Joint audit and financial scandal: The case of the French context

The European Commission issued a proposal that encouraged the practice of joint audit on a voluntary basis. It recommended that mandatory audit firm rotation is extended from six years to nine years if joint audits are performed. This is considered as a new mechanism that increases audit quality and auditor independence. This new orientation of the EC Green Paper has an important impact on investor perception during the financial scandal announcement. In order to answer this question, we investigate the stock market reaction of SBF 250 after the disclosure of the financial scandal in presence of joint audit. Our sample is composed of 140 French listed enterprises. We use event study method and OLS regression. The empirical results demonstrate that the stock market of Non Big-Four clients does not significantly react compared to the firms audited by at least one Big Four in France. Contrary to this finding, the stock market reaction of the firms audited by two Big Four was significantly reacted compared to the enterprises audited by one Big at least in France. Those results demonstrate that joint audit with at least one Non-Big encouraged and especially in the Financial scandal periods and resolve the problem associated with the audit market concentration.


Introduction
*Audit quality becomes a contemporary question through the world. Regulators, investors and all third party accorded to this concept an absolutely importance after the scandal series of 2002. For this reason, the significant reforms of the audit profession are the consequence of the government engagement to reduce the conflict of interest and improve the financial confidence on the financial market. The first reform after the contemporary scandal was the Sarbanes and Oxley (2002) in the United States; it was a significant reform of the audit rules since 1934.
In Europe the debate on the audit profession is a fundamental question for different authorities. Through different European directives, the different audit commission goals seek principally to improve the independence of statutory audit firms and auditors from the entity being audited, enhance the informational value to investors of audit reporting, contribute to a more dynamic audit market in the EU, enhance audit supervision and foster convergence and cooperation with non-EU countries. To ensure the achievement of this different goals, the joint audit is one of the mainly proposal, but the lack of the consensus on this suggestion is the most important characteristics of this different committee meeting in Europe. More than European countries required the publishing firms to designate two certified public accounts. France agreed to support this measure. This disposal finds his theoretical foundation that increases the auditor independence and audit quality and ensures a minimum of the financial statement credibility. Other raison explains the strategy that minimizes the audit market concentration and constitute for the small audit firms an opportunity to defend their services quality and their reputation. The different empirical research studied the French regulator orientation on the audit quality. Velte and Azibi (2015) studied the joint audit as instrument for increase the audit quality in Germany and French. They we use a sample of 306 Germany and French companies between 2008 and 2012. Empirical results demonstrate unclear effect of the joint audit on audit quality in these two countries. In the same perspective, Azibi and Rajhi (2013) studied the audit quality in French context after the Enron Scandal between 2002 and 2005. They find that are no difference of audit quality between Big and Non-Big audit firm.
The role of the joint audit became an important in the case of financial market crises due to the financial scandal. To identify this role, we studied the French market reaction after the shredding of the Enron document by Arthur Andersen in presence of two statutory auditors. The aim of this paper is to study the investor reaction in the French context in the presence of two different certified accountants.
The remainder of this paper is organized as follow: Section 2 describe the audit quality concept and develop the hypothesis related to joint audit and the shredding of the Enron document. Section 3 presents the methodology and the measures of different variables. Section 4 presents the interpretation and the discussion of the empirical results. The last section summarizes empirical findings and concludes.

