Affes H and Sardouk H*
Faculty of Economics and Management of Sfax, University of Sfax, Tunisia
Received April 21, 2016; Accepted May 05, 2016; Published May 15, 2016
Citation: Affes H, Sardouk H (2016) Accounting Conservatism and Corporate Performance: The Moderating Effect of the Board of Directors. J Bus Fin Aff 5:187. doi:10.4172/2167-0234.1000187
Copyright: ©2016 Affes H, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Journal of Business & Financial Affairs
The purpose of this article is to clarify the moderating effect of the board of directors in the relationship between the accounting conservatism and companies' performance. Previous studies have investigated the impact of corporate governance on the level of accounting conservatism. Our objective is to study the effect on the relationship between the level of accounting conservatism and performance of a central governance mechanism, namely, the board of directors. Using a sample of 60 French listed companies belonging to the SBF 120 index in all the period 2007-2012, we examine the effect of the size, the composition and the structure of such a mechanism on the relationship between the accounting conservatism and companies' performance. Therefore, most of our findings support the moderating impact of such a mechanism that has been ignored or left out by earlier studies.
Board of directors; Moderating effect; Accounting conservatism; Performance
We are currently witnessing a set of recommendations about governance and the quality of the accounting information disclosed by large companies. Such recommendations can detect the defects in order to deal with and control them . In this context, accounting conservatism is a tool which is often used to assess the quality of the accounting standards. According to Cadbury and Mallin, the governance system needs a good quality of information to eliminate or reduce the information asymmetry between the company’s executives and their stakeholders [2,3]. The governance systems can still counter balance the power and affect the relationship between the quality of the accounting information and performance. Its role is essential in terms of actors and tools. According to Gray, the accounting conservatism is a prudent accounting measure in the face of uncertainty of future events. This practice varies according to the European accounting or Anglo-Saxon systems . Bebchuck and Roe identified two sources of dependency on corporate governance . The first is guided by the structures that existed before, and the second is guided by the legal rules governing the relationships between the investors, the shareholders and the leaders. Deffains and Guigou tried to explain the persistent differences in the governance structures, especially, between the US and Europe and proposed different ways to be followed by the countries which are founded upon different legal systems .
According to Monks, corporate governance refers to the means by which the dominant decision makers (typically managers) are controlled by other stakeholders. The governance structure specifies the allocation of rights and responsibilities among the various participants in the corporation, such as the board, the managers, the shareholders and other stakeholders and clearly explains the rules and procedures required for decision making. It also describes the targets put in place and the necessary means to achieve these objectives and monitor performance (OECD 1999).
The empirical studies, which focused on the relationship between the governance structure and the information dissemination, concluded that there are links between some mechanisms of governance and voluntary information disclosure. Through these mechanisms, the information is involved in the alignment of the executive and stakeholders’ interests. For Jacquillat, by means of the internal audit, the board ensures the control over the financial reporting . On the other hand, John and Senbet and Fama argue that the characteristics of the board can have an effect on the quality of financial reporting [8,9]. Actually, the board is the main source of annual and quarterly information for the shareholders . Its role main function is to actively supervise the management team and redirect it when necessary [11,12]. It comes directly from the shareholders’ votes and can prompt the leaders to provide high quality information. According to Mizruchi, being a member of several boards provides administrators with an important source of information on practices and corporate policies . It is a determinant positively associated with the directors’ perception of their ability to contribute to board’s debates .
Since the board is a supervisory body the effectiveness of which is essential, it would be interesting to analyze its impact on the relationship between the quality of the accounting information represented by the level of the accounting conservatism and the company’s performance. In this sense, we seek to answer the following question: What is the contribution of the Board of directors to the relationship between accounting conservatism and corporate performance? For this purpose, it would be necessary to consider the moderating effect of this mechanism. Embedding this effect within our analysis helps improve the existing literature and complete the analyses dealing with this issue.
The remainder of the paper is structured as follows: we begin with a brief survey of the empirical literature. The description of the sample data and variables as well as the presentation and discusses of our empirical findings will be figured in a next part.
As an internal mechanism of the governance system, the board of directors is viewed as a means of internal control that ensures the conflict resolution, the procurement and the allocation of resources as well as the determination of the strategic choices. The theoretical analysis of the board of directors, distinguishes, in particular, between the contractual (financial and partnership) and the strategic theories .
