4.5 Additional Tests
4.5.1 Client Size
A sample of audit-engagements tends to be biased towards audits of higher complexity, as evidenced by the Big-4 market share dominance, in the large-client segment. Generally, SMB
& SME clients have a lower demand for technology and resources in comparison to large clients (Vanstraelen, & Willekens, 2008). In other words; the smaller a client, the less complex the audit, and thus, the more reliance placed upon personal knowledge and skills of the auditor.
It is reasonable to assume a positive relationship between size and complexity, as can be seen in Table 4 (Pearson correlations table of audit fee & competition). Table 4, shows univariate correlation that size is positive and economically significant with AFEE (𝛽 = 0.890).
Furthermore, size is positive and economically significant with all complexity variables (BUSSEG, GEOSEG, ISSUE, EXORD, FOREIGN). Therefore, it is reasonable to split the
9 Generally, a Psuedo R2 that exceeds 0.2, demonstrate a well fit measurement.
Variables
Full Sample RESTATE
(1)
Full Sample RESTATE*BIG4
(2)
BIG4 Sample RESTATE
(3)
NB4 Sample RESTATE
(4)
MRKT_COMP -0.0612 -0.159 0.000471 -0.00676
(-0.176) (-0.246) (0.0168) (-0.270)
SIZE 0.0868 0.134 -0.00102 0.00639
(1.309) (1.595) (-0.146) (0.662)
CITY_SPEC -0.164 -0.170 -0.00965 -0.00497
(-1.403) (-1.450) (-1.010) (-0.346)
NAT_SPEC -0.119 -0.107 -0.00811 0
(-0.584) (-0.523) (-0.470) (0)
AFEE 0.294*** 0.293*** 0.0169** 0.0269**
(2.998) (2.977) (2.184) (2.480)
FEE_RATIO 0.161 0.186 -0.00225 0.122**
(0.468) (0.540) (-0.0837) (2.176)
INFLUENCE -0.0818 -0.0658 -0.0271 0.0192
(-0.285) (-0.228) (-0.927) (0.775)
QUAL_SIZE -0.0369 -0.0351 -0.00281 0.00215
(-0.773) (-0.733) (-0.715) (0.474)
LEV -0.190 -0.188 -1.80e-06 -0.0127
(-1.109) (-1.097) (-0.000124) (-1.429)
ACCEL FILER 0.362*** 0.358*** 0.0240* 0.0267
(2.666) (2.635) (1.758) (0.899)
CA 0.174 0.175 0.0184 0.00633
(0.647) (0.651) (0.786) (0.354)
LOSS 0.227** 0.222** 0.0146* -0.00668
(2.274) (2.230) (1.688) (-0.789)
ROA 0.580*** 0.573*** 0.0227 -0.000587
(2.677) (2.648) (1.027) (-0.0647)
LIT 0.0268 0.0242 0.00551 -0.0124
(0.240) (0.217) (0.604) (-1.470)
INVREC -0.312 -0.328 -0.0124 -0.0224
(-0.854) (-0.895) (-0.391) (-0.930)
GROWTH 0.0126 0.0150 -0.000844 0.00152
(0.173) (0.205) (-0.129) (0.338)
MERGER -0.00451 -0.00481 -0.00285 0.00310
(-0.0532) (-0.0568) (-0.422) (0.324)
RESTRUCT 0.113 0.111 0.0100 0.000167
(1.298) (1.269) (1.463) (0.0148)
IC_WEAK 0.459*** 0.455*** 0.0560*** 0.0114
(3.530) (3.501) (3.382) (1.096)
Constant -7.824*** -8.391*** -0.107 -0.429***
(-7.907) (-7.426) (-0.864) (-5.405)
Observations 14,367 14,367 9,919 4,521
Pseudo R2 0.0514 0.0516 0.0259 0.0712
*, **, *** Represent statistical significance at the 0.10, 0.05, and 0.01 levels, respectively. This table presents logistic regression results examining the association between local market competition and client restatements after controlling for other determinants of restatements.
Regression models presented in the table use logistic regression and cluster standard errors by firm. For brevity, coefficients on MSA, Year, and Industry fixed effects are not reported.
Variable definitions are shown in Appendix A.
sample into groups based upon US business size classifications in this analysis10 (SMB, SME, and Large Enterprise, 2021).
Table 9 presents the descriptive statistics for the groups. Table 9, indicates a positive relationship between group size and the Herfindahl Index, consistent with the literature, indicating that the market share is a response to the demand for technology, and quality requirements (e.g., Dopuch et al., 1980; Sirois et al., 2011). Additionally, Table 9 reveals market competition is negatively associated with client size. This can be reasonably assumed, as the transaction costs incurred in attempting to switch auditors is significantly lower for small client. Also a positive association between group size and % of Big 4 auditors exists. Big-4 auditors represent auditors that have achieved large economies of scale, of which is required to
10 According to (SMB, SME, and Large Enterprise, 2021), client size is based off total revenue. The classification is provided below.