Joint audit and financial scandal: A brief review
Joint audit can be defined as an audit in which financial statements are audited and certified by two or more independent auditors in a way that involves: coordination of the audit planning; shared audit effort; cross reviews and mutual quality controls; and issuance of one single auditor's report signed by the auditors who are jointly liable (Ratzinger-Sakel et al., 2013). In France, joint audit has been mandated since 1966. Starting at 1984, joint audit has been mandatory for all companies preparing consolidated financial statements. The perception of the enterprise decision makers, auditors and other stakeholders on joint audit may differ. For example, Francis et al. (2009) andHarford et al. (2012) examined auditor choice for listed companies in France where two (joint) auditors are required by law. The joint audit requirement creates a unique setting to study if a firm's ownership structure affects its auditor-pair choice as well the consequences on earning quality. The results support predictions from agency theory that higher quality (Big 4) auditors are more likely to be used as external monitors when there is greater separation of ownership and control and increased information asymmetry between the firm and outsiders. A Big 4 auditor (paired with a non-Big 4 auditor) is more likely to be used for firms with more diversified ownership structures and less family block holdings, and these associations are even stronger for firms with two Big 4 auditors conducting the joint audit. For this reasons, some companies may prefer joint audit in order to signal a higher level of audit quality to users of the financial information. Their choice is connected to the Big 4 auditor. Previous studies discussed in the literature review demonstrate that the big audit firm has the ability to respect the standards of independence and competence compared to the small firm. For this reason they considered as the suppliers of the audit quality (Becker et al., 1998;Francis et al., 1999). The return of the investment of this big firm in human resources and their financial potential are determinant to guarantee their reputation and their share market. The big audit firm are enfacing to less litigation risk compared to the others group of auditor (Palmrose, 1988). Lennox (1999) finds that Big Four auditor clients have a level of fraud detection compared to the non-Big Four in UK context. While the audit firms have divergent views on joint audit . For example, Big 4 audit firms suggest that joint audit increases costs; and second-tier audit firms mainly stress the potential enhancement in audit quality and each group of auditors thus appears to seek to protect their private interests . Other stakeholders have mixed positions on joint audit. In their responses to the Green Paper, many investors and associations state that they fear the related increase in the costs of audits, the lack of clear lines of responsibility between the joint auditors and a general lack of advantages of this mechanism (EC (2011). André et al. (2016) studied the association between the joint audit and the audit fees. They compare the audit fees in French, British and Italian context during 2007 to 2010. They find significantly higher audit fees in France after controlling for welldocumented auditor, client, and engagement attributes, which vary across countries. This result raises the question on the relationship between the audit quality and the audit fees.  and  document that joint audits do not have an impact on audit quality. The lack of a positive impact of joint audit on audit quality is also supported by Ratzinger-Sakel et al. (2013) in a matched-pair study of France and Germany. Velte and Azibi (2015) found unclear effect of the joint audit in France and Germany during 2008 to 2012. Contrary to these findings, Zerni et al. (2012) studied the voluntary Swedish joint audit. They find that companies opting voluntarily for joint audits have a higher degree audit quality with two different measures (earnings conservatism and lower abnormal accruals). They also have better credit ratings and lower risk forecasts for insolvency than companies with only one auditor. Zerni et al. (2010) find in the Swedish setting that, compared to single audit cases, firms with joint auditors have the highest perceived audit quality, because the market values joint auditors as a monitoring mechanism that helps prevent the expropriation of minority shareholders. These results demonstrate that joint audit increase relatively the audit quality.
In Financial crises the association between the audit quality and joint is an important question. This question is very significant when one of these joint auditors is one of the Big firm. The scandal of the Enron and the implication of AA in this failure prove the limit of the association between audit quality and big audit firm. After the shredding of Enron document by his auditor, more than study examines the effect of this failure on the financial market reaction in different context. Chaney and Philipich (2002) studied the share reaction of AA clients in the United States. They found that the Enron scandal induces investors' perceptions of impairment of financial reporting and the AA clients suffered a negative abnormal return after the announcement of Enron failure news. This result demonstrates that the AA lost their reputation on audit market. Nelson et al. (2008) have studied the reaction of firms audited by Arthur Andersen and Big Four in the oil industries. The empirical results show a significant negative reaction of Arthur Andersen clients versus the Big Four. Krishnamurthy et al. (2006) (2005) have treated the effects of the announcement of the collapse of Enron and WorldCom on the reactions share firms audited by Arthur Andersen in the United States, United Kingdom and Australia. The empirical results indicate that during the week of adjustment benefits announcing of Enron from 1996 until 2000, the stock reaction of Arthur Andersen clients is not reacted significantly in the United States and UK. While in Australia, the market responds negatively for the companies audited by AA. In Spanish context, (Barbera and Martinez, 2006) studies the effect of Enron scandal on Spanish stock market through the Mercado Continuo Español sample (Madrid, Barcelona, Valencia and Bilbao). The empirical results achieved by authors show that the Spanish market does not react negatively around the announcement of the Enron scandal.
In French context, Evraert and Trebucq (2003) studied the influence effects of the official announcement of the Enron and WorldCom failure on the CAC40 reaction, but they ignore the audit joint scenario. The results show that during the Enron event (October 2001), the French market reactions have not been related to quantitative data, and companies with the clearest off-balance sheets and the best auditors' quality have been sanctioned. On the opposite, during the WorldCom event, market reactions have been linked to quantitative accounting data measuring performance, and to the quality of off-balance sheets. Abnormal returns have also been less negative for companies with a high performance and off-balance sheets of quality. Quantitative accounting data has then explained a significant part of abnormal returns, and financial statements have been obviously used with a higher confidence. In joint audit context, the impact of this failure depends on the joint audit scenario. This joint audit structure influence play an important signal for the investor perception in the scandal period. Joint audit is considered as the mechanism that minimizes the implication of one of the Big Four auditor in financial scandal and resolve the problems associated to the audit market concentration. For example, the Big Four audit firms audit all but one of the FTSE 100 companies, and represent 99% of audit fees in the FTSE 350. Switching rates are low (around 2% on average for FTSE 100 companies), and competitive tendering does not occur frequently. For this reason, the European Commission issued a proposal that encouraged the practice of joint audit on a voluntary basis following the green paper published in November 2011. It recommended that mandatory audit firm rotation be extended from six years to nine years if joint audits are performed (EC, 2011). H1: The stock prices of the enterprises audited by at least one auditor from the Big Four group is more significant than the enterprises audited by two Non-Big Four auditors after the announcement of the financial scandal. H2: The stock prices of the enterprises audited by two Big Four auditors are more significant than the companies audited by two Non-Big Four auditors after the announcement of the financial scandal.

