The Impact of CSR and Green Banking Practices on the Financial and Risk Sustainability of Commercial Banks in Bangladesh

Abstract

This study examines the impact of Corporate Social Responsibility (CSR) and green banking practices on the financial and risk sustainability of commercial banks in Bangladesh. Using panel data of 300 observations over the period 2013 to 2024, this study employed random effects regression models to assess how CSR and green banking influence profitability, financial stability, and risk exposure. A System GMM model is additionally used to address endogeneity concerns and validate the robustness of the results. The empirical findings indicate that CSR engagement by banks contributes positively to financial performance with superior shareholder returns. This positive relationship remains robust under the GMM framework. CSR also shows a significant negative association with non-performing loans, which becomes stronger after controlling for endogeneity. Green banking practices also exhibit a positive and statistically significant relationship with ROE under the baseline model. However, this relationship becomes statistically insignificant in the GMM estimation, indicating that the effects of green banking are sensitive to endogeneity.

1. Introduction

Corporate Social Responsibility (CSR) has increasingly emerged as a strategic priority for modern firms. It is widely used as a tool to attract investors and strengthen relationships with stakeholders (Yang et al., 2019). Nearly 90 percent of Fortune 500 companies publicly disclose information regarding their CSR initiatives (Sun & Cui, 2014). In the United States, a notable portion of professionally managed investment funds is invested in companies with strong CSR performance. Nearly one out of every nine dollars is allocated to such firms. Reflecting this practical importance, CSR has also become a prominent area of academic interest. Although there has been a huge volume of literature in the context of both developed and developing countries, there has been limited research in the context of Bangladeshi companies. This study contributes to the existing body of literature by bringing evidence from the banking sector in Bangladesh.

In recent times, Bangladesh’s banking sector has gone through major regulatory changes. These changes focused on sustainability and positive commitment towards the environment. In response to widespread concerns about environmental damage and financial sector stability, the Bangladesh Bank has adopted a structured policy guideline. These guidelines were introduced to enhance more responsible banking practices. Under these initiatives, the issues of CSR and green banking were incorporated into the regulatory framework for banks operating in Bangladesh. After these initiatives, the environmentally friendly and sustainability practices for banks in Bangladesh are not seen as a ‘merely’ voluntary practice, rather, they function within a regulatory and policy-driven framework. In this paper, we examine how the expenditure on CSR and green banking practices affects the financial and risk sustainability of commercial banks in Bangladesh. We assessed financial sustainability through profitability and financial stability. Risk sustainability is examined using non-performing loans for credit risk and the loan-to-deposit ratio for liquidity risk.

This study contributes to the current literature on sustainability in the banking sector. Firstly, we constructed a structured green banking practice index and we also used actual CSR data in our analysis. The index is aligned with the current regulatory framework. Secondly, using a stakeholder-related framework, we introduced a mediation analysis to understand whether CSR influences bank risk. Finally, this study focuses on an emerging economy context, i.e., Bangladesh, where sustainability initiatives are becoming policy-driven and increasingly important for financial institutions.

The paper is organized thus. Section 2 reviews pertinent research and develops research hypotheses. Section 3 describes the data and research methodology. Section 4 details the empirical findings. Section 5 concludes the study.

2. Literature Review

2.1. CSR and Financial and Risk Sustainability

CSR has been increasingly considered not only as a tool for enhancing corporate reputation but also as a tool of risk sustainability. There is a large body of literature that supports a positive relationship between CSR and risk sustainability. For example, Jo and Na (2012) found that higher CSR engagement is associated with lower firm risk. Similarly, Albuquerque et al. (2019) highlight that firms with stronger CSR performance enjoy higher firm valuation and reduced exposure to systematic risk. Further evidence indicates that CSR lowers stock price crash risk by encouraging greater transparency and limiting managerial incentives to withhold negative news. In other words, the empirical evidence supports the view that CSR predominantly serves a risk-mitigating function rather than an opportunistic managerial tool (Kim et al., 2014).

The relationship between CSR and financial performance is largely indirect (Surroca et al., 2010). Rather than generating immediate financial returns, CSR contributes to accumulation of intangible resources such as human capital development, reputational capital, and a strong organizational culture. These tangible assets enhance a firm’s competitive advantage and operational stability, particularly in growth-oriented industries. This perspective highlights CSR as strategic investment to improve financial performance while mitigating risk.

