The Shareholders Confidence and Effectiveness of the Joint Auditors: Empirical Validation in the French Context

The goal of our research work is to show the effect of the presence of joint auditors (BIG4) and a few mechanisms of governance on the shareholders’ confidence. We are trying to explore the relationship between its different variables in the French context during the period 2005-2010. The empirical results show a positive and significant relationship between, on the one hand, the presence of two auditors BIG4, the opinion of the auditors, the independence of the members of the audit committees and the boards of directors and the asset efficiency and the confidence of the shareholders of the counterpart. On the other hand, the audit fees, the collaboration period between the associate signatories, the size and the debts of the companies seem to have a negative and significant effect on the shareholders’ confidence.


Introduction
As an economic and social entity, the company is a place of intervention and meeting of several actors who are interested in its performance.The debates studying corporate governance state that the interests of its various stakeholders can, on the whole, be convergent (Jensen and Meckling, 1976;Fama, 1980).Indeed, the accounting scandals characterizing the business world today is a consequence of this divergence of interest.
However, to face the completely fraudulent behaviors of certain leaders (see for example, the Enron business, at the lead, Adelphia, and especially WorldCom), the American legislator adopted the law known as "Sarbanes-Oxley" which has been the most important reform in the United States since the crisis of the 1930's.Within this framework, several studies confirm that the "Sarbanes-Oxley" act can produce economic consequences by examining the reactions of the market to the relative legislative events.This law will have consequences on the American companies as well as on the companies in other contexts (Zhang, 2007;Vakkur et al., 2010).Thus, the law of financial security, which was diffused in France, is a consequence of the adoption of the law "Sarbanes-Oxley" in the United States.
In this regard, we can note that the effectiveness of the auditor's work constitutes a condition of reliability of the financial statements published by companies (Alles and al., 2004).It is to restore the stakeholders' confidence and, in particular, that of the shareholders in the credibility of the published financial information that the French legislator has to establish the law of financial security.This paper constitutes a part of a study conducted within the framework of a thesis in progress, on the audit quality in the presence of the co-statutory auditor.We try to present the shareholders' perception of the financial statements and the effectiveness of the co-statutory auditors.In short, our study will be organized as follows: The second section attempts to present a synthesis of the literature and the development of our assumptions.In the third section, we will present the methodology of the adopted research.The fourth section will present the results and the interpretation of the econometric analysis.Finally, the fifth section will be devoted to our conclusions.

Concept of Confidence: Definition
The idea of studying the effectiveness of the co-statutory auditors with its sociological dimensions is a challenge for most researchers.Perez (2003), states that this discipline is useful to understand corporate governance which "implies at the same time institutions, relations, rules and behaviours".This is why, in this article, we try to define the notion of confidence in order to apprehend the effectiveness of the co-auditor.
In fact, this concept is at the heart of concerns.C. Mothe (1999) suggested that: "Confidence rarely appears as a dependent variable, but mainly as a variable to explain either the performance of firms or their governance structure.Today, there is no unique measure of confidence, or a case study specifically dedicated to trust.The measures used by the authors are mainly drawn from the psychosocial literature with a strong predominance of the interpersonal aspect".In addition, such a definition is difficult to adopt.Literature offers us several definitions of trust either under the socio-philosophical angle, or, the managerial side (McAllister, 1995;Zucker, 1986;Barney and Hansen, 1994;Hosmer, 1995).Actually, in this research, we base ourselves on (McAllister 1995) who states that the recent development in sciences of the organizations shows the importance of interpersonal confidence in the organizational effectiveness.Similarly, the definition suggested by (Bidault, 1998, p. 34) seems interesting as it makes it possible to integrate most of the characteristics of confidence: "presumption that, in a situation of uncertainty, the other part, even in a case of unforeseen circumstances, will act according to the rules of behavior which we find acceptable".We will focus, in this article, on the concept of the shareholders' confidence against the effectiveness of the joint auditors.We are trying to explain this relationship by studying the effect of the auditors' opinion published in their auditing report and the situation of the stock prices in maintaining and restoring the shareholders' confidence.We try to explain this relationship by studying the effect of the auditors' opinion published in their auditing report and the situation of the stock market share prices in maintaining and restoring the shareholders' confidence.