For the former strategies, the function of the board of directors is to discipline the leaders, whereas for the latter ones, it is rather a cognitive instrument which helps building skills. From a disciplinary perspective, the board’s target is to protect the shareholders’ interests [9,16,17]. It has the power to engage, assess, pay, and lay off the leaders, as well as ratify and control the strategic decisions. The intervention of the board of directors consists of the executives’ incentive to be efficient either through remuneration systems (bonus, stock options...), or by threatening them with revocation. After non-quality financial communication, the board of directors can punish or consider a revocation that aims to penalize the incompetent or opportunist leaders .
In a spirit of a strategic perspective, the administrators’ aim is rather to increase the shareholders’ wealth by improving the company’s competitive position. Its role is manifested by the involvement and commitment of the directors in the definition, selection and implementation of the strategies of the firm . The obligation of the Council generally consists in the review and monitoring of the strategy of the undertaking .
Goodstein et al. defined the strategic role of the board as the major decision maker of the policy changes that will help the organization adapt to the environment significant changes . The administrators are themselves elected by the general assembly . They also play an ambivalent role between the shareholders’ sovereignty expressed through the general assembly and the company’s general direction on a daily basis . The board of directors forces the leaders to unveil their strategies and limit their informational power when the information is communicated to the shareholders.
According to the major theoretical perspectives, it is clear that the quality of a board of directors is designed on the basis of its important role in the value creation. Its effect on the information quality and its corresponding empirical justifications can be summarized in the following Table 1.
However, the board of directors’ effectiveness depends on its characteristics. According to John and Senbet and Fama, the size, structure and the combination of functions are the characteristics of the board of directors which may impact the information quality [8,9]. Several studies focused on the ability of the size of the board of directors to control the leader’s discretion [23-25]. According to Jensen, adding an additional administrator tends to raise the control capacity of the board however its role is moderated by the marginal cost in terms of communication and decision making. Jensen proposed a moderate board size composed of 7 to 8 members . The efficiency of a large board can still be limited by the communication and coordination difficulties of the board’s members, which gives the leader a margin of extra freedom. For Lipton and Lorsh, the size of the board weighed down the communication process and makes it more difficult to decision making . According to Zéghal et al. in a large-sized board, members often face coordination problems and a collusion risk between them . Godard and Schatt found that largesized boards are ineffective and may be fragmented after the emergence of coalitions and group conflicts and show difficulties regarding any consensus on the important decisions and generally characterized by the domination of the leaders, which reduces the protection of the interests of the shareholders . Some other authors supported the idea that a small board may be more efficient [29,30]. Yermack and Eisenberg et al found that the Board’s size is negatively correlated with the company’s performance [24,31]. This can be explained by the coordination problems and the collusion risk which can occur between the Board’s members, which complies with most of the studies [27,29,30]. On the basis of the group cohesion principles, other authors claimed that a small board can be more efficient [29,30]. Moreover, we should study the leadership structure so as to detect its impact on the information quality.
Klein showed that the structure of the board has a considerable impact on the disclosure of credible and relevant financial information . According to Peasnel, Pope and Yeung and Weibach, the Board of Directors is a key mechanism to control the leadership opportunism [33,34]. It should include internal administrators and independent external directors, who can help with the decision ratification since they know the firm very well, and independent outside directors who can ensure the leaders’ control. Tifafi and Dufour stated that the major role of the independent directors is to control the leaders so as to ensure the performance and safeguard the shareholders’ interests . In this context, the Board's effectiveness will be evaluated according to the degree of its independence from the leaders .
