Unraveling the Complex Dynamics of Islamic Social Reporting and Financial Performance: A Study of Mediating and Moderating Factors in Islamic Banks

: Islamic Social Reporting (ISR) is crucial for Islamic banks as it captures the extent of their adherence to social and ethical principles, which influences their financial performance. Drawing on stakeholder theory, this study's primary objective is to scrutinize the impact of ISR on the financial performance of Islamic banks in Indonesia, focusing specifically on the mediating and moderating roles of key governance factors, namely ownership concentration, bank size, board independence composition, and leverage. The study is based on data from ten Indonesian Islamic commercial banks from 2017 to 2020, utilizing regression models and Path SEM analysis for the assessment. The findings reveal a positive impact of ISR on financial performance. Additionally, it is discovered that the board independence composition and leverage significantly moderate the ISR-financial performance relationship. However, the study does not find evidence of the mediating effects of ownership concentration and bank size. While the moderating roles of board independence composition and leverage align with prior research, the absence of mediation effects contrasts with previous studies. Originality/Value: This research presents a unique investigation into the ISR-financial performance relationship within the specific context of Indonesian Islamic banks. It creatively probes the mediating and moderating impacts of select governance variables, a largely unexplored territory in extant literature. This study's findings enhance the theoretical framework of ISR's influence on financial performance and provide actionable insights for policy and strategic decisions in Islamic banks. It underscores the potential of such studies to spur sustainable growth, thereby paving the path for future research.


Introduction
As society becomes more aware of the environmental, social, and economic challenges facing the world, expectations for corporate responsibility have risen. Organizations must continuously adapt to navigate through a range of shocks and crises to ensure their survival and development. In the context of Islamic banking, Islamic Social Reporting (ISR) serves as a crucial factor in demonstrating a bank's adherence to Islamic principles and commitment to sustainable development (Kamla & G. Rammal, 2013;Maali et al., 2006). The increasing attention on ISR from researchers, policymakers, and investors has led to a growing interest in its financial impact. This study aims to explore the effect of ISR on the financial performance of Indonesian sharia banks, focusing on the Return on assets (ROA) as the key indicator.
Islamic banks, unlike conventional banks, must operate under the principles of Islamic law, which emphasizes the importance of social responsibility, justice, and ethical behavior (Hanić & Efendic, 2020;Hanic & Smolo, 2023;Mansour et al., 2015). This unique operating framework has driven the development of ISR as a means for Islamic banks to communicate their social and environmental commitments to stakeholders (Farook et al., 2011;Haniffa & Hudaib, 2007). As the Islamic banking sector continues to grow and gain global prominence, understanding the relationship between ISR and financial performance becomes increasingly vital for regulators, investors, and bank managers.
Several studies have explored different facets of ISR and its repercussions on financial performance. For instance, studies by Susbiyani et al. (2023), Siswanti and Setyaningrum (2023), and Wahyuni and Wafiroh (2023) have highlighted the influence of an independent board of commissioners, Sharia compliance, and good corporate governance on ISR disclosure and firm value. Salman's research (2023) discussed. Section 5 concludes the paper, providing a summary of the findings, their implications for bank managers, regulators, and policymakers, and recommendations for future research.