Methodology
The sample selected is composed of 140 French enterprises listed on SBF 250. We use the Enron failure as an example. The period of the study focuses on the day of shredding documents of Enron auditor by AA (January 10, 2002 only). Three criteria have been adopted for the selection sample in this study. First, we excluded the firms audited by AA. Second, every identified corporation must have all interest variables in Thomson Financial and Datastreem databases. Third, banks, insurance companies and financial enterprises are excluded due to their accounting specificities and financial legislation. If we apply these conditions, we find 140 French corporations examined over a period of 2001 (Table 1). To investigate the effect of the joint audit on the investor perception, we use the date of the shredding of Enron document by his auditor: 10 th January, 2002. We compute the abnormal return on day t (ARit) as following: where Rit is there turn for client i on day t, Rmt is there turn on the SBF 250, ᾰ and β are parameter estimates obtained from estimation of the market model for the period November1,2000 to October 31,2001.

Rit = ᾰi + βi Rmt + µit (2)
To investigate the influence of confounding control variables, we use the cross-section analyses. The dependent variable is represented by the cumulative abnormal returns for the window (-1, 1) around the announcement of the shredding Enron documents, calculated through the market model shown in Tables 2 and 3.
To test our first hypothesis, we use the Big Four variable dummy equal to 1 if the enterprise is audited by at least one auditor from the Big Four networks, 0 otherwise.
Based on previous research, we use control variables related to the business characteristics in our model. According to (Easton and Zmijewski, 1989), we use the turnover variation between 1999 and 2000 to capture the business recognition through aggressive revenue potential. For example, to occupy a high responsibility senior position, the offices of some enterprises will be able to increase business via creative action. According to this angle, the Enron officers reported a 151% increase in sales from 1999 to 2000. The long term debt is our second control variable. It is measured by the total longterm debt divided by total assets. This variable is an additional control variable to capture the effects of leverage on stock reaction. According to Chaney and Philipich (2002), the most indebted enterprises are more likely to use off-balance sheet items for funding. We expect a negative relationship between long-term debt and the dependent variable. The size of the firm measured by the natural logarithm of total assets is a supplementary control variable to accommodate the effect of agency problems on stock prices.
According to Nelson et al. (2008), the energy market was reacting negatively after the announcement of the Enron bankruptcy. Based on this argument, we use Oil variable, equal to 1 if the enterprise have oil and gas activity, 0 otherwise to detect the . Finally, we use the MON to control the reaction of the French listed companies on US stock market. This variable is dummy, which equal to 1 if the enterprise is listed on US stock market, 0 otherwise.

CARit = λ0 + λ1 Big4 it + λ2 LTD it + λ3 SIZE it + λ4 GRW it + λ5 Oil it + λ6 MOM it + ζ
( 3) where; CAR it: Cumulative abnormal return; Big4 it : Dummy variable, equal to 1 if the company is audited by at least one auditor which belongs to Big Four networks, 0 otherwise; GRW it: (turnover at the moment t / turnover at the moment t-1)-1; LTD it: Long term debt to total assets; SIZE it: Natural logarithm of total assets; Oil it: Dummy variable, equal to 1 if the company operate in energy sector, 0 otherwise; MOM it: Dummy variable, equal to 1 if the enterprise is listed on US financial market, 0 otherwise.

Results
In contrast to the Chaney and Philipich (2002) study, our research concentrates only on 10 January 2002. The choice of this date is the result of an unexpected behaviour of Arthur Andresen in Enron bankruptcy. According to the analysis of CAR (Cumulative Abnormal Returns, presented in Table  2, between firms audited by auditor at least Big Four and Non-Big Four. The average cumulative return is equal to -0.325 for the first days (0, 1), and at -0.699 with 54.21% of the returns are negative. This shows that the maximum has reached a pick in late second day after the announcement of disappearing of Enron files.
Similarly, the companies audited by Non-Big Four showed negative reactions on 10-January 2002. The maximum was recorded in January 15, 2002 (0, 3). The mean difference tests show that there is no significant difference in cumulative abnormal returns between companies audited by at least one Big Four and Non-Big Four. Table and graphs illustrating the results have been drawn up. We present the long-window CAR (-5, 5) to demonstrate the movement to the CAR during the event reporting period. Each window is marked by a vertical line. Fig.  1 shows that the CAR recorded a decline, but the parametric tests to (0, 3), demonstrate an insignificantly difference.  To test our first hypothesis, we use ordinary-least square with a dependent variable CAR. The results of the first regression are shown in Table 4. Empirical results indicate clearly that the Big 4 coefficient is negatively insignificant. This indicates that investors on SBF 250 do not draw a particular attention when the enterprise was audited by at least one Big or two Non-Big-Four.
By contrast, the maximum of CAR (1.981) was recorded on the second day after 10 January 2002, with a percentage of negative operation equal to 81.6. For, one Big-Four clients, the maximum of the abnormal returns (-0.356) was recorded the third days after January 10, 2002, with 63.14% of negative transactions. The mean difference tests show clearly that there is a significant difference between the average CAR of companies audited by two Big 4 and 1Big 4 at 1% for the three windows. Fig. 2 shows this difference.