Moreover, Fama and Jensen claimed that a Board of Directors composed mainly of independent directors is deemed to be more inclined to monitor the executives . Other authors found that the presence of independent outside directors on the Board reduced the agency conflicts and the opportunistic behavior proven by the leaders [37,38]. Rosentein and Wyatt argue that the presence of independent external directors positively affects the company’s performance . On the other hand, Weinbach, think that their presence can enhance the replacement of an ineffective leader . For Charreaux, Daily and Dalton and Mizruchi and Stearns, the objective behind appointment the outside directors is to have a better control of the leader and provide the Board of Directors with skills and objective judgment, something which can only be ensured by the internal directors who are too much involved in management [40-42]. In this context, the studies of Haniffa and Cooke performed on some Malaysian listed companies stated that the presence of a positive relationship between the proportion of the independent directors and the quality of the annual report disclosure . In short, internal administrators, too much involved in the company’s internal management, do not have enough power to oppose the leaders’ decisions. They can still be subject to their pressures to comply with their interests. Such pressure can be stronger when these administrators have some seniority within the company. In sum, the external directors are an effective tool to reduce the agency conflicts against the internal ones.
|Proposal||Hypotheses||Agency theory||Empirical justification|
|Board of directors||Size||Small||Abbot, Parker et Peters (2002) et Davidson et Dahalt (2002), Bedard, Coutreau et Chtourou (2001), Coulton, James et Taylor (2001), Peasnell, Pope et Young (1998), Beasley, (1996);Dechow, Sloan et Sweeney (1996); Lipton et Lorsh (1992).|
|Board of directors||Board of directors’ independence||Large||Cormier (2007), Mezghani et Ellouze (2007), Cheng et Courtenay andKrishnamurti, (2005), Chtourou, Bédard et Courteau (2001), Chen et Jaggi (2000), Haniffa et Cooke (2000), Beasly (1996), Malone, Fries et Jones (1993), Williamson (1983), Leftiwich, Watts et Zimmerman (1981).|
|Board of directors||Board of directors’ independence||Small||Eng et Mak (2003), Godard et Schatt (2000), Forker (1992).|
|Board of directors||Board of directors’ independence||Neutral||Coulton, James et Taylor (2001), Wright (1996).|
|Board of directors||Combination of functions||Large||Labelle et Schatt (2005), Gul et Leung (2004), Wong (2001), Forker (1992).|
|Board of directors||Combination of functions||Small||Godard et Schatt (2004), Godard et Schatt (2003).|
|Board of directors||Combination of functions||Neutral||Coulton, James et Taylor (2001) et Hanifa et Cooke (2000).|
Table 1: Impact of the board of directors on the information quality.
We must also study the cumulative impact of corporate executive positions and board of directors on the quality of information. In the literature on the separation of positions between the chief executive officer and the board’s chairman, there are diverging opinions. In fact, some denounce this separation while others support it. According to the agency theory, to improve the monitoring level, the control and decision functions must be separated. Jensen also supports this idea and states that the dual function of the Board’s leader and president helps strengthen the interest conflicts . Moreover, for Jensen and Meckling and Jensen, the function duality can limit the effectiveness of the governance mechanisms and can also be a source of conflicts of interest between managers and shareholders [16,23]. Godard and Schatt found that the combination of functions allows managers to more easily defend the projects they have initiated, even if they do not create value for shareholders . The separation of powers is therefore necessary to reduce the agency costs. For some researchers, such as Cannella and Lubatkin; Sridharan and Marsinko; the duality of functions is recommended since it helps avoid the miscommunication and the contradictions between the expectations and the actions and offers more flexibility to seize new opportunities [44,45]. By examining the accounting measures, Rechner and Dalton found that firms with separate management structure outperform those with combined management structure .
Based on the arguments cited above, we will assume that the Board of Directors and its characteristics have a moderator effect on the relationship between accounting conservatism and the enterprise’s performance. Hence, the foregoing discussion suggests the following hypothesis:
H: Board of directors has a moderator effect on the relationship between accounting conservatism and the enterprise’s performance.
Our study covers the period 2007-2012 and focuses on 60 French companies belonging to the SBF 120 index. Our choice of the study period is explained by the evolution of accounting conservatism level following the adoption of standards IFRS in 2005, particularly in the French context. Thus, our study focuses on a sample of 60 French firms observed over a period of 6 years. This will lead us to estimate panel data regression models that take into account the individual effects of companies and periods studied. In panel data, the model that takes into account the presence of heteroskedasticity and autocorrelation is the model Feasible Generalized Least Squares (FGLS) which explains our choice about using this search method. The bases "Thomson Financial", "Diane", "Worldscope" and "Dafsaliens" are the sources of our accounting data. Governance data were manually extracted from companies’ annual reports available on the internet. We have eliminated the companies for which data was missing, those belonging to sectors presenting a particular functioning such as banks and those with accounting practices requiring a specific treatment. The final sample consists of 360 firm- years’ observations.