Financial stability is considered an important determinant of firm performance and sustainability (Poliakova et al., 2024). While CSR enhances sustainability, financial stability enables firms to invest in sustainability initiatives, with their effectiveness depending on organizational structure and capabilities. Complementing this view, Mahmood et al. (2024) suggest that incorporation of digital technologies further strengthens CSR and financial performance by improving transparency, operational efficiency, and resource allocation. Digital technologies as a moderating factor enhance the risk-mitigating and value-creating effects of CSR, thereby supporting long-term financial sustainability.

Despite the growing influence of digital technologies, Nazir et al. (2024) find that financial return associated with ESG (Environmental, Social and Governance) is not always positive and consistent. Particularly in high-tech industry, ESG performance is negatively related to financial sustainability. Individual ESG components produce mixed impacts. While environmental and social initiatives may create short-term financial pressure, governance improvement tends to improve financial performance. These results suggest that the ESG initiative does not guarantee value creation; instead, it depends on the specific sustainability initiatives undertaken and the industry settings in which it operates. Consequently, aggregated ESG measures may hide important underlying differences. This highlights the necessity for a more detailed and specific understanding of how sustainability components affect firms’ financial performance.

Lopatta et al. (2024) argue that the financial performance of CSR is dependent on the level of intensity of a firm’s engagement. Moderate investment in CSR results in improved productivity, whereas excessive engagement leads to lower financial returns. Investing in CSR beyond an optimal level leads to the decline in firms’ marginal benefit, highlighting that optimal level of investment in CSR can lead to financial efficacy. Sharma and Chakraborty (2024) advocate that CSR initiatives that are integrated with the core business strategies result in stronger and positive financial performance. In contrast, evidence from emerging markets suggests that CSR may impact financial performance negatively, particularly under mandatory settings where compliance driven initiatives exacerbate extra cost and shifts management focus away from core activities (Sharma & Chakraborty, 2024). The findings suggest that the performance outcome of CSR is not dependent on the level of spending, but rather on how well the initiatives are integrated into the overall business strategy. Complementing this view, Licandro et al. (2024) highlight stakeholder satisfaction as a mediating variable that impacts the relationship between CSR and financial performance. As discussed earlier in the findings of Surroca et al. (2010), CSR influences financial performance indirectly. CSR improves financial performance by strengthening the relationship with key stakeholders including shareholders, employees, customers and suppliers. Increased stakeholder satisfaction enhances stronger loyalty, increases productivity, and portrays a positive business image. Empirical evidence suggests that this mediating effect is not only confined to large publicly listed firms but also across firms of varying sizes and ownership structures. These findings suggest the premise that CSR influences financial sustainability in an indirect way.

In the context of the banking industry, Wu and Shen (2013) find a positive relation between CSR and financial performance. This relationship is prominent in developed countries due to stronger investor protection and greater stakeholder orientation that reinforces the effectiveness of CSR initiatives (Belasri et al., 2020). However, in developing countries, such relationships are not consistently observed where divergent economic structures and weaker institutional framework affect the relationship between CSR and financial performance (Blowfield & Frynas, 2005). In the context of developing countries like Bangladesh, the CSR expenditures, particularly of banks, have been on an upwards trend (Kabir & Chowdhury, 2023). This growing trend is driven by shifts in corporate culture and policy incentives, i.e., tax rebates for CSR spending. However, CSR expenditure does not necessarily influence financial performance (Kabir & Chowdhury, 2023).

In contrast, recent empirical evidence from Bangladesh suggests that CSR exhibits a positive and statistically significant association with Return on Equity (Hasan et al., 2025). Similarly, as discussed above, the literature widely supports a positive relationship between CSR and financial performance. However, contrasting evidence in the literature highlights the negative relationship between CSR and financial performance (Hirigoyen & Rehm, 2015). The primary reason for this negative relationship can be attributed to the additional cost of CSR activities which can constrain short-term profit maximization and financial return.