Fig. 2:
Difference between the average CAR of companies audited by two Big 4 and 1Big 4 at 1% for the three windows In order to confirm this result, we use a second regression. The empirical outputs related to the second hypothesis are presented in Table 5. The results show that the coefficient Two Big4 is negatively significant at 1%. This shows that the stock market reaction of two Big clients was reacting negatively more than the enterprises audited by 1BIG and 1Non-Big. This clearly indicates that investors in the SBF 250 pay a particular attention to the financial statements of two auditors Big-Four clients. This provides information on a confidence problem relate to the financial reports of Big-Four clients. This supports our second hypothesis and confirms that the firm audited by two Big Four was largely reacted in France.
For the control variables, the coefficient of debt is not significant for both regressions. This shows that the investors on the SBF 250 revise slightly their estimates for indebted enterprises audited by the two groups of auditors. However, the market revises their estimate for the companies listed on the United States financial market as part of the first regression. Furthermore, the investors in the SBF 250 pay a particular attention for the enterprises of gas and oil.

Conclusion
The European Commission issued a proposal that encouraged the practice of joint audit on a voluntary basis following the green paper published in November 2011. It recommended that mandatory audit firm rotation be extended from six years to nine years if joint audits are performed. This is considered as a new mechanism that enhances audit quality and auditor independence. This position of the EC Green Paper has a significant impact on investor perception, exceptionally during the financial scandal crises. The debate in literature review reveals a lack of consensus on joint audit mechanism. The aim of this paper is to study the role of the joint audit in financial scandal period in France. To achieve this goal, we use the failure of the Enron Corporation as an example.
Since the end of the 2001, a number of major financial scandals were announced in the US. More than the question was related to the financial transparency, greater accountability and compliance with the GAAP.  Legend: *, **, ***: significant coefficient at the level of 10 %, 5 % et 1 % Variables definition : CAR it: Cumulative abnormal return; Big4 it: Dummy variable, equal to 1 if the company is audited by at least one auditor which belongs to Big Four networks, 0 otherwise; GRW it: (turnover at t / turnover at t-1)-1; LTD it: LT Debt/ total assets; SIZE it: Ln of total assets; Oil it: Dummy variable, equal to 1 if the company operate in energy sector, 0 otherwise; MOM it: Dummy variable, equal to 1 if the enterprise is listed on US financial market, 0 otherwise. Legend: *, **, ***: significant coefficient at the level of 10 %, 5 % et 1 % Variables definition : CAR it: Cumulative abnormal return; Big4 it: Dummy variable, equal to 1 if the company is audited by at least one auditor which belongs to Big Four networks, 0 otherwise; GRW it: (turnover at t / turnover at t-1)-1; LTD it: LT Debt/ total assets; SIZE it: Ln of total assets; Oil it: Dummy variable, equal to 1 if the company operate in energy sector, 0 otherwise; MOM it: Dummy variable, equal to 1 if the enterprise is listed on US financial market, 0 otherwise.
There is research that shows that one of the causes of the scandal series is the financial audit process. The announcement of the Enron scandal justifies this hypothesis. The first thematic studies have examined the stock market reaction in a US context after Enron failure. The vast majority of research is based on the study of the stock reaction of AA clients. The most of research confirms the negative stock reaction of AA clients after Enron shredding document. Our data was composed by the listed companies on the SBF 250. Banks and other financial institution are excluded.
The empirical results show that the stock market reaction of the enterprises audited by two auditors' type 1BIG and 1Non BIG or 2 Non-Big doses not reacted negatively after the announcement of Enron shredding documents. Whereas, the stock market reaction of the two Big Four clients presents a significant negative reaction after the destruction of Enron files. This result demonstrates that the choice of one auditor from non-big audit firm is an important strategy for the enterprises when financial scandal revealed. In conclusion, the joint audit with one at least Non-Big is encouraged by the European Green Paper. This mechanism reduce the audit market concentration and contribute to the knowledge transfer between auditors Big and Non-Big.