Research model and measurement
To study the moderating effect of the Board of directors in the relationship between accounting conservatism and enterprise’s performances, the following model will be tested:
The indices i and t correspond, respectively to the company and the year (2007-2012).
PERF: enterprise’s performance; C-Score: Accounting Conservatism; X: variables linked to the characteristics of the Board of directors.
Enterprise’s performance: PERF: The review of existing empirical studies reveals the dominant place occupied by accounting measures of Performance Company’s [47-50]. Our concern to compare our results with those of the majority of the research works led us to retain accounting nature indicator ROE (Return on Equity).
Accounting conservatism: C-Score: According to Khan and Watts, C-Score reflects the change in accounting conservatism between companies . To estimate the conservatism level, authors fit the model of Basu . The model of Basu adjusted by Khan, Watts, Ahmed and Duellman lead to the calcul of the conservatism level on each company and for each year [51,53].
Board of directors variables: X: According to Lobo and Zhou, companies are on average more conservative in their financial reports after the Sarbanes-Oxley Act . For Bessieux-Ollier C, French and German companies have higher levels of conservatism that the US companies . This can be explained by the characteristics of the Board of directors. In this regard, we can affirm that the agency problems lead to call for governance mechanisms as a regulator of the managers’ opportunistic behavior. For the purpose of a more thorough, study of the characteristics of the Board of directors and their effect on the relationship between enterprises’ performances and accounting conservatism becomes essential (Table 2).
In what follows we will discuss the obtained results about the moderator effect of the Board of directors in the relationship between conservatism and accounting performance of listed companies.
At First, we will report the descriptive statistics belonging to the entirety of variables. We will present descriptive statistics over the period: 2007-2012 such as; Means, Median, Maximum, Minimum, Std. Dev Skewness, Kurtosis and Jarque-Bera (Table 3).
Then, we will present the descriptive statistics belonging to dichotomous variables as following Table 4.
We note that the average size of the board of directors is 12,55. Respectively, independent administrators’ average is 6,380%. Indeed, we notice a small percentage of independent directors in French companies. The average attributable to the combination of the functions corporate executive positions and board of directors is higher than the average of separation of the two functions, respectively, 55,8 and 44,2. Accordingly, there is not a significant disparity in the intensity of the quality of accounting information measured by the degree of accounting conservatism (0,133).
|Measure of performance|
|Performance||PERF||ROE=Net results/shareholders ‘ equity||Xu and Wang (1999), Jiaet al. (2005) et Omran (2009)|
|Accountingconservatism||C-Score||C-SCORE i,t=λ0+λ1 (SIZE i,t)+λ2( (M/B) i,t)+λ3 ( LEV i,t)||Basu adjusted by Khan and Watts , Ahmed andDuellman|
|Ratio market-to-Book||M/B||Market value/book value||Khan and Watts |
|Size||SIZE||value log of the equity market||Khan and Watts |
|Leverageeffect||LEV||Total debts/Equity market value||Khan and Watts |
|Board of directors variables|
|Board of Directors’ size||TCA||The overall number of administrators||Zahra and Pearce (1989); John andSenbet|
|Independent administrators||ADM_IND||The proportion of independent administrators on the Board of Directors||HermalinandWeisbach (2003)|
|Dual function||CUMUL||Binary variable=1 if the CEO is at the same time the Board’s Chairman and 0 if not.||Baliga, and Moyer (1996);Rechnerand Dalton |
|Firm’s size||TAILLE||Debt ratio=total liabilities/total assets||Boubakriet al. (2005); Mak andKusandi (2002);Villalonga (2000)|
|Leverageeffect||ENDET||Binary variable=1 ifthe company is technological and 0 otherwise||Nget al. (2009); Sun et al. (2002)|
|Litigation||Contentieux||Binary variable=1 if the firm is supervised by a large Board and 0 otherwise||Field et al. (2005)|
|Auditing||Vérification||Binary variable 1 if the company belong to CAC 40 and 0 otherwise||Watts (2003)|
Table 2: Summary of variables definitions and measurements
Table 3: Statistics descriptive.