The Efficiency of the Joint Auditors
The agency theory, which results from a study by Jensen and Mekling (1976), lies within the scope of the contractual theory of companies.This theory considers the company as an entity playing the part of a "contracting center" and taking the shape of a node of contractor whose center is occupied by the leader.Thus, the organization in relation to this theory, which relies on methodological individualism and the assumption of rationality of the individuals, is analyzed starting from the behavior and the interactions between the individuals.In addition to the agency problems, there is the information asymmetry between the principal and the agent.Indeed, the parts do not hold the same information on the evaluation of the firm's, performance which could generate conflicts of interest.
This situation of superiority of information maintaining is often due to various accounting discrepancies.In this regard, Stolowy and Breton (2003, p. 3) argue that: "accounts manipulation is defined as the use of management discretion to make accounting choices or design transactions so as to affect the possibilities of wealth transfer between the company and society (political costs), funds providers (cost of capital) or managers (compensation plans)".In this kind of discrepancy, the direction acts artificially on the result in order to carry out a pre-established level of "expected" benefit.According to Healy and Whalen (1999), the management of result intervenes when the managers use their discretionary latitude in the process of financial accounting and the structuring of transactions which modify the financial statements, which can induce certain stakeholders in error on the real economic performances of the company or influence the contractual challenges which rely on the accounting figures.
The various accounting scandals throughout the world enable us to note that a crisis of confidence is striking the business world today.So Enron, WorldCom, Parmalat and Vivendi are only some examples to approve that the credibility of the leaders is reconsidered and that they are confronted with a challenge of renewal.In this respect, the firm is regarded as a legal system used to draw up a multiplicity of implied and explicit contracts governing the relations between the agents.These contracts are important since they define the rights of the various agents having divergent objectives and who seek the maximization of their own functions of utility.Then, resorting to an external auditor seems necessary to solve the conflicts of interests.According to Shapiro (1987), the auditor is the "guard of confidence" between the parts, which makes it possible to guarantee the transparency of the financial information.Bennecib (2004) indicates that "the effectiveness of the audit remains a current event.The accounting scandals, which marked out the history, have revived debates on the credibility of the financial information published and consequently on the reliability of the opinion stated by the auditor".Indeed, the former studies state that the membership of the auditor to one of the large international cabinets (BIG4) is an indicator of the effectiveness of the latter (Bennecib, 2004).In other words independence and competence measured by the membership of networks (BIG4) allows these cabinets to detect and reveal to the public the anomalies of the published financial information which makes it possible to have a good audit quality (Manita and al., 2008).

Mechanisms of Governance and Shareholders Confidence
The leaders who manage the company for the purpose of serving the interests of their office period have of specific competence and better information on the company and its environment.This information asymmetry enables them to direct the business management according to their own interests (Shleifer and Vishny 1997).Under these conditions, corporate governance can be defined as a set of mechanisms which discipline the leaders.According to the definition of Shleifer and Vishny (1997), corporate governance gathers the means by which the suppliers of company's capital can be assured of the return on their investment.According to this definition, the field of governance is limited to the conflicts of interests between the leaders and the owners (shareholders).Charreaux (1998) proposes a broader definition.Corporate governance covers the whole organizational mechanisms for the purpose of determining the powers and influencing the decisions of the leaders, those who control their behaviour and define their discretionary space.In sum, "the establishment of relations of trust with the various partners would have a positive effect on the creation of value in reducing the costs (agency costs, influence costs, rooting costs …) and widening the discretionary space of the leaders, the thing which would strngthen the implementation of a more profitable investment policy" (Charreaux 1998).In the same context, Zingales (2000) expects that, in order to avoid the risk of the power fragmentation which can affect the sustainability of the organizational capital, the design of a system of governance could protect the interest of each stakeholder.
Literature offers us various types of control mechanisms.According to Jensen (1993), the mechanisms of corporate governance could be classified into two broad categories: the internal mechanisms (the concentration of ownership, debt, the board of directors,…) and the external mechanisms which gather the main markets representing the environment of the company (the job market, that of goods and services and that of capital) and the specific legal system to each country (financial system, legal system…).These various mechanisms evolve together in a system to affect corporate governance.As it is stated by the various research tasks, we can deduce the following assumptions according to which: H1: The presence of two auditors BIG4 has a positive and significant impact on the shareholders' confidence.