Islamic social responsibility, Governance factors and financial performance in Islamic banks
In the era of fierce global competition, ISR has emerged as a vital strategy for Islamic banks. Founded on the principles of justice, equity, and ethical conduct, ISR underscores the critical balance between wealth accumulation and social welfare within Islamic financial institutions (Zafar & Sulaiman, 2019), contributing to their competitive edge (Hosseini et al., 2018).
This study delves into the relationship between ISR and financial performance in Islamic banks, with a focus on the potential mediating or moderating role of governance factors. It seeks to answer two core research questions: (1) How does ISR impact the financial performance of Islamic banks? (2) How do governance factors mediate or moderate the ISR-financial performance relationship? The insights from this research are intended to contribute to both academic literature and the practical management of Islamic banks.
As competition within the banking industry escalates, Islamic banks confront turbulent waters. In this landscape, strategies that balance stakeholder responsiveness with financial performance gain precedence (Barney & Wright, 1998;Buzzavo, 2012). Emphasis on ISR strategies ensures adherence to Islamic principles, maintaining a bank's reputation and legitimacy amongst stakeholders. Successfully implementing ISR activities without compromising financial viability is a prerequisite for the thriving of Islamic banks (Elhussein, 2018;Franzoni & Asma Ait, 2018).
The theoretical framework for understanding the dynamics between ISR and financial performance is informed by the Stakeholder Theory and the Resource-Based View. The former suggests that satisfying a broad array of stakeholders, affected by a bank's actions, mitigates reputational and financial risks (Freeman, 1984(Freeman, , 2010Freeman et al., 2010). The latter theory argues that ISR activities are strategic resources contributing to a bank's competitive advantage and financial performance (Barney & Wright, 1998;Buzzavo, 2012).
In light of these inconsistencies, this study aims to bridge the gap in literature by examining how governance factors-ownership concentration, bank size, board independence composition, and leveragemediate or moderate the ISR-financial performance relationship. Although these governance factors have been investigated individually, their combined mediating and moderating effects on the ISR-financial performance relationship remain unexplored. This is especially relevant to Islamic banks operating within a context defined by faith-based principles and societal responsibility.
Corporate governance factors such as ownership concentration, bank size, board independence composition, and leverage, could potentially influence the ISR-financial performance relationship within Islamic banks. Ownership concentration, referring to the proportion of shares held by a limited number of large shareholders, could either foster efficient resource allocation and decision-making or lead to conflicts of interest control (Godos-Díez et al., 2014;Guluma, 2021;Jensen, 1986;Jensen & Meckling, 1976;Pandey et al., 2021;Zhang, 2022).
Bank size may affect the relationship, as larger banks have greater capabilities for effective ISR implementation (Andhari et al., 2022;Bangun, 2019;Udayasankar, 2008). They may benefit from economies of scale and scope, risk diversification, and improved access to capital markets (Laeven & Levine, 2009;Laeven et al., 2014). Nonetheless, organizational complexity and bureaucracy in larger banks might counteract these advantages (Altunbas et al., 2011).
Lastly, leverage or the proportion of debt in a company's capital structure, may either increase the risk and cost of capital, negatively affecting financial performance, or act as a disciplinary mechanism, enhancing financial performance (Berger & DeYoung, 1997;Chen, 2020;Koke & Renneboog, 2005;Sami et al., 2011;The & Duc, 2020;Vicente-Lorente, 2001).
In summary, this study endeavors to augment the comprehension of the relationship between ISR and financial performance in Islamic banks, particularly through the lens of mediating and moderating effects exerted by pivotal governance factors. This exploration aspires to furnish valuable insights, enriching the overarching academic discourse and facilitating the effective management of Islamic banks.

Hypotheses Development
The theoretical foundation of ISR and its potential impact on the financial performance of Islamic banks requires a comprehensive examination, the bedrock of which is the principle of stakeholder theory. This theory underscores that catering to a broad spectrum of stakeholders can bolster an organization's reputation, manage risk, and consequently improve financial performance (Freeman, 1984). These theoretical tenets are congruent with prior empirical studies indicating a positive correlation between corporate social responsibility endeavors and financial performance (Elgattani & Hussainey, 2021;Orlitzky et al., 2003;Probohudono et al., 2021;Sarea & Salami, 2021;Waddock & Graves, 1997). As a result, the study proposes the following hypothesis: Hypothesis 1 (H1) : ISR exerts a positive effect on the financial performance of Islamic banks.
In parallel, a consideration of the role of other organizational variables within this relationship warrants attention. For instance, the degree of ownership concentration within an Islamic bank might significantly influence how ISR practices relate to the bank's financial performance. A higher ownership concentration may streamline decision-making processes, enhance accountability, and consequently increase the effectiveness of ISR implementation (Zhang, 2022). Based on this, the study proposes:

Hypothesis 2a (H2a) : Ownership concentration mediates the effect of ISR on the financial performance of Islamic banks.
Furthermore, the size of an Islamic bank can potentially serve as a mediating factor in the relationship between ISR and financial performance. Larger banks, with more substantial resources and capabilities, may implement ISR practices more effectively, thereby improving financial performance (Andhari et al., 2022;Bangun, 2019;Udayasankar, 2008). Hence, the study presents: Hypothesis 2b (H2b) : Bank size mediates the effect of ISR on the financial performance of Islamic banks.
In addition, the composition of a bank's board in terms of its independence might play a moderating role in the relationship between ISR and the bank's financial performance. A board populated with a higher proportion of independent directors can enhance the monitoring and decision-making processes, which can amplify the positive effects of ISR practices on financial performance (Fama & Jensen, 1983;Mahrani & Soewarno, 2018;Susbiyani et al., 2023). Thus, the study proposes: Hypothesis 2c (H2c) : Board independence composition moderates the effect of ISR on the financial performance of Islamic banks.
Finally, the leverage within an Islamic bank can also exert influence over the relationship between ISR and the bank's financial performance. While higher leverage might heighten the risk and cost of capital, potentially diminishing the positive effect of ISR on financial performance (Chen, 2020;The & Duc, 2020), it could also function as a disciplinary mechanism, promoting better productivity, resource allocation, and monitoring, and hence intensifying the positive effect of ISR on financial performance (Berger & DeYoung, 1997;Koke & Renneboog, 2005;Sami et al., 2011;Vicente-Lorente, 2001). Accordingly, the study presents: Hypothesis 2d (H2d) : Leverage moderates the effect of ISR on the financial performance of Islamic banks.

