The Impact of the H3C on Auditor Independence in French Context

The H3C was created by the LSF Act of 2003 (LSF) to regulate the auditing profession in France. It is considered as an external control authority. The auditor independence is considered as the first preoccupation of this new authority. This paper focuses on the impact of the H3C on auditor independence in France context from 2002 to 2007. The first hypothesis is related to the establishment of the H3C and auditor independence and the second deals with the relation between the start of the inspection program and the auditor independence. Our data was obtained from Thomson Financial Data for 140 enterprises. The empirical results show that the auditor independence increases only after the creation of the H3C. Contrary to this result, the auditors' independence did not increase significantly after the announcement of the inspection process because the inspection process started in 2007.


Introduction
At the end of the 2000s, a series of large company failures, assigned in majority to auditor independence, announced in the US context.This is due to the lack of the corporate governance law.The history marks the disappearance of the Arthur Andersen LLP (one of the Big auditor) following the bankruptcy of the Enron Corporation.To restore the confidence on financial market, several countries adopted law securities act.The majority of countries adopted the same approach of the Sarbanes-Oxley Act of 2002 approved by US-Congress.This recent act established the Public Company Accounting Oversight Board (PCAOB).This non-profit corporation oversees the audits of public companies in order to protect the interests of investors and further the public interest in the preparation of informative, accurate and independent audit reports.The PCAOB also oversees the audits of broker-dealers, including compliance reports filed pursuant to federal securities laws, to promote investor protection.
In the French context, the parliament adopted a similar act to that of the Sarbanes-Oxley Act in 2003.This new reform (Lois de la Sécurité Financière, LSF) established the Haut Conseil du Commissariat aux Comptes (H3C).It's considered as an external control authority to the audit profession.The High Council of statutory auditors (H3C) contributes to a better transparency of the police function auditors and strengthens its control.It's established under the Minister of Justice, it ensures the profession surveillance with assistance from the National Society of Certified Public Accountants (CNCC).To sum up, this new authority searches to inspect the auditors and guarantee their independence and the respect of the ethical code.This reform will help ensure improving the auditor independence and audit quality after a scandal series of 2002.
The modern academics studies focus on the effectiveness of these new reforms and its impact on auditor independence and audit quality.Prior research in accounting has treated the effects of inspections of those new authorities on audit quality and auditor independence in Anglo-Saxon context.Lennox & Pittman (2010) focused on the impact of the recent reforms on the external monitoring audit firms.They find that the PCAOB improve the quality of the inspected audit firms.In the same perspective, DeFond & Lennox (2011) treated the effect of SOX on small auditor and audit quality.They find that PCAOB inspections improve audit quality in the US context.Other researchers have examined audit firms' and clients' reactions to PCAOB inspection reports.They demonstrate that small audit firms with deficient inspection outcomes have deregistered with the PCAOB (DeFond & Lennox 2011;Daugherty et al., 2009).Results are inconclusive and depend on context on whether deficient inspection reports have influenced audit firms to adjust audit quality and their independence.In conclusion, the PCAOB inspection reports succeeded in providing meaningful information about audit quality in the US context.
In France, the lack of studies on this subject leads us to focus on the effectiveness of H3C.This control authority (H3C) influences the incentives applicable to the legal audit, ethical and deontology.To study this impact in the French context, we use two assumptions in connection to the creation of the H3C in the French context.The first deals with the impact of the establishment of the H3C, and the second study the effect of the starting of the H3C inspection program on auditor independence.
To study this impact of the H3C on auditor independence in French context, we use the data of the listed firms on SBF 250 obtained from Thomson Financial databases.These selection criteria yield a final sample of 140 firms issued between 2002 and 2007.To test our hypotheses, we use the total audit fees as a proxy of auditor independence.The empirical findings demonstrate that the announcement of the establishment of the H3C increase significantly the auditor independence.This suggests that the establishment of the H3C has a positive effect on the auditor independence.But announcing the start of the permanent inspection program does not have a significant influence on the independence of the audit in the environment characterized by the lack of investor protection.
The remainder of this paper is organized as follows.Section 2 provides relevant background on the institutional function of the H3C.Section 3 describes prior literature on the effects of PCAOB inspections.Section 4 outlines the research design and discusses the sample.Results are presented and discussed in section 5 followed by a conclusion in section 5.