The existing literature, therefore, suggests that the relationship between CSR and firm outcomes is complex and mediated by institutional, strategic, and governance factors. However, in the context of commercial banks, this relationship may be particularly consequential. Unlike non-financial firms, banks operate in a highly regulated and trust-sensitive environment. In banks, reputational capital, stakeholder confidence, and risk management practices play a central role in sustaining financial stability. CSR engagement in banks may, therefore, serve as a mechanism for strengthening depositor confidence, improving borrower relationships, enhancing regulatory legitimacy, and reducing information asymmetry. Accordingly, we formulate the following hypotheses.

H1: 

CSR engagement is negatively associated with bank risk and positively associated with financial sustainability.

H2: 

The negative association between CSR engagement and bank risk is mediated by stakeholder confidence.

2.2. Green Banking and Financial and Risk Sustainability

Climate change and environmental degradation have become significant concerns in global economic governance. These concerns have led to growing expectations for financial institutions to contribute to sustainable development. Commercial banks do not directly create environmental pollution. However, their lending and investment decisions strongly influence the environmental impact of the real sector. Banks act as financial intermediaries in the economy. They can direct funds to environmentally responsible projects and limit financing for harmful activities (Porter & van der Linde, 1995; Ambec & Lanoie, 2008). When environmental initiatives are aligned with operational efficiency and market strategy, they can create a competitive advantage. In this case, they become a source of strength rather than a financial burden (Orsato, 2006). Investors usually react more positively when firms with strong environmental performance make environmental investments (Rahman et al., 2025). However, similar investments by high-polluting firms often fail to generate similar valuation benefits, as investors anticipate potential future environmental costs and liabilities (Clarkson et al., 2004). Diouf and Boiral (2017) suggest that sustainability disclosures can sometimes be used as impression management tools. Firms may highlight positive achievements and minimize weaker environmental performance. Moreover, high quality disclosure can reduce information asymmetry between firms and investors. This is especially important in environments where information gaps are large and investors rely heavily on private information (Brown & Hillegeist, 2007). Higher information asymmetry can increase risk premiums. This can raise the overall cost of capital and its direct impact may vary depending on market conditions (Hughes et al., 2007).

Consequently, the role of banks in promoting sustainability has attracted increasing academic and regulatory attention. Green banking has emerged as a strategic response to these expectations. According to Ratnasari et al. (2021), environmentally oriented daily banking activities are positively associated with profitability. Banks that actively integrate green practices into their routine operations appear to benefit financially. Stronger capital adequacy and higher liquidity levels are linked to better profit performance. This suggests that financial stability enhances banks’ ability to improve earnings. However, Siswanti et al. (2024) suggest that the relationship between green banking initiatives and financial performance is not always straightforward. They find no direct positive impact of green banking practices on profitability. Asset quality, particularly the level of non-performing loans, also plays a critical role in shaping financial outcomes (Zheng et al., 2022). Furthermore, ownership structure plays a key role in shaping how sustainability initiatives translate into financial outcomes.

Gulzar et al. (2024) highlight the importance of green banking practices in enhancing banks’ performance. Evidence suggests that environmentally oriented financing activities and operational initiatives can meaningfully contribute to sustainability outcomes. Interestingly, not all dimensions of green banking are not equally effective. Operational practices and green financing initiatives often have a more pronounced environmental impact compared to internal policies or symbolic actions. Overall, the success of green banking depends on how it is implemented. Simply having sustainability frameworks in place is not enough. Empirical evidence from other studies shows that long-term financial sustainability depends on more than environmental commitment. It is also influenced by internal efficiency and careful cost management (Afandi & Purnamasari, 2025).

Recent studies show that green finance plays an important role in linking green banking practices to overall sustainability performance. It acts as a key channel through which these practices create broader environmental and financial benefits. Internal factors such as strong commitment from top management, effective operational systems and clear policy support can improve the success of green finance initiatives. In contrast, customer-focused practices usually have a weaker direct impact (Kumar et al., 2024). Additionally, prior research suggests that green banking initiatives can enhance a bank’s goodwill. Investments in environmental sustainability and responsible financing can improve a firm’s reputation. They also build stakeholder trust which indirectly supports financial performance (Inegbedion, 2024).