Large client: Annual revenue >1,000,000,000,000, SME revenue 10,000,000<x<1,000,000,000,000, SMB Revenue 5,000,000<x<10,000,000,000
TABLE 9: Relationship between audit complexity, audit market competition, audit fees, and audit quality.
Group. Category Obs.
Mean Total Assets
Mean Local Comp
Mean Herfindahl
Index
Mean AFEE
Mean RESTATE
% B4 audit
1 SMB 1033 41.9 0.089 0.185 11.923 0.023 57.1%
2 SMB 443 44.1 0.085 0.181 11.913 0.022 55.7%
3 SME 2850 107.1 0.089 0.189 12.570 0.035 58.8%
4 SME 1333 245.1 0.083 0.190 13.278 0.072 67.2%
5 SME 864 492 0.082 0.193 13.641 0.078 74.2%
6 SME 724 568.1 0.081 0.195 13.773 0.082 79.8%
7 SME 600 734.2 0.074 0.195 13.890 0.082 82.7%
8 SME 475 810.5 0.074 0.191 13.966 0.087 82.5%
9 SME 358 1009.1 0.070 0.194 14.033 0.086 82.3%
10 SME 352 1143.6 0.075 0.196 14.112 0.119 85.5%
11 Large 368 1281.5 0.066 0.194 14.166 0.103 84.7%
12 Large 296 1373.5 0.064 0.198 14.224 0.104 87.5%
13 Large 2801 2559.3 0.064 0.204 14.552 0.080 93.2%
14 Large 970 5421.2 0.060 0.203 14.993 0.088 97.5%
15 Large 2221 31,276.6 0.064 0.201 15.639 0.083 98.0%
Notes: This table displays the average values of client size, market structure measures, audit fees and absolute abnormal accruals for 15 size groups. Size category indicates whether a company is classified as a small or medium-sized client (SME client) or as a large client.
Within each size category, size groups are equally-sized. All variables are as defined in the appendix. The sample used in this table is sample C (n = 15, ).
undertake large complex audits. Therefore, it can be inferred that audit firms choose their supplier based on the technological requirements, agreeing with the intuition above.
A regression analysis, table 10, is utilized to display the relationship between local audit market competition, audit fees and audit complexity. Column 1, displays coefficient estimates for the full sample. Column 2, displays coefficient estimates for the SBE sample, column 3, displays the coefficient estimates for the SME sample and, column 4, displays coefficient estimates for the large client sample. Test scores in Table 10, indicate that the coefficient estimate between AFEE and local market competition for SME clients is negative and economically significant at the 10% level. This finding can be interpreted as evidence that SME clients situated in competitive counties are offered price discounts. These findings are consistent with Van Raak et al., 2020 in relation to audit fee’s for SME clients, however in the private market. Test scores, for SBE clients, column 2, and large clients, column 4, are both insignificant and for brevity I will not discuss the results for these values.
I attempt to separate the effect on Big-4 and NB4 auditors for SME clients, in table 11, to which I find no significant evidence. Once again, as Hypothesis 3 states, there is no significant evidence that the coefficients on local market competition in the against Big-4 and NB4 auditors are differentiated from each other, therefore not in support of Hypothesis 2.
Table 12, uses a logistic regression to display the results of the analyses examining the relationship between local market competition, audit quality and complexity. Column 1 displays coefficient estimates for the full sample. Column 2 displays coefficient estimates for the SBE sample, column 3 displays the coefficient estimates for the SME sample and column 4 displays coefficient estimates for the large client sample. Test scores in Table 11, indicate that the coefficient estimate for local market competition of SMB clients on audit quality is negative and economically significant at the 10% level. I interpret this finding as evidence that SMB’s improve the quality service, in competitive environments in order to entice the clients to the auditors services. Looking at SME and large clients, I see both values are insignificant, indicating similar to Hypothesis 2, in which I find that Industry Market Competition is highly insignificant and negatively associated with audit quality (although significance has improved).
This supports Hypothesis 2 for both SME and large clients. Unablated test results show Hypothesis 4 in the SME client segment to not hold.