H3C: The Institutional Aspect
After the failures at the end of 2001, the Sarbanes-Oxley Act was promulgated in the United States of America to improve the auditor's independence and to protect investors.To ensure the auditor independence and the financial statement credibility, the Sarbanes-Oxley Act (2002) established the Public Companies Accounting Oversight Board (PCAOB).According to paragraph 101 of the Sarbanes Oxley-Act the PCAOB is required to "oversee the audit of public companies that are subject to the securities laws, and related matters, in order to protect the interests of investors and further the public interest in the preparation of informative, accurate, and independent audit reports for companies the securities of which are sold to, and held by and for, public investors".Therefore, the PCAOB board considers the inspection program as an important determinant of the audit quality and auditor independence (PCAOB, 2010a).
In Europe, the series scandals of 2001-2002 accelerated the debate of the European Commission on corporate governance in 2002.The main purpose of this discussion is to prove new rules on auditor independence and to provide the stakeholders with a high-level of insurance through the audit.To guarantee the success of these new reforms, the European Commission has proposed the basic principals to ensure the independence of fact and appearance of the auditors (EC, 2002).But, the EC recommendations of 2002 have ignored the audit fees consulting from these principles.
In France, a similar security act to the Sarbanes-Oxley was passed in 2003.This new financial security law established new rules on auditor independence and ethical aspect.In order to guarantee the effectiveness of these new rules, the Haut Conseil du Commissariat aux Comptes (H3C) was established by LSF as the regulator of the accounting firms that inspect the auditor.The composition of the H3C is as follows: -Three judges from court of cassation -The court of auditors and the judiciary and the president being a judge of the court of cassation; -The chairperson of the financial market authority; -Representative of the ministry of economy; -A university professor specializing in legal, economic or financial studies.
-Three qualified persons in economic and financial domain: two of them have the expertise in an initial public offering and one has the expertise in the small and medium sized business or association field.
Three CPA, two of them has an experience in the auditing of the person using public offering or public domain.Given the passage to the Sarbanes-Oxley Act, Raghunandan & Rama (2006) studied the audit fees before and after the adoption of Sox for 660 manufacturing firms.They found that audit fees were about 86 percent higher in the post-SOX period.Wang & Zhou (2012) also studied the impact of the PCAOB Auditing Standard No. 5 on audit fees and audit quality.They used pre-and post AS5 period.The empirical results demonstrate that audit fees decreased following the implementation of AS5.However, there was no impact on audit quality (using abnormal accruals and meet or beat analysts' forecasts as measures of audit quality).
Other researches treated the impact of the disclosure of the PCAOB report on audit quality.The disclosure of the PCAOB reports on audit firms deficiencies has a negative effect on audit firms (Daugherty et al., 2011), and appeared more sever for the smaller audit firms (Gilbertson & Herron, 2009).Gramling et al. (2011)  In France, the regulator and the professional organization have incentives to perform auditor independence and quality audits.Specifically, the establishment of the new organism, like the PCAOB, will cause accounting firms to improve the quality of their services and their independence in order to guarantee their reputation and to avoid penalties imposed by regulators.The establishment of the H3C, in particular, produces strong incentives for the auditors to ensure their independence for more than one reason.For example, the reputational loss that can result from the detection of the economic dependence and subsequently the loss of the audit market share.Therefore, the auditor independence should be greater in the period after the establishment of the H3C compared to the period before the establishment of the H3C.Our first hypothesis, stated in the alternative, is: H1: Auditor independence is increased after the establishment of the H3C compared to the period before the establishment of the H3C.
The establishment of the H3C only, is an insufficient reaction to ensure the achievement of the objectives of the Law security Act of 2003 in France.But it required the establishment of a permanent inspection program by the H3C.For this reason the juridical ministry approved the H3C inspection permanent program at the starting of 2006 (H3C, 2006, 2007, 2008).In order to avoid the possibility of financial penalties and temporary or permanent prohibition, the auditor must be more independent after the announcement of the start of the inspection program by the HC3.In order to prevent a consequence of unfavorable report inspection and penalties, the auditors are expected to adopt a measure that can guarantee the audit quality (Carcello et al., 2011b) and auditor independence and unite their effort to prevent future deficiencies in their work (Carcello et al., 2011b).
The announcement of the starting of the H3C inspection in 2006 should also alter the difference in auditor independence between the establishment of the H3C and the announcement of the staring inspection program.In the period prior to inspections, the auditor independence may eventually be lower than that after the announcement of the start of inspection program.If inspections have the desired effect of improving auditor independence that should increase in the post-inspection period compared to the pre-inspection period.We therefore posit that in the post-inspection period, the auditor independence has improved more relatively to the pre-inspection period.Our second hypothesis stated in the alternative is: H2: Auditor independence is greater after the pre-inspection compared to the period before the post-inspection.