Table 4: Variables dichotomiques.
Before the estimation of any linear regression, it is advisable to ensure that there is no multicollinearity among the explanatory variables. Then, some tests, such as the heteroskedasticity and the error auto-correlation tests should be conducted to reach robust estimates. To check for this problem at the level of our sample, we will calculate Pearson correlation coefficients between these variables as well as the "Variance Inflation Factor" (Table 5).
The examination of the correlation matrix shows that all the correlation coefficients are below the 0.9 limit from which usually serious multicollinearity problems start to rise. In addition, according to this table, it can be noticed that all our explanatory variables have a value of the "variance Inflation Factor" below the 5, limit suggested by Gujarati and Kennedy [56,57]. These results make us conclude that there is no serious multicollinearity problem.
Our study covers a sample of 60 French companies observed over several years, which, by definition, leads to estimate the regression model defined above on panel data. Given the particular nature of the panel data, the order of some econometric steps should be necessarily followed. At first, it should be noted that the fixed effect model was rejected since regression includes invariant dummy variables in the temporal dimension. Therefore, a random model estimation was used and subsequently the 'Breusch-Pagan' test, called the "Lagrange Multiplier test for random effect” was ordered to identify whether there are or not specific effects. In other words, this test helps identify which model to be used: the "pooled" or the random-effect model. The result of this test (Table 5) is a significant Chi-square statistics (Prob>chi2 = 0.000), which makes us confirm the existence of individual effects. Therefore, the random effect model is retained for the estimation of the various regression models. Afterwards, heteroscedasticity was tested by conducting the Breusch-Pagan test. In the context of a heteroscedasticity test, the null hypothesis is homoscedasticity, which will be the case when all the residues in regression coefficients are zero. To carry out this test, we had to regress the squared residuals resulting from the random effect model with the explanatory variables of the various regression models. As a result of this test (see Table 5), there is a significant Fisher's statistics (prob>F=0.000). This makes us reject the null hypothesis and subsequently confirm the presence of a heteroscedasticity problem. In this case, the generalized least squares method (MCG or GLS) should be used, which enables the correction. However, to implement this method, we should first identify the form of heteroscedasticity. For this reason, a modified Wald test was conducted using the STATA program. This test checks whether heteroscedasticity is inter-individual. Under the null hypothesis, the test assumes that the error variance is the same for all the individuals and the statistics follows a Chi-square of N degree of freedom (60 in our case). On the basis of the assessment of the P-value associated with the Chi-square test, the null assumption cannot be accepted. The rejection of this hypothesis does not further specify the heteroscedasticity structure. The previous heteoscedasticity conclusion remains the same without any further specification. Afterwards, to test the error autocorrelation, we proceeded with the intra-individual autocorrelation test of Wooldrigde. The results of this test (see Table 5) confirm the presence of a serial autocorrelation (Prob>F is lower than 0.05). To sum up, we can say that there are heteroskedasticity and autocorrelation problems. In panel data, the model that takes into account the presence of heteroskedasticity and autocorrelation is that of the model Feasible Generalized Least Squares (FGLS) (Table 6).
Table 5: Correlation Matrix and VIF test`
In what follows, we will interpret the FGLS estimation results of the different regression models which test the moderating effect of the Board of Directors’ characteristics on the relationship between the accounting information quality measured by the level of the accounting conservatism and the company’s performance. We will try to check if the impact of the accounting conservatism on the company's performance is moderated by the characteristics related to the Board of directors. In other words, this point is to check if this impact varies with respect to the size of the Board of Directors, to the directors’ independence and to the function combination of the Chief Executive Officer and the Chairman of the Board of Directors.
The following table presents the FGLS estimates of the various regression models including the various variables related to the Board of Directors. From models 2 to 4, we test the interaction effect of each variable taken alone with the level of the accounting conservatism on the company’s performance. Model 1 includes simultaneously all the variables related to the Board of Directors and their interactions with the conservatism level.