Variables
Full Sample (1)
SBE Clients (2)
SME Clients (3)
Large Clients (4)
MRKT_COMP -0.0771 0.107 -0.120* -0.0231
(-1.530) (0.782) (-1.841) (-0.298)
SIZE 0.532*** 0.433*** 0.534*** 0.491***
(94.30) (16.72) (52.40) (44.04)
CA 0.530*** 0.312*** 0.449*** 0.522***
(9.687) (2.961) (6.467) (4.514)
INVREC 0.0608 -0.101 -0.0240 0.304**
(0.937) (-0.735) (-0.294) (2.326)
LEV -0.0852** 0.0671 -0.201*** 0.0102
(-2.407) (1.214) (-4.213) (0.148)
ROA -0.209*** -0.233*** -0.320*** -0.593***
(-6.331) (-4.833) (-6.881) (-4.889)
LOSS 0.105*** -0.0236 0.0809*** 0.0627**
(6.630) (-0.343) (3.895) (2.328)
CR -0.0420*** -0.0243*** -0.0397*** -0.0740***
(-12.19) (-4.138) (-9.630) (-6.596)
BTM -0.0573*** -0.00532 -0.0760*** -0.0523***
(-5.469) (-0.249) (-5.631) (-2.909)
GROWTH -0.0358*** -0.0286** -0.0326** -0.0491
(-4.006) (-2.183) (-2.573) (-1.608)
BUSSEG 0.178*** -0.242 0.0613 0.245***
(6.111) (-1.327) (1.288) (7.247)
GEOSEG 0.138*** -0.226* 0.145*** 0.154***
(6.902) (-1.959) (5.380) (5.913)
ISSUE 0.0736*** 0.130** 0.0544** 0.0529**
(4.192) (2.045) (2.414) (1.995)
EXORD 0.161*** 0.0508 0.156*** 0.156***
(8.744) (0.595) (6.238) (6.418)
FOREIGN 0.260*** 0.0477 0.231*** 0.340***
(13.03) (0.846) (9.558) (9.721)
PENSIONS -0.0452 0.0198 -0.0318 -0.0136
(-1.570) (0.367) (-1.027) (-0.161)
NAT SPEC 0.165*** 0.343*** 0.130*** 0.153**
(4.411) (3.872) (2.694) (2.503)
CITY SPEC 0.0106 0.00676 0.0375 -0.00803
(0.442) (0.0754) (1.029) (-0.262)
JOINT SPEC -0.0134 0.0639 0.00208 -0.130
(-0.245) (0.457) (0.0282) (-1.646)
LIT -0.0325 0.0253 0.0285 -0.128***
(-1.545) (0.438) (1.066) (-3.648)
IC_WEAK 0.0969*** -0.102* 0.146*** 0.183***
(3.394) (-1.834) (4.019) (3.064)
Constant 9.686*** 10.71*** 9.993*** 9.762***
(114.2) (35.96) (84.74) (59.29)
Observations 15,513 1,127 7,665 6,721
Adjusted R-squared 0.857 0.774 0.699 0.732
Notes: ***, **, * indicate 1%, 5% and 10% significance levels, respectively (two-tailed). t values are in parentheses. Standard errors are clustered by firm. For brevity, coefficients on MSA, Year, and Industry fixed effects are not reported. The dependent variable in the regressions i s the natural logarithm of Audit Fees. Variables are defined in the appendix.
Variables
Full Sample SME Clients
AFEE (1)
Full Sample SME Clients AFEE*BIG4
(2)
BIG4 Sample SME Clients
AFEE (3)
NB4 Sample SME Clients
AFEE (4)
MRKT_COMP -0.120* -0.0420 -0.0528 -0.0396
(-1.841) (-0.462) (-0.701) (-0.448)
SIZE 0.534*** 0.443*** 0.356*** 0.493***
(52.40) (40.43) (21.55) (35.26)
CA 0.449*** 0.300*** 0.201** 0.300***
(6.467) (4.814) (2.545) (3.263)
INVREC -0.0240 0.135* 0.427*** -0.0269
(-0.294) (1.863) (3.581) (-0.292)
LEV -0.201*** -0.153*** -0.0432 -0.215***
(-4.213) (-3.519) (-0.739) (-3.380)
ROA -0.320*** -0.265*** -0.174** -0.313***
(-6.881) (-5.982) (-2.250) (-5.835)
LOSS 0.0809*** 0.0757*** 0.0535* 0.0916***
(3.895) (3.859) (1.851) (3.649)
CR -0.0397*** -0.0348*** -0.0284*** -0.0395***
(-9.630) (-9.254) (-5.789) (-7.515)
BTM -0.0760*** -0.0611*** -0.0554*** -0.0591***
(-5.631) (-4.923) (-3.065) (-3.867)
GROWTH -0.0326** -0.0301*** 0.00338 -0.0955***
(-2.573) (-2.603) (0.281) (-4.617)
BUSSEG 0.0613 0.0727 0.0660 0.0587
(1.288) (1.621) (1.259) (0.855)
GEOSEG 0.145*** 0.147*** 0.101*** 0.169***
(5.380) (5.814) (3.155) (4.464)
ISSUE 0.0544** 0.0532** 0.0785*** 0.0300
(2.414) (2.542) (2.885) (1.040)