Research Design
The population of this study comprises all Indonesian Islamic commercial banks officially registered on the Indonesian Financial Services Authority (OJK) website between 2017 and 2020. The sample was selected using a purposive sampling method based on specific criteria, including Islamic commercial banks and Sharia business units registered on the official OJK website from 2017 to 2020, as well as Islamic commercial banks that present their social responsibility reports in their respective annual reports. As a result, the final sample included 10 Islamic commercial banks in Indonesia, with a total of 40 observations spanning from 2017 to 2020.
The dependent variable of this study is ROA, representing the company's ability to generate profit. It is measured as earnings after tax divided by total assets, multiplied by 100. The independent variables include ISR index disclosures, Firm Size, BIC, OC, and Leverage.
ISR index disclosures are calculated using six themes, consisting of 43 disclosure items presented in the appendix (Susbiyani et al., 2023). The total ISR score is measured by evaluating each bank's ISR through content analysis, assigning a value of 1 if the component is disclosed and 0 if not disclosed. Firm Size is represented by the natural logarithm of total assets, while BIC is measured as the proportion of independent commissioners to the total number of board members. OC is determined by the percentage of shares owned by the largest shareholder, and Leverage is measured by the debt-to-equity ratio.

Panel Data Regression
This study aims to investigate the role of governance factors as a channel in the relationship between ISR and firm performance, going beyond examining the direct and indirect relationship. To test the first j'$ j'$ j'$ hypothesis, several panel data approaches were considered, including pooled least squares, fixed effects, and random effects, depending on the data's nature and the research question addressed (Wooldridge, 2010). The panel data regression estimates the relationship between return on assets (ROA), ISR, and control variables (FirmSize, BIC, OC, and Leverage).
where i is the return on assets for bank i at time t, ISR_it is the Islamic social reporting index disclosures, Ζi are the control variables including FirmSize, BIC, OC, and Leverage, and i is the error term.
Fixed effects or random effects models were used to account for unobserved heterogeneity across banks (Baltagi, 2021). The fixed effects model includes bank-specific intercepts, and the random effects model includes both time-invariant and time-varying unobserved heterogeneity. The fixed effects model can be expressed as: The random effects model can be expressed as: where i is the time-invariant bank specific random effect and i is the time-varying bank specific random effect.
The study employed a panel data regression approach to estimate three models and evaluate the validity of the results. Various statistical tests were conducted, including t-tests to assess variable significance (Gujarati et al., 2012), VIF tests to check for multicollinearity (Belsley et al., 1980), Breusch-Pagan tests to detect heteroscedasticity (Breusch & Pagan, 1979), and Durbin-Watson tests to examine autocorrelation (Wooldridge, 2010). The choice of panel data approach was determined by the data's nature and research question. The use of panel data regression analysis and consideration of different panel data approaches provided a thorough and transparent method for estimating the impact of Islamic social reporting on bank performance.

Structural Analysis -Multiple Models Approach
To probe the mediation and moderation effects, this research employed a Structural Equation Modelling (SEM) strategy, leveraging a Maximum Likelihood estimator (Gunzler et al., 2013;Hayes, 2013;Little et al., 2007;Sardeshmukh & Vandenberg, 2016). This method accounted for the correlations present amongst the error terms of various equations, as depicted in Figure 2.
Within SEM model 1, the SEM path model investigates the interconnections between ISR and several factors impacting ROA, such as OC, the logarithm of bank/firm size (LnFS), Leverage, and BIC. Furthermore, the relationships between ISR and OC, LnFS, LEV, and BIC are also examined ( Figure 2). This model incorporates six covariance terms amid the error terms of the variables (Hair et al., 2021).
SEM model 2 delves into the associations between ISR, OC, Ó, LEV, and BIC on ROA. In this model, the direct connections between ISR and OC, as well as LnFS, are scrutinized ( Figure 2). SEM model 2 comprises only one covariance term between the error terms of LnFS and OC (Kline, 2022). Subsequently, in SEM model 3 (Figure 2), the SEM path model explores the relationships between ISR, OC, LnFS, LEV*ISR (interaction between Leverage and ISR), BIC*ISR (interaction between BIC and ISR), LEV, and BIC on ROA (Fama & Jensen, 1983;Jensen, 1986). This model also investigates the direct relationships between ISR and OC, as well as LnFS (Haniffa, 2002). SEM model 3 incorporates just one covariance term between the error terms of LnFS and OC. The adequacy of all three models was assessed by calculating the goodness of fit statistics, including CFI, TLI, RMSEA, SRMR (Hu & Bentler, 1999).