H2:
The auditors' opinion would positively and significantly affect the shareholders' confidence.

H3:
The audit fees have a positive and significant impact on the shareholders' confidence.

H4:
The independence of the members of the audit committees and of the boards of directors has a positive and significant effect on the shareholders' confidence.

H5: The Company debt would negatively and significantly affect the shareholders' confidence.
H6: There is a positive and significant relationship between the return on assets and the shareholders' confidence.H7: The company size has a negative and significant impact on the effectiveness of the audit job and therefore on the shareholders' confidence.Our original sample consists of all the French listed companies of the SBF 250 up to December 31st, 2010 in the Paris Stock Exchange during the period 2005-2010.The accounting and financial variables in our econometric model variables were extracted from the database Data Stream and annual reports published by the companies on the site: http://www.amf-france.org.From this data panel, the following selections and modifications were carried out: the financial institutions, off-finance government and the non-classifiable institutions (banks, insurance companies…) are excluded from the analysis sample.All the observations, for which one of the variables is missing, are also excluded from the panel.These various changes made us end up with a sample of 145 French listed companies, that is to say a total of 870 observations of company-years.

Research Model and Measurement of the Variables
Following the example of some researches associating the co-statutory auditor, corporate governance and confidence (Charreaux, 1998;Bennecib, 2004…), we propose to study the following model:

Dependent Variables
CONFIDENCE it : A binary variable measuring the degree of the shareholders confidence against the effectiveness of the joint auditors.In other words, it is equal to1 if the average daily closing market rates during the week following the publication of the auditors' report increases compared to the daily closing rates in the market during the week which precedes the publication by the auditors' report and 0 otherwise.-1, if the average of a week after the publication of the audit report> the average of one week before the publication of the audit report.-0, if not.

Independent Variables
AUDIFIRM it : coded 1 if the company is audited by a large international audit firm (BIG4) and 0 if not.
AUDIFI_A it : just as AUDIFIRM, but it concerns the second auditor.
OPINION it : equals 1 if the two auditors certify, without any reserves or remarks, on the financial statements of the company, and 0 if not.
ΔHONOR_AUD it : variation of the audit fees between (t) and (t-1) divided by the total fees (t-1); IND_COMITE it : it is the number ratio of the independent members compared to the overall number; CAD_IND it : This proxy detects the independence of the board of directors and measured by the number ratio of the independent administrators compared to the overall number; SIZE it : it is the Napierian logarithm of the total assets of firm i at time t; LEV it : measured by the ratio: debts (t-1)/total assets (t-1); ΔROA it : variation of the ROA ratio between t and t-1;