Concerning the moderating effect of the Board of Directors’ size, it is clear that the effect of the C-score * TCA interaction is negative and significant at 10 level % for model 1 and at 1% for model 2. Such a result suggests that Board of Directors’ size has a negative effect on the moderation of the relationship between the level of the accounting conservatism and the company’s performance. This thesis joins John, K. and Senbet, L and Lipton and Lorsch in their study [8,26]. Authors support the negative effect of the Board of Directors’ size. According to John K and Senbet L the size of the board is negatively associated with its ability to work effectively. For Lipton and Lorsch the size of the board weighed down and hinders any communication process [8,26].
Models 1 and 3 of the following table indicates that the C-score * ADM_IND interaction term is positive and weakly significant (at 10% level). Such a result suggests that Board’s independence has a positive impact on the relationship between conservatism and the company’s performance. Therefore, the fact that the Board of Directors is predominated by independent administrators leads to the improvement of performance. This is motivated by the fact that these independent administrators being free of any managerial influence are considered the best ones to control the leaders and limit their discretionary space. These administrators are essentially guided by the protection of the shareholders’ interests and the maximization of the company’s value. They actively participate in the decision-making process, the control and ratification of decisions taken by the leaders, in such a way that only the value-creator projects for the shareholders should be carried out [17,58].
The study of the effect of the combination of functions of the Chief Executive Officer and the Chairman of the Board of Directors on the relationship between thee accounting conservatism and the company's performance shows that this impact is negative and significant (at a level of 1%) according to the coefficient of the C-Score * CUMUL interaction term. This result suggests that the impact of the accounting conservatism on the company’s performance is less favorable in companies where the function of the CEO and of the Chairman of the Board are performed by the same person compared to the companies where there is separation of both functions. Such a conclusion is also supported by the previous literature review [28,59,60]. This is in line with the input supported by the Agency theory which considers the combination of functions as a potential source of conflicts of interest.
We can finally point out that our achieved results appear to be consistent with the studies conducted by John K and Senbet L, Lipton and Lorsch, Fama and Jensen, Zahra, Godard and Schatt, and Boyd which have demonstrates the impact of Board of Directors on the performance and accounting conservatism within companies [8,17,26,28,58-60].
|Cross-sectional time-series FGLS regression
Coefficients: PanelFeasible Generalized Least SquaresPanels: Heteroskedastic with cross-sectional correlation
Number of observations=360Number of groups=60Time periods=6
|Independent variables||Modèle 1||Modèle 2||Modèle 3||Modèle 4|
|Breusch-Pagan Lagrangian Multiplier Test for Random Effects|
|Breusch-Pagan Test for Heteroskedasticity|
|Wald Test for Heteroskedasticity|
|Wooldridge Test for Autocorrelation|
|*Significant at 10% level** Significant at 5% level*** Significant at 1% level|
Table 6: The results of FGLS estimates for the analysis of the moderating effects of the Board of Directors on the relationship between the accounting information quality and the company’s performance.
In this paper, we tried to present the moderating effect of the Board of Directors on the relationship between the accounting conservatism and the company’s performance. The objective of this work is twin. We have apprehended the effect of accounting conservatism on the performance of the company and identify the moderating effect of the board of directors on the relationship between the accounting conservatism and the performance of the company. The practical implication of our study consists in studying the effect of the interaction of the Board of Directors and the accounting conservatism on the company's performance. In this sense our moderating variable namely the board of Directors modulates the direction and the force of the effect of accounting conservatism on the company's performance. Identify the moderating effect of the board of Directors is of a great importance since moderation increases the significance of the relationship between accounting conservatism and company performance. Yet the ignorance of such effect may lead to the conclusion about the likely impact of the low level of accounting conservatism on the company's performance. An empirical study was conducted to test the validity of the assumptions which emerged from the previous literature to confirm or overturn the reached conclusions. Our results are in line with those of some previous studies. In fact, they enable to decide on the importance of the Board of Directors as a central mechanism of governance in the moderation of the relationship between the accounting conservatism and the company’s performance. As a result, despite the found results, it seems that the recourse to the independent directors seems more effective in moderating the relationship between the accounting conservatism and the company’s performance. Moreover, it would be more efficient to separate the CEO and the Board Chairman’s to better moderate the relationship between the accounting conservatism and the company’s performance. Similarly, small tips would be more effective in moderating the relationship between the accounting conservatism and the company’s performance.
Make the best use of Scientific Research and information from our 700 + peer reviewed, Open Access Journals