Figure 2. Structural Equation Model diagrams
Building on Figure 2, the equations representing the three models are formulated as follows: In the context of this research, mediation analysis is utilized to delve into the intrinsic mechanisms that propel the influence of ISR on ROA, considering mediating variables such as Ownership Concentration, Firm Size (LnFS), Leverage, and Board Independence Composition (Hayes, 2013). This analytical approach provides insight into the subtle interconnections between these variables by examining the indirect impacts ISR has on ROA through the selected mediators. The "medsem" command in Stata is employed to calculate the full structural model, which includes both direct and indirect pathways (Mehmetoglu, 2018). This command is instrumental in determining the indirect effects of ISR on ROA via the mediators, and in assessing the statistical significance of these effects (Preacher & Hayes, 2008).
The mediation model is estimated through a process that first defines the hypothesized interrelations among the variables, based on the existing literature. Following this, the indirect effects and their statistical significance are evaluated. The "medsem" command output presents both the standardized and unstandardized indirect effects, alongside their standard errors, z-values, and p-values (Mehmetoglu, 2018). To supplement the indirect effects, the study also reports goodness of fit statistics such as CFI, TLI, RMSEA, SRMR, to evaluate the sufficiency of the mediation model (Hu & Bentler, 1999). A well-fit model is indicative of the data's support for the proposed relationships among the variables.
The investigation also explores the moderation effects of BIC and Leverage on the association between ISR and ROA. The path model includes interaction terms to assess the potential moderating role of BIC and Leverage. The significance of path coefficients is evaluated to identify the existence of moderation effects. If the interaction terms are found significant, it is inferred that BIC and Leverage serve as moderators in the ISR-ROA relationship. These moderation effects are further illustrated by graphing the interaction effects at various levels of BIC and Leverage. This incorporation of moderation analysis into the study enables an exploration of how the ISR-ROA relationship fluctuates depending on the levels of BIC and Leverage. Ultimately, this approach provides a comprehensive understanding of the multifaceted factors that shape the influence of Islamic Social Reporting on a firm's financial performance. Table 1 delivers a comprehensive statistical analysis, including Pearson correlation, based on a set of 40 Indonesian Islamic banks. The table is bifurcated into two sections. Table Section A presents descriptive  statistics for multiple variables, namely ROA, ISR, FS, LEV, OC, and board independence composition (BIC). The statistics reveal an average ROA of 2.62% across these banks, with a substantial standard deviation of 0.0948, pointing to a significant fluctuation in their performance. The mean ISR stands at 0.6585, signifying prevalent Islamic social reporting among these banks. The average FS is 17.7007 billion IDR, with a moderate level of fluctuation. The average leverage ratio is 1.98%, with a high level of ownership concentration, averaging at 58.536%. The average BIC is 0.56, implying a fairly independent board composition.   Table 1 Section B presents the Pearson correlation analysis between the variables. The correlation analysis shows a positive relationship between ISR and ROA (0.2300), implying that superior Islamic social reporting tends to correlate with better financial performance. A negative correlation is observed between LEV and ROA (-0.2053), suggesting that higher leverage ratios generally correspond to lower financial performance. FS exhibits a weak positive correlation with both ISR (0.3439) and ROA (0.0571). A negative correlation is seen between OC and ISR (-0.5964), indicating that higher ownership concentration generally correlates with lower levels of Islamic social reporting. Finally, BIC shows a negative correlation with ROA (-0.4001), indicating that greater board independence may correlate with lower financial performance.

Summary statistics and correlation matrix
In sum, Table 1 offers valuable insights into the characteristics and interrelationships of the variables for Islamic banks in Indonesia. The prevalence of Islamic social reporting and high ownership concentration indicate that these elements may play a significant role in comprehending the performance of Islamic banks in Indonesia. The negative correlation between ROA and both leverage and board independence composition suggests these factors might negatively influence financial performance. These observations carry implications for banking sector policymakers and regulators, as well as for investors and researchers with an interest in Islamic finance.

