Disclosure, Shariah governance and financial performance in Islamic banks

Purpose – This study aims at investigating the impact of the disclosure and the Shariah governance on the financial performance in MENASA (Middle East, North Africa and Southeast Asia) Islamic banks. Design/methodology/approach – We use the Generalized Least Squares (GLS) regression models to check the interdependence relationship between the disclosure, the Shariah governance and the financial performance of 47 Islamic banks (IBs) from ten countries operating in MENASA region. The sample period is from 2012 to 2019. In these regressions models, Return on Assets (ROA) and Return on Equity (ROE) are the dependent variables. The disclosure and the Shariah governance indicators are the independent factors. To measure the Shariah governance, we use the three sub-indices, which are the Board of Directors (BOD), the Audit Committee (AC) and the Shariah Supervisory Board (SSB). Size, Leverage and Age of the bank are used as control variables. We also used The Generalized Method of Moments (GMM) and the three-stage least squares (3SLS) estimations for robustness check. Findings – Resultshowsanegativerelationshipbetweenthedisclosureandthetwoperformancemeasuresin IBs. Furthermore, as far as the governance indicators are concerned,we foundthat the BOD andAC, as wellas the BOD and SSB, have a positive and significant impact on the ROA and ROE, respectively. This reveals that good governance had a significant association with higher performance in MENASA IBs. Originality/value – The paper considers both IBs that adopt mandatory as well as voluntary AAOIFI standardsandtheGLSmethodtoinvestigatetheimpactoftheAAOIFIdisclosureandtheShariahgovernanceonROAandROE.Also,itusestheGMMandthe3SLSestimationsforrobustnesscheck.Itisrelevantforresearchers,policymakersandstakeholdersconcernedwithIBs ’ performance.


Introduction
Islamic banks (IBs) and conventional banks (CBs) live up to the clients' desires for long-haul benefits. However, based on RIBA and risk-sharing practices, they are different (Mushafiq and Sehar, 2021). IBs are the highest proportion of the Islamic financial institutions (IFIs) that extend locally and internationally across both Muslim and Western countries (Sarea and Hanefah, 2013). According to Abdullah et al. (2015), the percentage of IB assets worldwide AJEB 5,3 234 JEL Classification -G21, G34 increased with an annual rate of approximately 17% from 2009 to 2013. Islamic banking is a model of financing which is based on equity, not on debts. Therefore, it is very important for all the Islamic finance corporations, particularly the IBs, to disclose item in their reports and their financial statements to shareholders for reduce agency and clash of interest's issues and to achieve a good bank control. The advantages of disclosure in the banking systems are first correcting any bank misevaluation and secondly increasing the liquidity and institutional interest of the bank (Tabash, 2019). For this reason, disclosure is the one of most significant factors that helps to the enlargement of the IBs. Paino et al. (2011) announced that transparency, disclosure and accountability are the three keys of corporate governance, which ensures the protection of stakeholders and shareholders. According to Mohd Zain et al. (2017), IBs must disclose all truthful information to stakeholders. Due to the specific contracts between IBs and their stakeholders, the disclosure in IFIs can reduce the information asymmetry (Tabash, 2019;Neifar et al., 2020). Thus, the increasing disclosure by banks is prospective to witness a growing in the performance. Therefore, disclosure assists the bank to enhance its value (Tabash, 2019).
In addition to disclosure, the corporate governance is also a very important factor for IBs as it maximizes the value of banks. Indeed, the good governance is a set of rules and regulations that provide transparency and accountability. It also protects the shareholders (Sheikh et al., 2018) and increases trust among stakeholders (Srairi, 2015). Sayari and Marcum (2017) and Chazi et al. (2018) showed that governance leads to the good financial performance because it minimizes the internal risk and enhances the resilience of institutions to external stakeholders. Agency theory proposes that governance refers to reduce in the agency costs and consequently it improves the governance, disclosure and financial performance (Fama and Jensen, 1983).
The effect of disclosure on the financial performance has been studied by different studies and the results are mixed. For instance, Tabash (2019) showed a significant and positive effect between the disclosure and the financial performance in IBs from 2009 to 2013. However, Ellili and Nobanee (2017) found that the disclosure has a significant and negative association on the financial performance of the Islamic banks from 2003 to 2013. For Elgattani and Hussainey (2020), they showed that Accounting and Auditing Organization for Islamic Financial Institutions disclosure does not affect the IBs' performance measured by ROA (Return on assets). Furthermore, recent studies have investigated the link between the corporate governance and the IBs performance (Embi and Shaffi, 2018). Generally, it revealed a positive impact between the Shariah governance and the bank financial performance. For instance, Albarrak and El-Halaby (2019) proved this impact for 120 IBs across 20 countries during three years.
From the above discussion, earlier studies have mixed views on the relationship between the Islamic bank performance, the disclosure level and the Shariah governance. Therefore, the goal of this research is to reevaluate this relationship in the IBs operating in the MENASA (Middle East, North Africa and Southeast Asia) region, where the debate is still limited. Specifically, we study the link between the disclosure, the attributes of corporate governance and the ROA and ROE of these IBs. The sample is from 2012 to 2019. Ten counties namely, Bahrain, Qatar, Sudan, Syria, Oman, Palestine, Yemen, Jordan, Pakistan and Bangladesh are selected according to the data availability. This paper has several contributions to the previous literature. Firstly, it considers both IBs that adopt the mandatory and voluntary AAOIFI standards. Secondly, it is based on all the AAOIFI's governance (standards 1 to standards 7). Thirdly, we use the Generalized Least Squares (GLS) estimation to explore the impact of the AAOIFI disclosure and the Shariah governance on ROA and ROE (Return on Equity). Also, for robustness check, we apply the Generalized Method of Moments (GMM) and the 3SLS estimations.
The end of article is orderly as follows: Section 2 presents the literature review and develops the hypotheses. Section 3 analyses the data and methodology. Section 4 summarizes Shariah governance and financial performance results and discussions. However, section 5 checks the robustness. Besides, section 6 concludes the study.