In the context of developing countries like Bangladesh, green banking performance is positively related to financial performance (Bose et al., 2021). The improvement in profitability appears to be largely driven by the cost efficiency gains associated with green banking activities. In other words, environmentally focused practices improve financial results when they increase operational efficiency. However, the presence of corporate political connections alters this relationship (Uddin et al., 2022). Previous studies suggest that political connections can encourage firms to engage more in CSR activities (Zheng et al., 2022). However, evidence from the banking sector shows that investors may view green banking initiatives differently when banks have political ties (Zheng et al., 2022). In these situations, the positive link between green banking performance and financial performance becomes weaker. This may happen because stakeholders doubt the true motives behind the sustainability initiatives. One important limitation is that there may be endogeneity between green banking performance and financial outcomes. More profitable banks may have more resources to invest in sustainability initiatives. Another limitation is the use of disclosure-based measures to assess green banking performance. These measures may not fully reflect the actual environmental practices of banks.

Ambec and Lanoie (2008) suggested that well-designed environmental initiatives can improve efficiency and competitiveness, rather than merely imposing operational costs. Consistent with this view, prior studies indicate that environmentally oriented operations and green financing activities may contribute to stronger financial performance. However, the evidence remains inconclusive. In the context of Bangladesh, increasing regulatory attention on sustainability is expected to promote green banking practices which may result in enhanced financial sustainability. Accordingly, we formulate the following hypothesis.

H3: 

Green banking practices are positively associated with financial sustainability.

Green banking may also influence banks’ risk profiles. Banks can include environmental factors in their lending decisions. This can help them reduce exposure to risky borrowers. Improved disclosure quality can further reduce information asymmetry. This can lower risk premiums and reduce the cost of capital.

H4: 

Green banking practices are negatively associated with bank risk.

3. Methodology

In this study, we implemented a quantitative research design using panel data to assess the impact of Corporate Social Responsibility (CSR) and green banking practices (GBP) on the financial and risk sustainability of commercial banks in Bangladesh. This study employed random effects regression models to assess how CSR and green banking influence profitability, financial stability, and risk exposure. The random effects model is particularly ideal for banks where the data involves multiple entities, data variations within and across banks. This model enhances the generalizability and reliability of the results. We selected more than 300 observations, covering the period from 2013 to 2024. This period was selected because it coincides with the systematic integration of structured green banking and CSR reporting in the Bangladesh banking sector. Bangladesh Bank introduced Green Banking Policy Guidelines in 2011 (Bangladesh Bank, 2011). These guidelines require scheduled banks to disclose their green initiatives. Later in 2012, Bangladesh Bank provided a standardized reporting format.

To provide a comprehensive view of sustainability, we conceptualized the sustainability variable in two dimensions: financial sustainability and risk sustainability. Financial sustainability means a bank’s ability to generate stable and efficient returns over time. Risk sustainability refers to its ability to manage financial risks without harming long-term stability. Financial sustainability is evaluated using both profitability and stability indicators. Return on Equity (ROE) measures profitability and managerial efficiency. In addition, the Z score is used as a combined measure of financial stability. It includes profitability, capitalization and earnings volatility. Higher values show lower insolvency risk and stronger resilience.

Risk sustainability is assessed through multiple indicators—credit risk and liquidity risk, to ensure a comprehensive analysis. Credit risk is represented by the non-performing loan ratio. Liquidity risk is measured by the loan-to-deposit ratio. Employing multiple risk measures gives a clearer and more comprehensive understanding of bank risk. It helps show how CSR and green banking practices influence different dimensions of risk.

Corporate Social Responsibility is one of the main independent variables in this study. It is measured using the CSR expenditure reported in the annual reports of the banks, scaled by total assets. Green banking practices are measured using a green banking practices index. This index shows how much banks include environmental factors in their policies, financing activities, internal operations and reporting practices. The GBPI comprises 18 disclosure items organized into four dimensions: green policy, green financing activities, internal environmental management, and green reporting and risk disclosure. The green banking practices index is constructed using a four-point scoring system. A score of 0 is assigned when there is no disclosure. A score of 1 is given when only a generic statement is provided. A score of 2 indicates quantitative disclosure. A score of 3 is assigned when the disclosure includes quantitative information along with specific targets, progress and comparison. The index reflects compliance with the green banking guidelines issued by Bangladesh Bank. It also incorporates international green banking practices. In this way it captures both the regulatory and strategic dimensions of green banking.

The data on financial and risk-related variables are collected from the annual reports of the banks. Information on CSR and green banking practices is obtained from the descriptive sections of annual reports, sustainability reports and directors’ statements. Macroeconomic variables, including GDP growth and inflation, are obtained from the World Bank’s database.