Methodology
Audit  To examine changes in fees in response to the establishment of the H3C and the start of the inspection program, it is important to control changes in underlying client characteristics.We estimate the following pooled audit fees regression model for the sub-period using ordinary least squares regression: In this model, LNFEE is measured as the natural logarithm of audit fees.The client-specific explanatory varaiables are selected based on the meta analysis of audit fees studies by Hay et al. (2006) and Amir et al. (2010).As a measure of client size, we include SIZE, the natural logarithm of total assets and the growth (total sales of next year divided by the total sales of current years) 1).We include as additional control variables leverage, which equal to the company's long term debt divided by total assets.Indicator of client risk is ROA, measured as net income divided by total assets.Risk 1 and Risk 2 (view the annexe).To account for client complexity, we include ASTR equal to the gross, property, plant and equipment to total assets and MOM, which equals 1 if the firm is listed on more than one market, 0 otherwise.Finally, we include an indicator of auditor reputation, BIG, which equals 1 if one of the two legal auditors is one of the Big 4 network, 0 otherwise.

Results
Table 2, Panel A to C, provides an overview of the descriptive statistics for the period 2002 through 2007.Audit fees paid by the clients lie between 40.000 euro and 44 million euro with a mean of 1,3 million Euro.The mean size, measured by the natural logarithm of total assets, is equal to 14.093.50% of companies have total assets greater than 13.76 or the equivalent of 918 043 euro.While the mean of asset structure is equal to 16.9%, this demonstrates that French firms tend to invest in intangible assets.For frequency variables in our model, we find that 242 observations are related to companies listed on several exchanges.Variable for Big 4, we find 220 observations related to the Non-Big Four auditors.Finally, most companies have an audit committee.measured as the natural logarithm of audit fees.SIZE, the natural logarithm of total assets.GRW (total sales of next year divided by the total sales of current years) 1).LEV, which equal to the company's long term debt divided by total assets.ROA, measured as net income divided by total assets.Risk 1 and Risk 2 (view the annexe).ASTR equal to the gross, property, plant and equipment to total assets.Table 3 presents our univariate analysis.It includes tests of independent sample comparison.The objective of this analysis is to compare the audit fees variable before and after the establishment of the H3C.The mean and the median of the sub-period are presented respectively in the column A and B. The C column represents the difference of mean and median reported in the column A and B. We use the t-test and the Wilcoxan two-sample test to compare the difference of mean and median between the two sub-periods.For the sub-period 2002 and 2003 (Pre-H3C), the mean of natural logarithm of total fees is equal to the 13.808 compared to the Post-H3 (14.036).This result demonstrates that audit fees increased in the psot-H3C.The mean and the median test of difference is equal to 1.643 and is significant at 5%.This result demonstrates that LNFEE increased after the establishment of the H3C.However, the differences tests of mean and median for the period of the start of the inspection program does report the significant increase of audit fees for 2006 to 2007.The correlation matrix presented in table 4 shows that there are several statistically significant correlations between independent variables (audit fees and firm size =0.80; and size and audit committee= 0.46).The Variance Inflation Factors (VIF) is low (1.12), which shows the presence tolerant multi-collinearity between variables and does not affect negatively the estimation.The correlation matrix further indicates that there is a negative correlation between size and firm performance.This result provides an idea on agency problem.Finally, the positive correlation between Big Four audit, asset structure and firm size demonstrate that firms with agency problem and high level of assets choice Big Four auditors (Azibi et al., 2011).This is due to their audit expertise compared to the other auditors.measured as the natural logarithm of audit fees.SIZE, the natural logarithm of total assets.GRW (total sales of next year divided by the total sales of current years) 1).LEV, which equal to the company's long term debt divided by total assets.ROA, measured as net income divided by total assets.Risk 1 and Risk 2 (view the annexe).ASTR equal to the gross, property, plant and equipment to total assets.MOM: Dummy variable, which equals 1 if the firm is listed on foreign financial market, 0 otherwise; BIG: Dummy variable, which equals 1 if one of the two legal auditors is one of the Big 4 network, 0 otherwise.Notes.