Literature review and formulation of hypotheses
This study investigates the impact between the disclosure, the Shariah governance and the financial performance of IBs in the MENASA region. Consequently, this part is limited to issues attached to the formulate of hypotheses concerning the link between the disclosure and the performance measures on the one hand (1), the Shariah governance and the financial performance on the second hand (2), and the disclosure, the Shariah corporate governance and ROA and ROE on the third hand (3).

The impact of disclosure on financial performance
The disclosure decreases the agency expenditure by motivating stakeholders to engage at a cumulative and transparent level (Albarrak and El-Halaby, 2019). The increasing disclosure level for business generates better-interested parties' commitment and decreasing profitability.
The link between the AAOIFI disclosure and the performance has been examined by different studies. Consequently, several researchers showed a significantly positive link between these two variables. For example, Tabash (2019) examined the link between the AAOIFI disclosure and the IBs performance in the UAE from 2009 to 2013. Using the Two-Stage Least-Square regression method, they showed that the AAOIFI disclosure has a significant impact on performance. Hence, IBs, with a higher disclosure level, leads to a higher operating performance, reduces the equity cost and increases their financial market values. In addition, Albarrak and El-Halaby (2019) investigated the influence of the Shariah disclosure on the financial performance of 120 IBs across 120 countries during the year 2016. They showed that the disclosure has a positive effect on the performance based on ROA. This result suggests that transparency and responsibility improved the interest parties' trust that plays a role in improving the financial performance. However, other researchers resulted that there is a significantly negative impact among these two variables such as Ellili and Nobanee (2017) who investigated the disclosure impact on the Islamic and conventional banks' performance. They used the banks' annual reports between 2003 and 2013. The panel data regression shows that the disclosure affects negatively the IBs performance. These findings confirm that the corporate disclosure degree has no effect on the performance of the UAE Islamic banks. However, Elgattani and Hussainey (2020) showed an insignificant impact between AAOIFI and the financial performance. These authors studied the influence of the AAOIFI disclosure on the IBs performance in eight countries that adopted mandatory AAOIFI standards for the years 2013-2015 through using the OLS regression analysis. Their result highlights that the AAOIFI disclosure does not affect the IBs performance measured by ROA or ROE. As for the first hypothesis, we propose that the disclosure has a positive association with the IBs financial performance.
H1. There is a positive association between AAOIFI disclosure and the IBs financial performance.

Impact of Shariah governance on financial performance
The good corporate governance leads to a higher performance (Klapper and Love, 2004), and it protects the trust of investors (Zhang, 2012). According to Sarea (2020), more than 100 standards of AAOIFI were issued in 2018, and 7 standards were provided on the Shariah governance.