The model estimates the relationship between CSR, green banking, and sustainability:

$${Sustainability}_t^i={\alpha }_i+{\beta }_1{CSR}_t^i+{\beta }_2{Controls}_t^i+{\epsilon }_t^i$$

(1)

$${Sustainability}_t^i={\alpha }_i+{\beta }_1{GBPI}_t^i+{\beta }_2{Controls}_t^i+{\epsilon }_t^i$$

(2)

Here, ${Sustainability}_t^i$ represents financial or risk sustainability. ${CSR}_t^i$ is measured using the actual CSR expenditure disclosed in the annual reports, scaled by total assets. This adjustment accounts for differences in firm size. It also ensures comparability across firms. ${GBPI}_t^i$ represents the green banking practices index. ${Controls}_t^i$ denotes a set of control variables—bank size, and macroeconomic variables commonly used in empirical banking research. Bank size is measured as the natural logarithm of total assets to account for bank size effects. ${\epsilon }_t^i$ captures the movements that are not explained by the equation.

To address potential endogeneity concerns arising from reverse causality and omitted variable bias, this study employs a dynamic panel data estimation technique using the System Generalized Method of Moments (System GMM). In the context of banking, financial performance and sustainability indicators may be influenced by their past values. Therefore, a lagged dependent variable (${Sustainability}_{t-1}^i$) is included in the model specification.

The empirical model is specified as follows:

$${Sustainability}_t^i={\alpha }_i+{\beta }_1{Sustainability}_{t-1}^i+{\beta }_2{CSR}_t^i+{\beta }_3{GBPI}_t^i+{\beta }_4{Controls}_t^i+{\epsilon }_t^i$$

(3)

CSR and green banking practices are treated as endogenous variables because of the potential for reverse causality, while macroeconomic variables are considered exogenous. To address these endogeneity concerns and enhance the robustness of the estimates, the study employs the System GMM approach.

To assess whether the relationship between CSR and bank risk is associated through indirect channels, we adopt a mediation framework. In commercial banks CSR may not affect financial risk in a direct way. Instead, its impact may operate through intermediate factors such as depositor confidence. Accordingly, the analysis follows a two-step panel regression approach to examine the mediating role.

First, the mediator is regressed on CSR while controlling relevant bank-specific characteristics.

$${Mediator}_t^i={\alpha }_i+{\beta }_1{CSR}_t^i+{\beta }_2{Controls}_t^i+{\epsilon }_t^i$$

(4)

Second, bank risk is regressed on both CSR and the mediator.

$${Risk}_t^i={\alpha }_i+{\beta }_1{CSR}_t^i+{\beta }_2{Mediator}_t^i+{\beta }_3{Controls}_t^i+{\epsilon }_t^i$$

(5)

Here, ${Mediator}_t^i$ represents the depositor’s confidence, which is measured using the depositor’s growth. ${Risk}_t^i$ refers to the NPL ratio. A significant coefficient on the mediator ${(\beta }_2)$ suggests that the mediator influences bank risk. If CSR significantly affects the mediator and the mediator significantly affects bank risk, we can confirm that mediation has occurred.

4. Findings

4.1. Summary Statistics

The descriptive statistics with the key variables used in this study are represented in

Table 1. Summary statistics.

	Mean	Min	Max	Std. Dev.	Skewness	Kurtosis
CSR	1.742 × 108	0	3,273,690,000	3.226 × 108	6.238	51.60
Loan-to-Deposit	0.899	0.108	1.166	0.127	−1.597	9.471
ROA	0.007	−0.06	0.023	0.007	−4.5	37.05
ROE	0.09	−0.7	0.26	0.104	−5.42	41.08
NPL Ratio	0.074	0	0.671	0.083	4.35	25.717
Total Assets	3.623 × 1011	5.966 × 109	2.475 × 1012	2.980 × 1011	3.764	22.465

The mean return on asset (ROA) for banks in Bangladesh is around 0.7% with the distribution being negatively skewed with high kurtosis. Similarly, the average Return on Equity (ROE) is 0.09, indicating that banks generate an average return of approximately 9%. The ROE ranges from −0.7 to 0.26, implying moderate variation in profitability across banks. We winsorized the ROE data to limit the effect of extreme observations.