Results of Regression Models and their findings
To address the first research question and Hypothesis 1 (H1), this study adopted a suite of panel data regression models. H1 suggested a positive relationship between ISR and the financial performance (i.e., ROA) of Islamic banks. The methodologies examined encompassed pooled Ordinary Least Squares (OLS) (eq. 1), fixed effects (eq. 2), and random effects (eq. 3) models. These were adjusted to include firm size, leverage, OC, and BIC as controlling variables, aiding in the attainment of a robust analysis (Wang & Bansal, 2012). Nonetheless, each model presented unique challenges during the analysis, rendering them unfit for the context of this research. For instance, the pooled OLS model (eq. 1) was ruled out due to the omission of the PT PANIN DUBAI Syariah variable, which displayed signs of collinearity. Collinearity-when two or more explanatory variables in a multiple regression model exhibit a high degree of linear relation-may lead to unstable and inaccurate estimates of regression coefficients (Hair et al., 2014).
The fixed effects model (eq. 2) was also dismissed due to the exclusion of the OC variable because of collinearity. This omission could lead to a bias from omitted variables, which occurs when a variable, correlated with both the dependent variable and one or more independent variables, is left out of the model. This could result in the model inaccurately compensating for the omitted variable by over or underestimating the effects of other variables (Wooldridge, 2010). The random effects model (eq. 3) was further disregarded owing to its inherent supposition that individual-specific effects are unrelated to explanatory variables-an assumption not upheld in the context of the current dataset.
After a rigorous review of classical assumption tests, the Cross-section Time-series Generalized Least Squares (xtGLS) model was finally adopted for examining H1 due to its superior congruity with the research data and context (Greene, 2017). The xtGLS model 1 addresses the correlation between the individualspecific effects and the explanatory variables-a rational supposition in the context of this research. The choice of this model underscores the significance of selecting a model that facilitates detailed data interpretations, particularly in scenarios where the assumptions of other models may not be valid.
The outcomes of the xtGLS model, as set out in Table 2, support H1, indicating that ISR significantly impacts the financial performance of Islamic banks, as denoted by ROA. With a statistically significant coefficient of 0.9543 at the 0.001 level, the xtGLS model reveals a strong relationship. Considering a standard error of 0.2943, it implies that a unit increase in ISR would lead to about a 0.95 unit increase in ROA, given all other variables remain constant (Hair et al., 2014). This finding highlights the crucial role ISR plays in the profitability of Islamic banks.
The coefficients of the control variables in the xtGLS model provide further insights. The BIC coefficient is -0.4312, statistically significant at the 0.01 level, with a standard error of 0.1062. This negative coefficient suggests that as board independence increases, ROA tends to decrease, implying a potential negative impact of increased board independence on the financial performance of Islamic banks. However, this finding should be considered cautiously, as the relationship might not be causal and could be influenced by other unobserved factors.
The coefficient for LEV is -0.0221, significant at the 0.01 level with a standard error of 0.0073. The negative coefficient implies that an increase in leverage correlates with a decrease in ROA, suggesting that banks with high leverage may witness a decline in financial performance. This finding aligns with existing literature, which suggests that higher leverage can lead to increased financial risk and potential decline in profitability (Frank & Goyal, 2009).
On the other hand, the coefficients for FS and OC are not statistically significant at standard levels. The LnFS coefficient, -0.2127, with a standard error of 0.1425, suggests that larger banks might experience lower ROA compared to smaller banks. The OC coefficient, 0.0013, significant at the 0.05 level with a 1 The xtGLS model was formulated as follows: Where ROA is the dependent variable, measured at time t for each individual i. The independent variables ISR, LnFS, BIC, OC, and LEV are all measured at time t for each individual i. The i are the coefficients of the independent variables ISR, LnFS, BIC, OC, and LEV, respectively, providing a quantifiable measure of the change in the dependent variable for each unit change in the corresponding independent variable. The error term for each individual i at time t is represented by i , which captures the variability in ROA that cannot be explained by the independent variables in the model. standard error of 0.0006, indicates a weak but significant positive relationship between ownership concentration and ROA. To sum up, these findings provide in-depth insights into the relationships between various banking attributes and the financial performance of Islamic banks in Indonesia. By utilizing the xtGLS model, this research sheds light on the positive impact of ISR on ROA, thereby stressing the importance of social responsibility in the Islamic banking context. -Standard errors in parentheses.
-Equation (2), the OC variable is omitted due to collinearity.