Variables of Control
DUREE_COL it : duration of collaboration between the associate signatories and the company (on a monthly basis); REMUN it : Variation of the leaders' remuneration, calculated by the variation of the share variable between (t) and (t-1) divided by the overall remuneration paid on a yearly basis (t-1); ΔSALES it : It is the variation of the turnover between (t) and (t-1) divided by the total assets (t-1); AGE it : Napierian logarithm of the number of years of the company being listed.Hausman's test enables to determine if the coefficients of the two estimators (fixed and random) are statistically different.Indeed, the probability of this test is higher than the threshold of 10% (Prob>chi2 = 0.7067) which implies that the random effects model is preferred to the fixed-effects one.For each variable, the number of observations is of 870.Several results are debatable.Initially, the results of the descriptive statistics lead us to make the following observations.The statistics of our dependent variable (CONFIDENCE) as well as the explanatory variables (AUDIFIRM, AUDIFI_A and OPINION) tend to point out to the effect of the law enforcement on financial security in the French companies since 2003.This could explain the positive result of the average (0.66; 0.58; 0.57 and 0.57) over the studied period.Thus, we noticed that almost 66% of the shareholders belonging to our study sample have confidence in the presence of two auditors BIG4 in the control and the checking of the company's financial statements.Moreover, we noted that the French market is dominated by large international audit cabinets since, on average, 58% of the firms of our sample are audited by Big4.In the same way, more than 57% of the auditors certify, without any reserves or remarks, on the regularity and the sincerity of the financial statements and give a favorable opinion to it.It is also interesting to note that in France, the Button report ( 2002) insists on the importance of the independence criterion.Thus, the percentage of the independent members of the board of directors is, on average, 53%, whereas 48% of the members of the audit committees are regarded as independent (Geraldine, 2009).Besides, the observation of the table shows that, on average, 60% of the companies forming our sample are involved in debt.To avoid the problem of multi-collinearity, in other words, checking the assumption of the independence of the explanatory variables, we have used the correlation matrix.In fact, table 4 presents Pearson's matrix of correlation in our sample.The results presented in this table indicate that the explanatory variables (AUDIFIRM AUDIFI_A and OPINION), as well as most of the control variables, such as (ΔHONOR_AUD, DUREE_COL, IND_COMITE, CA_DIND, REMUN, SIZE, ΔSALES, ΔROA and AGE, are positively correlated to our dependent variable (CONFIDENCE).The explanatory variables and the control variables, on their part, seem to be globally and slightly correlated, which means that the obtained results are not skewed.