. It is observed that the average CSR expenditure is around BDT174 million with relatively high standard deviation. It appears that a handful of banks in Bangladesh spend very little on CSR, while a few banks spend extremely large amounts. The mean loan-to-deposit ratio is 0.899 meaning that banks tend to lend about 89.9% of their total deposits. The standard deviation indicates relatively stable lending behavior with negative skewness.

According to the table, non-performing loans account for 7.4% of total loans on average. The maximum value of 0.671 suggests that certain banks experienced significant credit risk problems. The positive skewness and kurtosis indicate that most banks have relatively lower NPL levels, but a few banks report very high default rates. Finally, Total Assets have a large standard deviation. That indicates a significant difference in bank size within the sample.

4.2. Impact of CSR and Green Banking on Sustainability

To examine the impact of CSR on financial and risk sustainability, we use random effect panel regression analysis, where ROE and Z scores are used as measures for financial sustainability. In addition, NPL ratio and loan-to-deposit ratio are used as proxies for risk sustainability to understand the dynamic effect of CSR.

Table 2. Impact of CSR and green banking on sustainability.

	Financial Sustainability				Risk Sustainability
	ROE		Z Score		NPL		Loan-to-Deposit
CSR	60.9 ***		−1076		−52.03 ***		40.83 **
	(23.32)		(873)		(14.41)		(19.72)
Green Banking		0.003 ***		−0.032		−0.001 *		0.0003
		(0.0009)		(0.037)		(0.001)		(0.001)
Bank Size	−0.016	−0.02	−9.95 ***	−9.747 ***	0.026 ***	0.019 *	0.096 ***	0.063 ***
	(0.017)	(0.017)	(0.75)	(0.831)	(0.009)	(0.011)	(0.014)	(0.015)
GDP Growth	0.006	0.004	−0.236	−0.214	−0.004	−0.002	0.013 ***	0.011 **
	(0.005)	(0.005)	(0.209)	(0.205)	(0.004)	(0.004)	(0.005)	(0.005)
Inflation	−0.009 *	−0.012 **	0.258	0.319 *	0.011 ***	0.014 ***	−0.001	0.001
	(0.0047)	(0.005)	(0.173)	(0.173)	(0.003)	(0.003)	(0.004)	(0.004)
Constant	0.52	0.64	288.67 ***	283.09 ***	−0.593 **	−0.482 *	−1.73 ***	−0.843 **
	(0.45)	(0.43)	(19.59)	(21.484)	(0.252)	(0.276)	(0.373)	(0.378)

*** p < 0.01, ** p < 0.05, * p < 0.1, Standard errors are in parentheses.

shows the findings. CSR is found to have significant and positive effects on ROE. This finding is consistent with those of Herremans et al. (1993) and Albuquerque et al. (2019). It supports the notion that corporate socially responsible activities enhance reputation, stakeholder trust, and long-term financial performance. However, CSR, although negatively related to Z score, is not significant. This implies that CSR activities do not seem to reduce the risk level and overall financial stability of banks. This finding is consistent with that of Poliakova et al. (2024), who suggest that CSR alone may not be sufficient to strengthen long-term financial stability unless supported by other internal factors.

When we test on risk sustainability, CSR shows a significant and yet negative relationship with NPL ratio, meaning higher CSR engagement is associated with lower non-performing loans. Banks that are more socially responsible tend to manage credit risk more effectively in the case of Bangladesh. This finding also suggests that CSR is likely to improve monitoring quality, borrower relationships, and overall credit discipline. These results strongly support the idea that CSR helps reduce risk consistent with the findings of Kim et al. (2014), who explain that CSR lowers risk by improving transparency and limiting managers from acting opportunistically.

When we measuring loan-to-deposit ratio, we find that CSR has a positive and significant effect, implying that the CSR-active banks tend to extend slightly more loans relative to deposits. Overall, our results indicate that CSR can improve asset quality (lower NPLs) but with slightly increased lending intensity.

Next, we investigate the impact of green banking practices on both financial and risk sustainability issues. Again, financial sustainability is measured by ROE and Z-score, and risk sustainability is proxied by NPL ratio and loan-to-deposit ratio. We find that green banking has a positive and statistically significant impact on ROE. This positive relation suggests improved financial performance is achievable through increased environmentally involved banking activities, consistent with Bose et al. (2021), who articulate that green banking performance is positively related to financial outcomes.