AJEB 5,3
The effect of the corporate governance on the IBs financial performance shows mixed results. For example, Harisa et al. (2019) reported the impact of the corporate governance on the IBs profitability in Indonesia and Malaysia from 2011 to 2017. Using the panel data regression analysis, they showed that the governance does not affect the profitability measured by ROA. This result was confirmed by Ajili and Bouri (2018), who studied the influence of governance mechanisms on the financial performance of 44 BIs operating in six countries between 2010 and 2014. Based on a multivariate regression assessment, their result shows that high governance does not necessarily maximize shareholder performance.
In this study, we are interested particularly in three mechanisms of the Shariah governance namely, the Board of Directors (BOD), the Audit Committee (AC) and the Shariah Supervisory Board (SSB).
2.2.1 Board of directors (BOD) and performance. BOD protects shareholders and helps managers because it maximizes the banks' profitability (Vu et al., 2018).
Therefore, Nawaz (2019) showed the effect of BOD size on the financial performance in 47 IBs operating in different regions between 2005 and 2010. Using the multivariate analysis, the result indicated a positively significant link between the BOD size and the ROA and ROE. This finding is supported by Darwanto and Chariri (2019) who investigated the good corporate governance impact on the financial performance using panel data for 14 Indonesian IBs from 2014 to 2017. Whereas, using OLS estimation, Naushad and Malik (2015) found a negative and significant effect of BOD on the 24 IBs' financial performance. This result means when the financial performance of IBs is better; BOD size is smaller. This result is the same for Grassa and Matoussi (2014), Mollah et al. (2017) and Aslam and Haron (2020a). Therefore, we propose the second hypothesis: H2. There is a positive association between the BOD and the IBs' financial performance.
2.2.2 Audit committee (AC) and performance. The AC ensure governance for all stakeholders (Velte, 2017) and protect shareholders internally and externally (Kallamu and Saat, 2015;Islam et al., 2020). Therefore, the AC has a positively association with the IBs performance following previous research. For instance, Aslam and Haron (2020) examined how the corporate governance mechanisms affect the financial performance of 129 IBs from 29 countries (Middle East, South Asia and Southeast Asia) from 2008 to 2017. The AC has a positively significant association with financial performance. Thus, the AC large-size improves the IBs performance. This conclusion is contradicted by Ajili and Bouri (2018) for GCC. The third hypothesis is as follows: H3. There is a positive link among the AC and the IBs' financial performance.
2.2.3 Shariah supervisory board (SSB) and performance. The main distinction between the conventional and Islamic banks is the existence of SSB. But, if the SSB of the Islamic bank does not esteems the law, the bank will misses the confidence of its investors and customers, and consequently its financial performance will decrease (Grassa, 2013). Empirically, Mollah and Zaman (2015) checked the effect of SSB on the financial performance for 86 IBs and 86 CBs between 2005 and 2011. Using GLS, GMM and the three-stage least square (3SLS) techniques, they showed that SSB has a significantly positive effect on the financial performance. These findings show that SSB protect the shareholders interest and influences the IBs financial performance. This result has been confirmed by Hassan et al. (2017) for Pakistan, Nomran et al. (2018) for Malaysia. As far as Ajili and Bouri (2018) are concerned, they found no significant relationship between the two variables (SSB and financial Performance). While Aslam and Haron (2020a) showed that SSB is negatively related to intellectual capital efficiency in IBs. Based on these earlier empirical evidences, we present the next hypothesis: Shariah governance and financial performance H4. There is a positive link between the SSB and the IBs' performance.
3. Data and methodology 3.1 Sample selection and data collection This research examines the impact between the disclosure (DCI), the Shariah governance mechanisms (Shariah Supervisory Board (SSB), the Board of Directors (BOD) and the Audit Committee (AC)), and the financial performance of IBs in the MENASA region. The financial performance is the dependent variables as measured by ROA and ROE, while the independent variables are the disclosure and the Shariah governance. Besides, the control variables are the bank size, the bank leverage and the bank age. Our sample concerns 47 Islamic banks operating in 10 countries with 17 banks in Bahrain, 4 banks in Qatar, 2 banks in Jordan, 2 banks in Palestine, 2 banks in Yemen, 2 banks in Oman, 3 banks in Syria, 6 banks in Sudan, 2 banks in Pakistan and 7 banks in Bangladesh. The definitive sample comprises 322 bank-year observations from 2012 to 2019. Data are hand-collected from the English and Arabic versions of the annual reports existing on the official bank websites. Table 1 summarizes and shows all dependent, independent and control variables.