Econometric Estimate of the Model and Interpretations of the Results
Table 5. Econometric estimation of the model of shareholders confidence In this paragraph, we confront our model described before with the data panel in order to empirically test the impact of the co-statutory auditor on the shareholders' confidence.Table 5 presents the result of econometric estimation of the model of shareholders confidence.It is advisable to specify, first of all, that the model, as a whole, is statistically significant (Wald chi2 (13) = 98.65;Prob > chi2 = 0.0000).In other words, the independent variables contribute to the explanation of the shareholders' confidence (CONFIDENCE) at a rate of 98.65%.This seems to indicate that the co-statutory auditor and the mechanisms of governance can be regarded as a guarantor of the shareholders' confidence.
In our analysis, we have considered the level of internationalization of the audit cabinets (membership of the network of large cabinets of audit BIG4, or not).The results reveal that AUDIFIRM and AUDIFI_A variables seem to be positively significant at the risk threshold of 1%.This implies that firms seek to convince the market of the reliability of their revealed information.Actually, as it is difficult or even impossible for the public to observe the opportunist behavior of the leaders of the companies, we suggest that the presence of two auditors BIG4 can help solve the problems and the specific dysfunctions within a given company, besides, and by taking account of its power of detection and explanation, it can solve the problem the company is facing.Therefore, using audit firms BIG4 helps restore and maintain the shareholders' confidence against the effectiveness of audit job.
As to the variables OPINION and ΔHONOR_AUD, they are respectively positive and significant of the order of 1% and 5% (P>|Z|=0,000; P>|Z|=0,021).Indeed, the opinion of the auditor is the concrete image of his independence and competence.Moreover, this variable can be interpreted by the fact that the leader of the company cannot exert pressure on both auditors at the same time.This allows either of them at least, to give an opinion reflecting the company's real economic and financial situation (Sundgren, 2009).Similarly, literature makes us confirm that the independence of the auditor is influenced by the fees he receives and consequently by the quality of his audit work (Malone and Robert, 1996).However, the shareholders receive audit fees as a tool in controlling and limiting the opportunistic behavior of the leaders.Thus, the opinion and the audit fees are used mainly as indicators of the audit quality affecting the shareholders' confidence.In addition, the period of collaboration between the signatory partners (DUREE_COL) is negatively significant (-0.0179) with the order of 10% (P>|Z|=0.058).Carey and Simnett (2006) suggest that, during the first two years, the audit quality can be low because the company does not know its environment.During the second phase, (beginning of the 3rd year until the end of the 5th year), the audit quality becomes better due to the acquired expertise.The third phase is characterized by the absence of the auditor's critical view which weakens his audit quality.In this way, we can say that there is a negative effect of the period of collaboration between the partners on the shareholders' confidence.Within this framework, the change of one of the signatory partners, before collaborating together for two years at least, negatively affects the job and the efficiency of the audit task, which explains the shareholders' mistrust.
Variables IND_COMITE and CA_DIND are positive (3.6022 and 6.6915) and significant with the order of 10% and 1% (P>|Z|=0,055; P>|Z|=0,008).Indeed, the board of directors and the audit firm represent control mechanisms of the leaders' actions which help check that the decisions taken match the shareholders' interests.
Hence, the independence of the members of the audit firm and of the board of directors can be regarded as mediating mechanisms which help with the resolution of conflicts between the external auditors and the management (Stewart and Munro, 2007).Moreover, they are two factors that affect the ability of the external auditor to resist the pressures of the audited.This helps the restoration and the maintenance of the shareholders' confidence against the co-statutory auditor's effectiveness.
The result of the other control variables shows a negative effect (-0.2115 and -2.2957) and significant of the size (SIZE) and debt (LEV) of the company at a rate of 10% and 5% (P>|Z|=0,067; P>|Z|=0,022) on the confidence of the shareholders.As to the characteristics of the company, we notice that the larger a company is, the more the auditor finds difficulties in carrying out his job of checking.In addition, the results suggest that the more one company is indebted, the less societal information it will disseminate in its annual report, which explains the shareholders' mistrust.However, the shareholders seem to be interested in the returns on assets.Hence, variable (ΔROA) has a positive (8.7607) and significant impact with the threshold of risk of 5% (P>|Z| = 0,024).
Regarding the remaining control variables, executive compensation (REMUN), the variation of sales (ΔSALES), trading status (AGE) include, are not determinants factors of shareholder confidence in the French context.

Conclusion
In this study, we tackle the question of the shareholders' confidence against the effectiveness of the co-statutory auditor in the French context.Thus, motivated by the recent work of (Perez, 2003;C. Mothe, 1991;Shleifer and Vishny, 1997) and recently (Bennecib, 2004), our research contributes to the current debate about the concept of confidence.Moreover, our explanatory model of the shareholders' confidence is based on the effectiveness of the joint auditors and the governance mechanisms using a sample made up of 145 French companies.
From the exploratory phase, emerge several important results.Initially, our results confirm the assumption that the presence of the joint auditors (BIG4) has a positive and significant impact on the shareholders' confidence.
Secondly, the results of our research affirm the significant effect of most of the chosen governance mechanisms on our dependent variable.In this sense, the efficiency of joint auditors ensures the confidence of shareholders.Thus, trust emerges as a mechanism pushing the whole economy.
This study, like any other research, suffers from some deficiencies.Indeed, the non availability of the data, the size and the choice of the companies forming our sample are possible insufficiencies for our research.These limits can reduce the explanatory power of our selected model, which could be the issue of a future research.

Table 1 .
Characteristics of the sample

Table 2 .
Result of Hausman's specification test B = inconsistent under Ha, efficient under Ho; obtained from xtlogit.