EXORD 0.156*** 0.171*** 0.143*** 0.196***
(6.238) (7.046) (4.289) (5.984)
FOREIGN 0.231*** 0.207*** 0.207*** 0.181***
(9.558) (9.290) (6.609) (5.913)
PENSIONS -0.0318 -0.0102 -0.0201 0.0194
(-1.027) (-0.363) (-0.604) (0.437)
NAT_SPEC 0.130*** -0.0353 0.0216 0
(2.694) (-0.800) (0.468) (0)
CITY_SPEC 0.0375 -0.0227 0.0117 -0.0762
(1.029) (-0.719) (0.308) (-1.472)
JOINT_SPEC 0.00208 0.0821 0.0476 0
(0.0282) (1.274) (0.733) (0)
LIT 0.0285 0.00362 -0.0294 0.0231
(1.066) (0.147) (-0.949) (0.632)
IC_WEAK 0.146*** 0.172*** 0.286*** 0.116***
(4.019) (5.081) (5.324) (2.927)
Constant 9.993*** 10.16*** 11.03*** 10.14***
(84.74) (91.70) (75.73) (63.44)
Observations 7,665 7,665 4,121 3,544
Adjusted R-squared 0.699 0.736 0.497 0.699
Notes: ***, **, * indicate 1%, 5%, and 10% significance levels, respectively (two-tailed). t-Values are in parentheses. Standard errors are clustered by client firm. Fixed effects for years and industries are included but not tabulated for reasons of bre vity. The dependent variable in the regressions is the natural logarithm of Audit Fees. Variables are defined in the appendix.
TABLE 12: Regression Analysis - Local Competition & Audit Quality & Complexity Full Sample
(1)
SBE Clients (2)
SME Clients (3)
Large Clients (4) Variables
MRKT_COMP -0.0612 -0.0611* -0.0828 0.0166
(-0.176) (-1.716) (-0.169) (0.428)
SIZE 0.0868 -0.00410 0.0342 -0.00881
(1.309) (-0.757) (0.359) (-0.938)
CITY_SPEC -0.164 0.00551 -0.173 -0.0139
(-1.403) (0.246) (-1.012) (-1.051)
NAT_SPEC -0.119 -0.0370 -0.0680 -0.00373
(-0.584) (-1.301) (-0.261) (-0.141)
AFEE 0.294*** 0.0244* 0.581*** 0.0163
(2.998) (1.875) (4.262) (1.429)
FEE_RATIO 0.161 0.390** 1.159*** -0.0403
(0.468) (2.378) (2.578) (-1.154)
INFLUENCE -0.0818 -0.0189 0.189 -0.0280
(-0.285) (-0.957) (0.528) (-0.673)
QUAL_SIZE -0.0369 0.00113 0.101 -0.00564
(-0.773) (0.106) (1.030) (-0.930)
LEV -0.190 -0.000993 -0.157 -0.0192
(-1.109) (-0.134) (-0.697) (-0.898)
ACCEL FILER 0.362*** -0.0426 0.713*** -0.0129
(2.666) (-1.565) (3.811) (-0.901)
CA 0.174 -0.0152 0.0646 0.0999**
(0.647) (-0.610) (0.195) (2.480)
LOSS 0.227** 0.0117 0.0106 0.000403
(2.274) (1.130) (0.0761) (0.0312)
ROA 0.580*** 0.00890 -0.110 -0.164***
(2.677) (1.030) (-0.296) (-2.663)
LIT 0.0268 0.00850 -0.221 0.0177
(0.240) (1.302) (-1.553) (1.338)
INVREC -0.312 0.00539 -0.569 -0.0903*
(-0.854) (0.190) (-1.160) (-1.819)
GROWTH 0.0126 0.00208 0.0730 -0.0117
(0.173) (0.346) (0.811) (-0.662)
MERGER -0.00451 -0.0274** 0.0545 -0.00283
(-0.0532) (-2.192) (0.474) (-0.322)
RESTRUCT 0.113 -0.0248 -0.0525 0.0201**
(1.298) (-0.866) (-0.425) (2.197)
IC_WEAK 0.459*** 0.000249 0.416** 0.0351*
(3.530) (0.0229) (2.429) (1.671)
Constant -7.824*** -0.0704 -12.04*** 0.0725
(-7.907) (-0.500) (-8.857) (0.516)
Observations 14,367 962 6,750 6,423
Pseudo R2 0.0514 0.118 0.0884 0.0331
*, **, *** Represent statistical significance at the 0.10, 0.05, and 0.01 levels, respectively. This table presents logistic regression results examining the association between local market competition and client restatements after controlling for other determinants of restatements.
Regression models presented in the table use logistic regression and cluster standard errors by firm. For brevity, coefficients on MSA, Year, and Industry fixed effects are not reported. Variable definitions are shown in Appendix A.