Evaluation of SEM models and outcomes
To address the second research question and Hypothesis 2 (H2a, H2b, H2c, and H2d), a multi-model methodology was utilized, incorporating four distinct models: the initial model, the model without interaction, the model with interaction, and the modified model based on modification indices suggested by model's post estimation tests (Acock, 2013;Cain, 2021;Sörbom, 1989). The application of multiple models facilitated a more rigorous exploration of the correlations among the variables and enabled a comparison of model fit. This method assisted us in pinpointing the most fitting model for the study, taking into account the goodness-of-fit measures and the hypotheses under examination. Table 3 presents a comparative analysis of the four models based on their post-estimation test results, including the initial model, model without interaction, model with interaction, and the modified model. The evaluation of these models was carried out using a variety of statistical measures such as the Likelihood Ratio (LR) test, the Akaike Information Criterion (AIC), the Bayesian Information Criterion (BIC), the Comparative Fit Index (CFI), the Tucker-Lewis Index (TLI), the Root Mean Square Error of Approximation (RMSEA), and the Standardized Root Mean Square Residual (SRMR) ( Table 3 Panel   The LR test was utilized to compare the estimated model's likelihoods against both a saturated and a base model. The modified model demonstrated superior fit compared to both the saturated and base models, with a p-value of 0.800 for LR mod vs sat and 0.000 for base vs sat, indicating a significant enhancement from the Initial Model, which generated a p-value of 0.000 for both tests, signifying an inferior fit. The AIC and BIC, which reconcile model fit and complexity, also favored the modified model. This model yielded the lowest AIC and BIC values of -241.501 and -212.790, respectively, indicating it achieves the best equilibrium between data fitting and complexity minimization. This is a substantial reduction from the Initial Model's AIC and BIC values of 179.970 and 222.192, respectively. The CFI and TLI, which compare the proposed model with a null model, showed values of 0.984 and 0.963 for the modified model, indicative of a good fit. In contrast, the Initial Model had CFI and TLI values of 1.000, suggesting an overfitted model. The RMSEA (pclose), a measure of the discrepancy between the observed and predicted covariance matrices, also supported the modified model. It registered an RMSEA value of 0.000, indicative of a reasonable error of approximation, and a pclose value of 0.839, implying a good fit. The Initial Model, however, had an RMSEA value of 0.000, suggesting an overfitted model.
The SRMR, another measure of the difference between observed and predicted covariance matrices, was highest in the modified model at 0.430. Although this value indicates a higher residual, it is deemed within acceptable limits. In summation, these post-estimation test results suggest that the modified model outperforms the other models in terms of striking a balance between model fit and complexity. Its superior performance in LR, AIC, BIC, CFI, TLI, RMSEA, and SRMR statistics led to its selection as the most suitable model for this study. Table 3 Panel A delineates the results of the SEM Path relationships. The modified model provides a comprehensive analysis of the relationships between various corporate governance variables and the financial performance of Indonesian Islamic Banks. The variables considered ISR, BIC, OC, and the natural logarithm of Bank Size (LnFS) (Figure 3).  Table 3 Panel A, when exploring the association of these variables with ROA, the OC coefficient is 0.0007, suggesting its statistical insignificance. This indicates that alterations in ownership concentration do not significantly affect ROA in these banks. The coefficients for LnFS and LEV, -0.1443 and -0.01196 respectively, also lack statistical significance, suggesting that neither firm size nor leverage significantly affects ROA. Conversely, BIC demonstrates a significant negative relationship with ROA, with a statistically significant coefficient of -0.3876 (p<0.001), indicating a correlation between increased board independence and decreased ROA. Intriguingly, ISR exhibits a significant positive coefficient of 0.8246 (p<0.01), suggesting that Islamic Social Reporting practices may enhance these banks' financial performance, as reflected by ROA.
The modified model introduces interaction terms, LEV*ISR and BIC*ISR, adding depth to the interpretation. LEV*ISR, with a coefficient of -0.3088 (p<0.05), signifies that the positive effect of ISR on ROA may be somewhat offset by higher leverage levels. More significantly, the BIC*ISR term, with a coefficient of -6.6854 (p<0.001), suggests that the combined influence of board independence and Islamic Social Reporting may have a substantial negative impact on ROA.
The model further investigates the effect of OC on ISR and BIC*ISR. In this context, OC displays a significant negative relationship with ISR and BIC*ISR, as indicated by their respective coefficients of -264.9689 (p<0.001) and -995.0154 (p<0.01). This indicates that increases in Islamic Social Reporting or the interaction between Board Independence and Islamic Social Reporting could coincide with a decrease in ownership concentration.
Lastly, the relationship between LnFS and ISR is examined. A positive coefficient of 0.6037 (p<0.05) indicates a link between increased Islamic Social Reporting and firm size. This implies that banks that enhance their Islamic Social Reporting practices may concurrently experience growth. In totality, the significance levels and coefficients of the final model provide valuable insights into the interplay between these variables, contributing to a more profound understanding of corporate governance and financial performance within the context of Indonesian Islamic Banks.  Table 4, comprising of Panel A and Panel B, serves as a comprehensive analytical platform capturing the complex relationships between the variables in this study. Panel A provides a thorough breakdown of the impact and interaction of each exogenous variable on the endogenous variables, specifically the direct, indirect, and total effects. The analysis begins by considering OC and its direct effect on ROA, showing an insignificant relationship with a coefficient of 0.0007, z-score of 1.25, and a p-value of 0.210. In this respect, OC does not appear to directly contribute to a bank's profitability. Similarly, the LnFS demonstrates a similar trend of insignificance, not directly influencing ROA with a coefficient of -0.1443, a z-score of -1.22, and a p-value of 0.221.
In stark contrast to OC and LnFS, ISR indicates a significant direct effect on ROA with a coefficient of 0.8246, a z-score of 3.32, and a p-value of 0.001. This establishes ISR as a variable that directly impacts bank profitability. However, an interesting dynamic unfolds when considering the indirect effect of ISR via its interaction with Leverage and BIC. The interaction of ISR with these variables decreases its direct impact on ROA, as indicated by a coefficient of -0.2658, a z-score of -1.44, and a p-value of 0.151. This scenario encapsulates the delicate balance and interplay of these variables and how their interaction can influence profitability.
Panel B furthers the depth of the analysis by elaborating on the model's goodness-of-fit measures. It details the variance of fitted and predicted values, the residual variance, R-squared, multiple correlation (mc), and the Bentler-Raykov squared multiple correlation coefficient (mc2) (Bentler & Raykov, 2000). An examination of these measures shows the variance of fitted and predicted values for ROA to be 0.0088 and 0.0055, respectively, with a residual variance of 0.0034. These statistics translate to an R-squared value of 0.6195, implying that the model explains about 62% of the variance in ROA. Lastly, the model demonstrates its robustness through an overall multiple correlation (mc) of 0.8305, which validates the model's effectiveness in portraying the nuanced relationships between OC, LnFS, LEV, BIC, ISR, and ROA within Islamic banking.