Variables description
3.2.1 Dependent variables: ROA and ROE. The main bank financial performance measurements are Tobin's Q, ROA and ROE (Srairi, 2015). In this research, the IBs are not all listed on the exchange stock. Therefore, the Tobin's Q ratio measure is not applicable. Following Mollah and Zaman (2015), the ROA is calculated by the ratio of the net income and the total assets, and the ROE is measured as the net income on the total shareholders' equity. 3.2.2 Independent variables: disclosure and Shariah governance. Srairi (2015) and Mnif Sellami and Tahari (2017) agreed that the annual reports generally used a disclosure index as a way to evaluate the standard of compliance with AAOIFI. For scoring, each checklist item was coded as disclosed (1), not disclosed (0) or not applicable (NA). The Disclosure Compliance Index (DCI) computes as a number of the required disclosures revealed by the bank to the total number of applicable disclosures.
In this paper, we developed three indices of corporate governance namely BOD, AC and SSB following Ajili and Bouri (2018). The overall CG-index contains 10 CG attributes, which are classified under the Islamic banks' main governance mechanisms. The BODindex is an indirect indicator of board effectiveness, and it contains three keyboard attributes. First, the BOD size who is proxies by the number of BOD members, while Mollah and Zaman (2015) used the logarithm of number of BOD members. Second, the ratio between the number of non-executive members' directors and the total number of directors shows the board's independence (Aslam and Haron, 2020a). Third, referred to Ajili and Bouri (2018) who announced that the most of BOD members would be calculated by the number of non-executives. For the three attributes of AC-index, first, the CG codes required that the board should establish an AC (Ajili and Bouri, 2018). Second, the AC should have a Charaique Audit Department (SAD). Third, we used a dummy variable for showing the attendance of the Big Auditor (Big 4) like the work of Mollah and Zaman (2015). The SSB is among the important GC sub-index in IBs. First, we included the SSB size as a proxy for the number of SSB members (Mollah and Zaman, 2015;Ajili and Bouri, 2018). Second, we used the reputation of SSB members (using a dummy). Third, to determine the cross membership, we used their average number. According to Al-Malkawi et al. (2014), Ajili and Bouri (2018) and Aslam and Haron (2020a), each dichotomous variable takes approximate value of 1 (presence of attributes) or 0 (otherwise). In addition, following these two studies, the firstly was employed to estimate a score for each sub-index. Secondly, we calculated the summation of the three sub-indices to measure the overall CG score for each IB. IBs with a low CG score were considered to get a lower quality of governance and vice versa (Ajili and Bouri, 2018).
3.2.3 Control variables. We consider some control variables (bank size, leverage and age). Therefore, Abdul Rahman and Bukhair (2015) reveal that the size of bank influenced its financial performance. Previous studies such as Al-Malkawi and Pillai (2018) found that the influence of leverage and ROA and ROE is significantly negative. Finally, some studies propose that older banks were more efficient than younger ones because their members are more experienced and more qualified. In accordance with some earlier works, we calculate the bank size (SIZE) as the natural logarithm of the total assets (Ajili and Bouri, 2018;Elgattani and Hussainey, 2020). The bank leverage (LEV) is measured by the total debt ratio to total assets (Abdul Rahman and Bukhair, 2015). Finally, the bank age (AGE) is measured by the number of years since foundation (Ajili and Bouri, 2018).

Empirical models
This study uses panel data of 47 Islamic banks in a period of 8 years from 2012 to 2019. Firstly, we tested the disclosure impact on the ROA and ROE. Secondly, we checked the Shariah governance sub-indices' effect on the financial performance. Finally, we analyzed the disclosure impact likewise the Shariah governance variables on the financial performance. The regression equations were developed to determine these impacts.
Shariah governance and financial performance where ε it is the error term, β 0 is the constant and β 1, β 2 , . . ., β 7 , are the vectors of coefficients estimates. Perf it is the proxy of bank level performance variables of bank i at time t. It is measured by ROA it (Return on Assets: Profit to Total Assets) and ROE it (Return on Equity: Profit to Total Shareholders' Equity). We have chosen these two variables for two reasons. First, ROA is defined as a basic measure of the profitability. It indicates whose bank's assets are being utilized to create gains. Second, ROE is an internal performance measure of the shareholder value. A higher ROE can reflect a higher level of profitability and can also reflect more limited equity capital. BOD it , AC it and SSB it , are the proxy of the governance level variables of bank i at time t. SIZE it , LEV it and AGE it are the control variables of bank i at time t.