We also observe that green banking does not have a statistically significant impact on Z-score, implying that green banking practices do not directly influence overall financial stability, consistent with Siswanti et al. (2024), who also demonstrate that green banking initiatives do not always result in immediate improvements in financial stability. When analyzing risk sustainability, we find that it has a negative and statistically significant impact on NPL ratio which is in line with the argument that green finance reduces exposure to risky borrowers by incorporating environmental screening into lending decisions.

To address potential endogeneity concerns arising from reverse causality and omitted variable bias, we employed the System GMM estimator.

Table 3. Dynamic Impact of CSR and green banking on financial and risk sustainability (System GMM estimation).

	Financial Sustainability		Risk Sustainability
	ROE	Z Score	NPL	Loan-to-Deposit
ROEt−1	0.81 ***
	(0.12)
Z scoret−1		1.01 ***
		(0.06)
NPLt−1			1.072 ***
			(0.21)
Loan to depositt−1				0.61 ***
				(0.13)
CSR	49.21 *	−978.58	−81.01 ***	16.47
	(27.24)	(1036.99)	(22.29)	(32.70)
Green Banking	−0.0013	0.01	0.0013	−0.001
	(0.0013)	(0.05)	(0.001)	(0.002)
Bank Size	0.014	0.91	0.043 ***	0.0065
	(0.017)	(0.72)	(0.01)	(0.02)
Constant	0.38	−25.13	−1.12 ***	0.22
	(0.44)	(19.84)	(0.36)	(0.56)

*** p < 0.01, * p < 0.1.

presents the dynamic panel results and provides a robustness check against the baseline random effects estimates reported in Table 2.

As expected in a dynamic model, the lagged dependent variables remain positive and strongly significant across all specifications. This justifies the use of a dynamic modeling framework and indicates that past performance plays a crucial role in shaping current bank outcomes.

The GMM results indicate that CSR continues to have a positive and statistically significant effect on ROE, which supports the conclusion that CSR engagement enhances profitability and strengthens shareholder returns. However, consistent with the random-effects model, CSR remains statistically insignificant for the Z-score. These findings suggest that while CSR supports financial performance, it does not directly contribute to overall financial stability. In terms of risk sustainability, CSR maintains a negative and highly significant association with the NPL ratio. The stronger magnitude in the GMM model implies that once endogeneity is controlled for, CSR’s risk-reducing effect becomes even more pronounced. On the contrary, CSR becomes insignificant for the loan-to-deposit ratio, suggesting that the earlier positive effect was driven by endogeneity.

The results for green banking change noticeably under the GMM model. In Table 2, green banking had a positive effect on ROE and helped reduce NPL. But in the GMM results, green banking becomes insignificant in every model. This means that once endogeneity is addressed, green banking does not directly affect financial performance or risk. This means the earlier positive results may not reflect a real effect. Stronger banks simply have more resources to spend on green initiatives. This suggests that green banking may not be the actual driver of better performance.

4.3. Mediation Analysis of the CSR–Risk Relationship

Next, we investigated whether CSR affects bank risk through an indirect channel—depositor confidence. We conducted a mediation analysis using a two-step panel regression approach (Equations (4) and (5)). In this analysis, depositor confidence is proxied by deposit growth and bank risk is measured using the NPL ratio. In the first step, we regress depositor confidence on CSR along with all control variables. We find that CSR does not have a statistically significant impact on deposit growth. This suggests that CSR does not influence depositor behavior. In this context, deposit decisions are likely driven by other factors. Finally, in the second step, we regress bank risk (NPL ratio) on both CSR and deposit growth. We find that CSR remains negative and statistically significant. Deposit growth shows a negative and statistically significant effect on NPL, indicating that higher deposits are associated with lower credit risk.

As shown is

Table 4. Mediation analysis of the CSR–risk relationship.