***, ** and *denote significance at p < 1%, 5% and 10% respectively.LNFEE is measured as the natural logarithm of audit fees.SIZE, the natural logarithm of total assets.GRW (total sales of next year divided by the total sales of current years) 1).LEV, which equal to the company's long term debt divided by total assets.ROA, measured as net income divided by total assets.Risk 1 and Risk 2 (view the annexe).ASTR equal to the gross, property, plant and equipment to total assets.MOM: Dummy variable, which equals 1 if the firm is listed on foreign financial market, 0 otherwise; BIG: Dummy variable, which equals 1 if one of the two legal auditors is one of the Big 4 network, 0 otherwise.Table 5 reports the regression results.The empirical results show that for three regressions, the variable size is positively significant at 1%.The accounting theory confirms this result.This means that when the size of the company is important, auditors carrying out an additional systematic control in order to evaluate the accounting system.In this situation, the audit fees are in fact improving.However, the growth variable is only significant at 1% for the two first estimations.This is an indirectly indicator of risk.This means that firms with high client credit have a potential risk and influence directly on the fees of auditor.The variable Risk 2 confirms this relation after the establishment of the H3C.

Variables
For listed companies in foreign financial market, the results of the regression show the presence of a positive and significant relation at 10% only for the first two estimations.This clearly indicates that if a company is quoted on foreign financial market, this requires extra efforts needed to be made to present the financial statement in conformity with the national and International standards.Otherwise, the auditors prepared with IFRS norms should make an additional effort to prepare a consolidated financial statement with US-GAAP.This necessary hypothesis increases the audit fees.
Contrary to our expected results, ASTR variable is not significant this shows that the criterion of industrial expertise is not taken into account in the discussion of the audit fees.This coefficient becomes positive and insignificant for the three sub-period regressions.This fact shows clearly that industrial expertise becomes an insignificant additional determinant of audit fees.
The variable Big is positively significant only for the third estimation at 10%.This result shows that after 2003 and the role of the big auditors in the scandals series, this group of auditor has regained their reputations on audit market.This situation favors the bargaining power of audit fees between the Big Four and the enterprises.In other words, audit fees are positively correlated with the Big Four choice after 2005 For our hypothesis, the empirical results presented in Table 5 show that the R2 before the establishment of the H3C (during 2003 and 2003) is equal to 69.84% and Adj.R2 equal to 68.82%.These two indicators allow the comparison of the three outputs.The estimated output shows an increase during the period 2004-2005 (R2 = 71.29%),just for post-H3C.The estimation year per year show that the annual maximum Adj.This shows that just after the establishment of the H3C, the independence of auditors increased in 2004 and 2005 compared to 2002 and 2003.This result confirms our first hypothesis and demonstrates that auditor independence increased after the creation of the H3C.However the announcement of the inspection starting activity of the H3C does not have a positive effect on auditor independence compared to 2004 and 2005 period.The R2 decreased significantly (R2 = 36,59%) This empirical finding rejects our second hypothesis.This result is consistent with the practice, because the beginning of the real inspection program and permanent control activity of the H3C was in 2008.