Estimation methods
To estimate the earlier equations, we used the GLS random and the fixed effects models. The Hausman test is used to select the best model. The presence of the individual effects leads to verify if it is fixed or random. The conclusions drawn of these tests are listed in Table 2.
For the three models, the p-value of the Hausman test is fewer than 10%. This proves that the zero hypothesis of equal coefficients is accepted. Consequently, we adopted the fixedeffects models. We tested heteroscedasticity by using the Breusch-Pagan test. The p-values of these models are less than 0.05; we dismiss the null hypothesis and deduce that heteroscedasticity is present in the data. We applied the Durbin-Watson test to verify the autocorrelation. In the current study, the Durbin-Watson test reports a value from 0.99 to 1.57, which means that there is a positive autocorrelation in the sample. For this reason, we used the GLS regression method. Recent studies such as Ajili and Bouri (2018) and Mollah and Zaman (2015) used the fixed-effects GLS estimation. Further, for a robustness check of our findings, we used a two-step of GMM and 3SLS methods. Table 3 reports the descriptive statistics for the sample consisting of 47 MENASA region IBs from 2012 to 2019. Here, ROA and ROE are dependent variables. DCI, BOD, AC and SSB are the independent variables. Size, leverage and age are the control variables. The medium values of ROA and ROE are 0.37 and 0.47, respectively. The maximum values of ROA and ROE are 47.32 and 12.32, respectively. The minimum values of these measurements are À0.10 for ROA and À0.54 for ROE. The average worth of the AAOIFI disclosure (DCI) is 77%. This means that 77% of the disclosure is revealed on average per annual reports that revealed that the IBs disclosure level is higher. For the minimum value, it is 0.15 and the maximum is 1. This indicates that several annual reports are disclosing very little information (approximately 1.5%), whereas others disclose 100% of the item. The full information disclosure may be caused that the disclosure of AAOIFI standards is mandatory. The mean worth of the BOD-index, AC-index and SSB-index are 12.95, 5.12 and 7.6, respectively. We conclude that IBs need to convince stakeholders to get better their compliance with SSB and AC. Concerning the control variables; the size mean value is 20.64. The average worth of leverage is 0.51. The mean score of age is 20.30 years with a minimum of 5 years and a maximum of 44 years. The p-values of J-B show that all series are normally distributed. Also, ADF test results show that whole variables are stationary at 1% level. AJEB 5,3

Correlation matrix
The correlation matrix examines the direction of association between the studied variables, and it shows how significant is this relationship. Also, it gives an indication of the absence and existence of problem of multicollinearity.  Note(s): ***, ** and *, indicates significant at 1%, 5% and 10% levels, respectively Table 4. Correlations matrix AJEB 5,3 and the ROA and ROE is negatively significant. The Shariah governance mechanisms (BOD, AC and SSB) have an insignificantly positive association with ROA, while only the AC has a significantly negative influence on ROE. For leverage, the correlation between LEV and the ROA and ROE is positively insignificant. The link between size and ROA and ROE is negatively significant. The association between age and ROA is positively significant, whereas for ROE is insignificant. The findings illustrate that all correlation coefficients between the independent variables are very weak. This indicates the absence of multicollinearity. This result is confirmed by VIF values, where the highest is 1.66 (Gujarati, 2003).