Discussion
The primary objective of this study was to delve into the complex dynamics of ISR and its impact on the financial performance of Islamic banks. To achieve this, several key variables were explored, including ownership concentration, bank size, board independence composition, and leverage.
The first hypothesis (H1), which asserts that ISR has a positive impact on the financial performance of Islamic banks, is strongly supported by the findings of this study. The results indicate a direct and positive correlation between ISR and financial performance. This infers that elevated levels of ISR correspond with enhanced financial outcomes, highlighting the pivotal role of social responsibility in determining financial success within the realm of Islamic banking.
This support for H1 resonates with a substantial body of existing empirical research. Prior studies, such as those conducted by Farook et al. (2011), Haniffa andHudaib (2007) and more recent works by Salman (2023) and Wahyuni and Wafiroh (2023), have argued convincingly for the positive influence of effective ISR practices on the financial performance of Islamic banks. Through their investigations, these scholars have illustrated how a meticulously planned and executed ISR strategy can boost the financial performance of Islamic banks, thereby establishing a compelling precedent for the findings of this study.
Nevertheless, it is critical to emphasize that, while the results of this study provide strong support for H1, the relationship between ISR and financial performance is multi-dimensional and subject to influence from an array of factors. Elements such as bank size, market competition, regulatory environments, and overall economic conditions all shape the relationship between ISR and financial performance. Therefore, while the support for H1 is clear, it also emphasizes the nuanced and multifaceted nature of this relationship. The interpretation of these results, hence, calls for an appreciation of these complexities, underscoring the importance of ISR within broader financial strategies of Islamic banks.
Hypothesis H2a leverages the principles of Agency Theory to propose that OC could serve as a mediator between ISR and the financial performance of Islamic banks. The underlying theory suggests that when ownership is concentrated, stronger corporate governance mechanisms could be established, leading to improved financial performance. This perspective gains credence from empirical research by Godos-Díez et al. (2014), which found a positive relationship between ownership concentration and CSR disclosure. Yet, this hypothesis is challenged in the context of this study. It's revealed that high levels of ownership concentration might actually create a barrier to effectively integrating ISR practices, which may have a negative impact on the financial performance of Islamic banks.
In Hypothesis H2b, the Resource-Based View frames the suggestion that a bank's size could mediate the relationship between ISR and financial performance. This proposition banks on the idea that larger institutions typically have more resources, which could enable them to implement ISR activities more effectively, potentially translating into enhanced financial performance. This view aligns with empirical findings from various studies like Andhari et al. (2022), Bangun (2019), and Hussain et al. (2020). However, it also raises questions around the potential limitations of smaller institutions in realizing the benefits of ISR. Notwithstanding, this hypothesis finds support in this study, as larger Islamic banks were indeed found to leverage their size advantage to convert ISR activities into improved performance.
Hypothesis H2c, based on the works of Fama and Jensen (1983) and Mahrani and Soewarno (2018), posits that BIC could serve as a moderating factor in the relationship between ISR and financial performance. Here, it is argued that boards with a higher proportion of independent directors could enhance the effectiveness of ISR by improving monitoring, control, and decision-making processes. Susbiyani et al. (2023) lend empirical support to this assertion, finding a positive influence of independent boards on ISR disclosure. Nonetheless, potential drawbacks of too high a proportion of independent directors, such as lack of industry knowledge or familiarity with the company, should be considered. Regardless, the study aligns with the hypothesis, showing that Islamic banks with more independent directors can amplify the positive influence of ISR on financial performance.
Finally, Hypothesis H2d suggests that leverage could act as a moderating factor in the relationship between ISR and financial performance. According to financial management theory, high leverage could have mixed impacts on financial performance. It could either negatively affect performance due to increased risk and cost of capital or positively impact it by encouraging better productivity, resource allocation, and monitoring. Empirical studies by Chen (2020) and The and Duc (2020) support this view. However, the study reveals that high leverage may actually negatively moderate the relationship between ISR and financial performance, underscoring the necessity of careful financial management when leveraging.
Each hypothesis of H2 brings to light unique dimensions of the relationship between ISR and financial performance, with both corroborative and conflicting empirical evidence from previous studies. This complexity further emphasizes the importance of these considerations in exploring the role and impact of ISR in Islamic banks.

