0.1177; between = 0.0039; overall = 0.0247
Panel B: Earnings quality for Companies Audited by Non-Specialist
Model 3.1
Model 3.2
Model 3.3
Variable
PS
ßit
P>|z|
ßit
P>|z|
ßit
P>|z|
intercept
?
.2056728
0.744
.2911793
0.643
.3060552
0.621
SHORT
-
-.0191145
0.713
-
-
-
-
MEDUIM
-
-
-
.0690179
0.214
-
-
LONG
-
-
-
-
-
-.0781096
0.186
SIZE
+
.0088635
0.798
.0019082
0.956
.0034054
0.921
DEBT
-
-.1821387
0.202
-.1635973
0.250
-.1708795
0.225
Number of observation = 67
Model 3.1: Wald chi2 = 1.91 ***; within = 0.0435; between = 0.0005; overall = 0.0181
Model 3.1: Wald chi2 = 3.40; within = 0.0599; between = 0.0130; overall = 0.0549
Model 3.1: Wald chi2 = 3.46; within = 0.0560; between = 0.0321; overall = 0.0897
Are reproduced in this table the coefficients of the linear estimate (ßit) and the coefficients of regression (statistical of Wald chi2 `Z') relative to each variable included in the corresponding model. *, ** and *** indicate a bilateral critical probability to the threshold of 10%, 5% and 1% respectively. The statistics of khi-deux Wald and test and the number of observation are also presented. PS corresponds to the predicted signs.
SPEC= 1 if the firm is audited by a specialist, 0 otherwise. TENURE= the numbers of consecutive years that the auditor proceeded to the audit of the firm. SHORT is equal to 1 if TENURE ≤ 3, and 0 otherwise. MEDUIM is equal to 1 if 4 ≤ TENURE ≤ 6, and 0 otherwise. LONG is equal to 1 if TENURE ≥ 7, and 0 otherwise. SIZE= Ln market value of equity. DEBT= Ln of the ratio of short and long term debt to total equity for fiscal year t.
EQ = adjusted R-square for the following model :
6. Conclusion
Prior literature and professional press suggest that the auditor's specialization could increase the perceived quality of firm's reported earnings (Carcello and Nagy, 2002; Balsam and al., 2003; Krishnan, 2003; Velury, 2003; Jenkins and al., 2006; Gul and al., 2009). Recent scandals have spawned intense scrutiny of the length of audit tenure. Among the most widely debated issues is the potential for the he length auditor-client relationships lower audit quality. A major research's support the hypotheses that the length of auditor tenure could lead to reduced auditor independence and, subsequently, has led to lower audit quality (Meyer and al., 2007; Fairchild, 2008; Davis and al., 2009). Other research concern expressed is that extended auditor-client relationships are associated with higher earnings quality (Geiger and Raghunandan, 2002; Gul et al., 2007) and higher audit quality (Libby and Frederick, 1990; Myers et al., 2003; Carcello and Nagy, 2004; Iyer and Rama, 2004; Mansi and al., 2004; Ghosh and Moon, 2005; Jackson and al., 2008). Our study contributes to this important debate by providing empirical evidence on the association between audit tenure and earning quality, as measured by return-earnings relation. Furthermore, we examine the effects of auditor's industry specialization on the association between audit tenure and return-earnings relation. Based on a sample of 20 companies all listed in the Tunisian Stock Exchange, with full financial data and information on the audit quality from 2000 to 2005, our empirical findings suggest that auditor's industry specialization product higher-quality services, and thus provide better earnings quality than non-specialists audit firms. Additionally, results support the assumption that medium auditor tenure (auditors in their fourth fifth and sixth year with a client) is the best period of audit mission. Indeed this period is characterized by a better audit quality which enhances greater earnings quality. More specifically, we suggest that auditors in their primary, second and third years with a client have statistically negatively and significantly effect on return-earnings relation. By comparing the relative effects of audit industry specialization on the association between audit tenure and earnings quality, we find that audit-firm specialization plays a key role in determining the quality of earnings.
Our study contributes to the literature on audit tenure by suggesting that the association between lengthy auditor-client relationships and earnings quality is affected by auditor's industry specialization. We believe that our finding is subject to the valuable implications of empirical studies of this type. Nonetheless, our findings may not fully reflect audit industry specialization and audit tenure effects in other countries, our results are only verified in the Tunisian context. Second, validity of our finding depends in general on the appropriateness of the proxies for earnings quality. However, in the accounting literature, there is no consensus on the superiority of model for estimating discretionary accruals or earning conservatism (Bartov and al., 2000; Dechow and al., 1995; Nichols, 2000). Third, our tests are based on a sample that include firms audited by auditor's industry specialization versus non-specialization and not consider the impact of auditor's Big and auditor's non-Big on earnings quality. Finally, our evidence is can be improved by a use of different measure of earnings quality and a comparison between countries.