Regression analysis
Tables 5-7 report fixed-effects GLS regression results of the link between AAOIFI disclosure, the Shariah governance variables and the financial profitability for the studied banks. Table 5 reports the findings of model 1 through testing the first hypothesis namely the positive effect of AAOIFI disclosure on performance in IBs. Using ROA as the dependent With BOD-index 5 average of the value assigned to each administration member; AC-index 5 average value assigned to each audit committee member; SSB-index 5 average value assigned to each Shariah council member; Size 5 Log Total Asset; Lev 5 Total Debt to Total Assets; Age: number of years since foundation. ***, ** and *, indicates significant at 1%, 5% and 10% levels, respectively  Shariah governance and financial performance variable, the Wald-chi 2 is equal to 3548.03 (0.000) showing that model 1 is significant. The Rsquared (R 2 ) is equal to 7.36% implies that the independent factors explain 7.36% of the variation regarding the impact of AAOIFI disclosure on ROA. For the ROE performance measure, the model is significant (p-value 5 0.000). The R 2 value is 5.46% implies that the independent factors explain 5.46% of the variation regarding the impact of AAOIFI disclosure on ROE. Table 5 shows also a significantly negative effect of the disclosure on the ROA and ROE. This earlier finding indicates that disclosure negatively affects the financial profitability. Indeed, disclosure might have a negatively value, even if its production is costless to the bank. Investors may perceive themselves to be worse off because they consider that the bank is disclosing information which might be exploited to their detriment. This finding was confirmed by Hassan et al. (2009). According to this impact, we reject the first hypothesis (H1). However, this finding contradicts some of the works like Albarrak and El-Halaby (2019) and Tabash (2019). These earlier studies found positively significant link between the two variables, meaning that IBs, who disclose more information to their stakeholders, gain the trust of shareholders. Consequently, it guides to an improving financial performance. While Elgattani and Hussainey (2020) found a positively insignificant association between disclosure and performance. It can be caused by two reasons. Firstly, the main objective of BI is to abide by the Shariah principles to have an active part in the community and to provide the value expectation to the customers and shareholders by focusing on the respect of the Shariah. Second, other factors like the attributes of the SSB or the disclosure of corporate social responsibility have more influence on the performance of IBs than AAOIFI governance disclosure (Aslam et al., 2018).
For the control factors, Table 5 indicates that size, age and leverage have a significant impact on ROA. This result is contradictory as Elagattani and Hussainey (2020) who found no significant impact between bank size and performance. However, Albarrak and El-Halaby (2019) showed no significant impact between age and performance. For ROE, the two variables, size and age are significant except for leverage which is insignificant. Table 6 reports the results of model 2 by investigating the impact of the Shariah governance indicators on the ROA and ROE for the analyzed banks. The model is significant 0.0000*** 0.0000*** Note(s): This table presents the regression results of different measures of disclosure and corporate governance on performance. With: DCI: Disclosure index; BOD-index 5 average of the value assigned to each administration member; AC-index 5 average value assigned to each audit committee member; SSBindex 5 average value assigned to each Shariah council member; Size 5 Log Total Asset; Lev 5 Total Debt to Total Assets; Age: number of years since foundation. ***, ** and *, indicates significant at 1%, 5% and 10% levels, respectively Table 7. The impact of disclosure and Shariah governance on performance in Islamic banks (fixed-effects GLS regression) AJEB 5,3 and R 2 are 0.0632 and 0.0590 for ROA and ROE, respectively. The R 2 value indicates that the three indices of GC explain 6.32% of the variation regarding the impact of BOD, AC and SSB on ROA. Also, the R 2 value of 0.0590 (5.90%) indicates that the three indices of GC explain 5.90% of the variation regarding the impact of BOD, AC and SSB on ROE. We discuss results when the dependent variable is ROA. Then, we consider the ROE results. Table 6 shows that BOD has a positively significant effect on the IBs' performance. This impact ameliorates the qualifications of the BOD members and the Islamic banking requirements. Also, it explains that the rise in the council size results in maximizing the performance. The larger BOD size is better for performance of bank and is more capable for monitoring the management. According to this impact, we accept the second hypothesis (H2). This result confirms Darwanto and Chariri (2019) and Nawaz (2019) findings. Al-Malkawi and Pillai (2018) found a negative but significant link between these two variables. The negative sign is delineated by the larger board