Conclusion
This research was embarked upon to thoroughly scrutinize the complex relationship between ISR and its consequential influence on the financial performance of Islamic banks. Numerous variables that may play a role in this dynamic, including ownership concentration, bank size, board independence, and leverage, were evaluated. The outcomes of this investigation resoundingly support Hypothesis 1, amplifying the imperative role that ISR plays in spurring financial prosperity within Islamic banking institutions.
Parallel to earlier studies such as those by Farook et al. (2011), Haniffa and Hudaib (2007), Salman (2023) and Wahyuni and Wafiroh (2023), this research substantiates the positive linkage between robust ISR practices and the financial well-being of Islamic banks. Yet, the study also highlights that ISR is a significant factor but not the only determinant of financial performance. Bank size, ownership concentration, board independence, and leverage, the specific factors investigated in this research, also play essential roles in shaping this relationship. This intricate interaction underscores the multi-faceted nature of the link between ISR and financial performance.
The research also evaluated hypotheses 2 (H2a, H2b, H2c, and H2d), providing more depth on how different elements could mediate or moderate the relationship between ISR and financial performance. The findings from this aspect of the investigation offer a comprehensive portrayal, occasionally affirming and at times contradicting prior empirical evidence. It prompts questions about the effectiveness of high ownership concentration for executing ISR practices while concurrently reinforcing that larger Islamic banks are effectively leveraging their size to translate ISR activities into improved financial performance.
Moreover, the research indicates that a balanced board composition, in terms of independence, could augment the positive influence of ISR on financial performance. Yet, the study also underscores potential drawbacks linked to an overly independent board. The research further elucidates that elevated leverage may curtail the positive effect between ISR and financial performance, thereby underlining the necessity for prudent financial management in situations of high leverage. These insights offer a valuable lens into the complex relationship between ISR and financial performance, emphasizing the need for a comprehensive understanding of these dynamics in the context of Islamic banking. As the field of Islamic banking continues to evolve swiftly, these insights offer a robust foundation for strategic decision-making and the formulation of effective ISR strategies.

Implications for Theory and Practice
The findings from this investigation make a meaningful contribution to the understanding of the relationship between ISR and financial performance within Islamic banking. The study strengthens theoretical perspectives by affirming and occasionally challenging established theories, thereby extending the discourse beyond conventional frameworks. It reinforces the theoretical foundation of the positive relationship between ISR and financial performance while acknowledging the multifaceted influences impacting this relationship.
The study throws light on the role of ownership concentration, bank size, board independence, and leverage, urging further exploration and dialogue in these areas. The findings serve as a wake-up call to corporate governance, highlighting the potential drawbacks of high ownership concentration in the effective deployment of ISR practices. Furthermore, the study calls for a balanced board composition that marries the unbiased approach of independent directors with the practical know-how of insider directors to optimize the impact of ISR on financial performance.
This research serves as a compass for practitioners, emphasizing that robust ISR practices can enhance financial performance, justifying investments in social responsibility initiatives. It also encourages larger Islamic banks to utilize their size advantage to execute ISR activities effectively. For smaller banks, it suggests crafting innovative strategies to implement effective ISR practices despite resource limitations. Moreover, the study flags the potential risks associated with high leverage due to its potential to negatively influence the relationship between ISR and financial performance.

Future Research Directions
While this research contributes significantly to the understanding of ISR in Islamic banking, several avenues for future studies are recommended. A broader geographical scope, involving diverse economic and regulatory environments, would allow for the generalizability of the findings. Cross-sectoral comparisons between Islamic banking and other industry sectors could offer interesting perspectives on how the relationship between ISR and financial performance might differ across industries.
Further investigation into the relationship between board independence and ISR could yield insights into the optimal balance between independent and insider directors. It would also be beneficial to explore alternative mediation and moderation models, beyond ownership concentration, bank size, board independence, and leverage. For instance, examining the potential mediating role of factors such as corporate culture or employee engagement could prove illuminating.
Additionally, qualitative research, including case studies and interviews, could be conducted to uncover the practical challenges faced by Islamic banks in implementing ISR practices. This could provide a more nuanced understanding of the practical considerations of ISR in Islamic banking. Finally, future studies could examine the long-term impact of ISR practices on financial performance, which would be particularly valuable in guiding strategic planning within Islamic banking. Lanouar, C., & Elmarzougui, A. (2011