Conclusion
The announcement of the financial scandals of 2001 accelerated the reform of the audit profession.To restore the investor confidence in financial markets, legislators through the world have established new rules and procedures to ensure the credibility of financial statements and auditor independence.From this perspective, the SOX created the Public Companies Accounting Oversight Board (PCAOB) in the U.S. In France a financial security act (LSF) was promulgated in 2003.The Act has introduced new rules of governance and has devoted a large paragraph for the ethics and auditor independence aspect.In order to ensure the auditor independence, the LSF created the H3C.The primary focus of the H3C is to control the legal audit profession in France.The aim of this paper is to study the impact of the H3C on auditor independence after the establishment and the disclosure of the inspection start activity.The empirical findings demonstrate that auditor independence was improved only after the creation of the H3C.For the announcing of the disciplinary penalties, the empirical results demonstrate that audit independence does not change compared to 2004 to 2005.This result is consistent with the practical aspect, because the announcement of the penalties program was announced in 2008.This new aspect can be explored to test the announcement of the sanctions against auditor and the impact on audit quality on the context which was characterized by the joint audit.
Figure 1.The o fees are observable and provide some indication of the economic relation between the auditor and his client.The fees depend on the auditor reputation and the risk factor.The announcement of the establishment of the H3C and the start of inspection program may cause a change in the auditor's assessment of expected losses arise from potential penalties.Using the audit fees as a proxy of the auditor independence is based on the assumption that total audit fees signals the firm relation firm-auditor and the LSF imposes a restriction on non-audit fees.The sample selection is based on the Thomson Financial Databases.Two criteria have been adopted for the selection.First, every identified firm must have all interest variables in Thomson Financial databases between 2002 and 2007.Second, banks, insurance companies and financial enterprises are removed due to their accounting specifications and their financial regulation.This yields a final sample of 140 firms conducted between 2002 and 2007.Our final sample comprises 840 company-year observations divided to three sub-periods as follow: 2002 and 2003: before the establishment of the H3C.2004 and 2005: After the establishment of the H3C.2006 and 2007: After the start of the inspection program.
variable, which equals 1 if the firm is listed on foreign financial market, 0 otherwise; BIG: Dummy variable, which equals 1 if one of the two legal auditors is one of the Big 4 network, 0 otherwise.
Lamoreaux (2013)with inspection deficiencies are less likely to receive on going-concern opinions.Defond &Lennox (2011)demonstrate that small audit firms with deficient inspection outcomes have deregistered with the PCAOB.Also,Carcello et al. (2011)studied the audit quality subsequent to the first two PCAOB inspections.They use the changes in abnormal accruals between 2004 and 2006 for 4719 firms Big 4 audit-years.They found a significant reduction in abnormal accruals after the PCAOB first inspection and a further reduction in abnormal accruals in the year following the second PCAOB inspection.Gunny & Zhang (2012)find that PCAOB inspections are associated with lower audit quality in the case of the clients of auditor with serious deficiencies.Lamoreaux (2013)studied the PCAOB inspection exposure from 2001 to 2010.He found that audit firms that are subject to the PCAOB inspection process tended to have going concern opinions and report more material weaknesses in internal controls in the case of foreign firms listed in the United States.

Table 1 .
Sample characteristics

Table 3 .
Mean differences TestPanel A: Mean differences test H1