of directors that can lead to free rider issues, time consuming decisions and increasing agency costs. In addition, there are high information asymmetries among the board members, when the large size of BOD has a negatively influence on performance. However, it contradicts the previous findings by Abdul Rahman and Bukair (2015), Sheikh and Kareem (2015), Abdallah and Ismail (2017) and Ajili and Bouri (2018) who found that BOD has a little influence on the financial performance. The insignificant impact is explained by the struggle between the qualifications of BOD members and the Islamic banking requirements or by the negligence of the system of recruitment of BOD members. Our result differs from those of Haddad et al. (2020), who investigated that the BOD negatively affects the financial performance for conventional banks. Table 6 shows also that the AC-index has a significantly positive impact on ROA. This result explains that the large size of AC helps to increase the performance of IBs. Also, the larger audit committees are able to detect the potential problems through the increase in resources and they provide more skills in evaluating the accounts of IBs. The latter would enhance the performance. According to this impact, we accept the third hypothesis (H3). Thus, this finding is coherent with Aslam and Haron (2020b). Also, it is like for the conventional banks where Haddad et al. (2021) concluded that the correlation between the conventional banks' ROA and AC is statistically significant. But the AC has a significantly negative impact on ROE. This explanation was proposed by Wild (1996), who argued that the main focus of the AC was to guarantee a better quality of disclosed financial reporting but not to increase the financial performance. The large size AC assists to ameliorate the performance of IBs (Aslam and Haron, 2020b). However, this study contradicts Ajili and Bouri (2018) and Elgattani and Hussainey (2020), who found no significant relationship between AC and financial performance.
From Table 6, the SSB-index has also a negatively significant effect on ROA. This means that SSB members lack full information related to the internal actions of the IB. This result may be manifested in performance as the higher value relate to the processing and preparation of the data. Furthermore, the small size of SSBs proved to be effective in developing consensus, reducing agency cost, bettering communications and having good control. It leads toward an improving performance. According to this impact, we reject the fourth hypothesis (H4). This finding is confirmed with Alsartawi (2019).The SSB has positively and significantly impact on ROE. This result implies that SSB members were generally specialized jurists in the Islamic commercial jurisprudence. The principal responsibility of these members is to assure the accordance of operations and dealings via Islamic rules. Also, large SSBs have different expertise and skills allied to the various schools of fiqh. Therefore, it prompts a superior understanding of services, which improves performance (Khan and Zahid, 2019). The importance of SSB is coherent with Mollah and Zaman (2015), Nomran and Haron (2019) and Aslam and Haron (2020b) who disclose that the larger members of SSB are enhancing the overall financial performance in IBs. Also, larger Shariah governance and financial performance SSB tends to perform better because the large size of SSB has more innovative ideas and skills to ameliorate the profitability. However, this result is opposite to Ajili and Bouri (2018).
Regarding control factors, the size of bank has negatively and significantly association with ROA and ROE. This result suggests that small Islamic banks were more possible to have good performance. This finding is coherent with Abdul Rahman and Bukair (2015) and Ajili and Bouri (2018). Moreover, the leverage has an insignificantly positive correlation with the ROA and ROE. Therefore, larger IBs with smaller debts were more likely to have a higher performance than smaller indebted. Finally, bank age has a positively and significantly association with performance, which is explained that IBs were very old in age and had an important experience. Table 7 concludes that the link between AAOIFI disclosure and performance measured by ROA and ROE has a significant and negative. We notice that this result is equivalent to the finding of Table 5. Our result is like Buallay et al. (2020), who showed that this association is significantly negative for conventional banks' performance. For sub-indices of BOD and AC, the outcome of Tables 6 and 7 is similar even in the existence of the disclosure as the independent variable. But, the outcome of SSB in Table 7 differs from the outcome of Table 6 only for ROA. They have a positively insignificant association. This explains that the SSB was not impacted with the IBs performance. The insignificance of SSB is high because they not only take care of the Shariah compliance but they also involve in the Shariah audit, distribute income to investors and answer the issues raised by the stakeholders (Noordin and Kassim, 2019).