	Mediator	Risk Sustainability
	Deposit Growth	NPL Ratio
CSR	−50.84	−58.58 ***
	(36.91)	(14.57)
Deposit Growth		−0.07 ***
		(0.022)
Bank Size	−0.12 ***	0.013
	(0.021)	(0.011)
GDP Growth	0.0006	−0.004
	(0.01)	(0.003)
Inflation	0.016 *	0.013 ***
	(0.008)	(0.003)
Constant	3.24 ***	−0.292
	(0.57)	(0.298)

*** p < 0.01, * p < 0.1, Standard errors are in parentheses.

, the results do not support a mediation effect. For mediation to exist, CSR must significantly affect the mediator. In addition, the direct effect of CSR on NPL remains strong even after including the mediator. This indicates that CSR reduces credit risk through a direct channel instead of through depositor confidence.

We conclude that socially responsible activities do not directly affect depositor trust as much as deposit decisions are influenced more strongly by other factors such as interest rates, service quality, or macroeconomic conditions rather than CSR initiatives. Overall, our mediation results from the banking sector in Bangladesh do not support the findings of Licandro et al. (2024). Therefore, the empirical evidence fails to support Hypothesis 2.

5. Conclusions

In this study, we examined the impact of Corporate Social Responsibility (CSR) and green banking practices on the financial and risk sustainability of commercial banks in Bangladesh. Using panel data, with 300 observations over the period ranging from 2013 to 2024, we employ random effects regression models to assess whether CSR and green banking influence profitability, financial stability, and risk exposure in the case of Bangladesh. In addition, we extend the analysis by incorporating a mediation framework to investigate whether CSR affects bank risk indirectly through depositor confidence. To address potential endogeneity, we further extend the analysis by using a System GMM approach as a robustness check.

Our empirical findings indicate that CSR engagement contributes positively to financial performance, as measured by Return on Equity (ROE). Banks with higher CSR expenditure tend to generate superior shareholder returns. Importantly, this positive relationship remains robust under the GMM estimation. CSR, however, does not show a statistically significant impact on the Z-score implying that CSR does not directly enhance overall financial stability.

When testing for risk sustainability, we find that CSR demonstrates a significant negative association with the non-performing loan ratio. We infer that those banks that engage in CSR activities tend to manage credit risk more effectively. This negative relationship becomes even stronger under the GMM framework. It emphasizes the role of CSR as a reliable risk-mitigating mechanism. However, unlike the baseline results, CSR does not demonstrate a statistically significant relationship with the loan-to-deposit ratio after controlling for endogeneity.

Our research also examines the green banking practices and finds that they exhibit a positive and statistically significant relationship with ROE. Similar to CSR, we also observe that green banking does not significantly influence financial stability as captured by the Z-score. More importantly, we observe a negative relationship between green banking practices and NPL ratio. However, the GMM results present a notably different perspective. After accounting for endogeneity, relationships are no longer statistically significant. This means green banking does not show a strong or reliable impact on financial performance or risk sustainability. The previous positive results may not reflect a true effect. They may simply indicate that more profitable banks have more money to invest in green initiatives.

Thus, it is clear that in the case of the Bangladesh banking sector, CSR consistently enhances profitability and reduces credit risk. But the impact of green banking practices appears sensitive to model specification and endogeneity concerns. The findings contradict prior evidence from Bangladesh. In the Bangladeshi setting, Islam and Faruquee (2022) state that green banking investment positively influences financial performance, which is consistent with our baseline results. Similarly, Gazi et al. (2024) highlight that green banking practices, daily operations, and green policies play a more substantial role in shaping sustainability outcomes. Then again, our dynamic analysis suggests that these effects may not be causal when endogeneity is properly addressed. Prior studies rely on perceptual sustainability measures and green CSR outcomes based on survey data. In contrast, our study provides objective data-driven evidence by linking sustainability practices to financial and risk indicators. Finally, our research contributes to the growing literature on sustainability in the banking sector by providing new evidence from the emerging market context. We conclude that CSR practices enhance profitability and reduce credit risk.

This study has a limitation that should be acknowledged. The model does not incorporate certain bank-specific financial control variables such as Tier 1 capital and cost-to-income. Data for these indicators was not available across all sampled banks for the study period, which limited their inclusion in the empirical model. Despite this limitation, the study controls for key bank-specific and macroeconomic factors.

As for the future direction of research, we propose that incorporating an independent board of directors, age of CEOs, and gender of board of directors and CEOs can be used as control variables to further investigate whether CSR activities improve financial performance and risk stability in any emerging markets. Furthermore, the inclusion of key financial controls such as Tier 1 capital and cost-to-income would allow for a more comprehensive assessment.