Robustness checks
To verify the robustness of our findings, first, we estimated model 3 for the ROA using a twostep system of the GMM method (Table 8) (Aslam and Haron, 2020a, 2020b, 2020b. Second, we considered a global governance index and a 3SLS estimation.

Two-step system of generalized method of moments (GMM)
We use the two-step system of GMM adopted by Arellano and Bover (1995) and Blundell and Bond (1998) for endogenous tests. It creates a double equation for the first difference of all series and uses GMM to test the model using the lagged values of the variables. First-order differentiation eliminated unobserved heterogeneity and omitted variable bias. The result is analyzed with two tests: Sargan and serial correlation (Arellano and Bond, 1991 in the equation. Like the serial correlation test, the zero hypothesis of no first-order serial correlation (AR1) must be rejected, while the zero hypothesis of no second-order serial correlation (AR2) must not be rejected. We used the Roodman's (2009) "xtabond2" module in Stata to obtain the system's GMM estimates.
The diagnostic tests reported in Table 8 show that the model is statistically insignificant for the first-order autocorrelation in first differences (AR (1)), second-order autocorrelation in second differences (AR (2)) and the Hansen J-statistics of over-identifying restrictions. The residuals in the first and second difference AR (1) and AR (2) are not correlated. This result indicates that the instruments are applicable in the two-step system of GMM estimation. Indeed, disclosure and sub-indices of the Shariah governance are related to the ROA and ROE in IBs. For the empirical results of Table 8, we find that the disclosure (DCI), BOD and AC variables have the same sign and significance of those of Table 7 (model 3). For the SSB, we observed the same sign and significance founded in Table 6 (model 2). Consequently, we conclude that the GMM results confirm the GLS one.

Three-stage least squares (3SLS)
To eliminate the endogeneity problem from simultaneity bias (if any), we omitted both corporate governance (summation of BOD, AC and SSB) and disclosure by developing the following regressions: The definitions of the variables presented in the above models are described in Table 1.
Whereas there is a variation in the levels' significance for AAOIFI disclosure and CG, results are still similar to those reported in Tables 6 and 7. The coefficients for disclosure and corporate governance variables are the same as the results presented through using the twostep of the GMM (Table 8) and the GLS-Fixed Effect models (Tables 6 and 7). Thus, results from 3SLS (Table 9) are coherent with the two-step of the GMM and the GLS-Fixed Effect models.
Overall, the robustness' checks in this section with the two-step of the GMM and 3SLS found that the disclosure and the Shariah governance influence on the performance of IBs are the same as the main results presented earlier in Tables 6 and 7 0.0000 0.0000 0.0000 Note(s): ***, ** and *, indicates significant at 1%, 5% and 10% levels, respectively Table 9. 3SLS model-CG, SSB and bank performance Shariah governance and financial performance

Conclusion
The main focus of this research is to check whether AAOIFI disclosure and the Shariah corporate governance help IBs to do better and to create shareholder value. In especial, our aim is to study the link between (1) AAOIFI disclosure, (2) Shariah corporate governance (3) AAOIFI disclosure, and Shariah governance and the performance of IBs. Our objective is motivated by the paucity of research that examined the association between AAOIFI disclosure and Shariah corporate governance and the performance of IBs.
This study investigates this association on 47 IBs from 10 countries operating in the MENASA region from 2012 to 2019. In this paper, the disclosure and Shariah governance index are used as independent variables. ROA and ROE are used as financial performance measurement. The size, the leverage and the age are used as control variables. Across different methods, our findings are robust.
Findings showed that the link between disclosure and performance has a negatively significant (Tables 5 and 7). This result is conflicting with our hypothesis. Therefore, we concluded that disclosure negatively affects IBs' performance. Concerning the relation between corporate governance and performance, we conclude a positively significant association between BOD and the bank performance measured by ROA and ROE in Tables 6 and 7. This result means that the Board of Directors plays a higher role in the performance of the IBS. However, in the GLS method (Tables 6 and 7), AC has a significantly negative influence on the performance (ROE) of IBs but has a positively significant impact on ROA. This could explain that the effective AC enhances the level of financial reporting and transparency as a safeguard of its reputation. This ultimately enhances its performance. The statistical regression tests (Tables 6 and 7 for ROE) found that the effect of SSB has a significant on the performance of IBs. These findings mean that higher CG of IBs in the MENASA countries was oriented to maximize the production of shareholders.
As far as IBs findings are concerned, the implication is necessary to guarantee that SSB works in coordination with the BOD. The current research may help and encourage IB to show more information. Besides, this research may be a valuable source of knowledge for policymakers, regulators and stakeholders to the skill of the governance practices and disclosure of IBs. Consequently, it leads to a higher level of performance. In addition, this research has several limitations. First, we focus only on two measurements of performance (ROA and ROE). In future study; we will use other measures such as earning per share. Secondly, only 322 annual reports were tested over eight years. Also, we will conduct researches on several years. Finally, this research examines only IBs, further study may check other IFIs.