2.1. The relationship between CSR and core competence

There are several explanations for why CSR efforts positively influence core competencies. According to stakeholder theory, it is assumed that profitability and engaging in CSR activities are not mutually exclusive and that acting in the best interests of all stakeholders can ultimately lead to improved organizational performance [14]. Firms that engage in socially responsible activities may gain a competitive advantage by attracting more investors and customers while reducing operating costs [15]. Additionally, Belasri et al. [16] argue that bank performance and CSR are positively associated. They claim that CSR initiatives can help organizations build a strong reputation, which brings several benefits, such as a greater ability to attract and retain high-quality employees. Furthermore, Wu and Shen [4] also found that CSR positively impacts banks’ financial performance (FP). They argue that when banks engage in large-scale CSR, customers may be more willing to accept lower interest rates on their deposits; from the banks’ perspective, lower-cost deposits represent lower-cost funding sources. Outstanding CSR performance and the resulting reputation enhancement can foster customer loyalty and attract clients from competitors, thereby increasing market share [16]. Kim et al. [17] also found that CSR efforts improve a firm’s financial performance, particularly when competition among firms is intense. Therefore, banks engaging in CSR activities may derive various benefits, including increased appeal to customers and investors, reduced cost of capital, enhanced brand value, and improved financial outcomes [15].

In contrast, neoclassical economic theory argues that funds allocated to CSR are inefficient. They should be redirected toward initiatives that maximize corporate value, and banks’ performance may decline due to CSR engagement [16,18,19]. Zhou et al. [7] argue that banks incur specific costs when fulfilling their social responsibilities due to short-term resource limitations. Banks that overinvest in CSR programs may undermine their core operations. Therefore, firms with limited resources tend to allocate their funds more prudently, investing only in CSR initiatives that are deemed essential and serve to reduce internal and external conflicts [1]. In addition, CSR can be beneficial, particularly for organizations operating under resource constraints. Nguyen and Nguyen [1] found that banks with abundant resources may engage in excessive CSR spending or treat CSR participation as more discretionary, which could increase the risk of failure rather than ease stakeholder tensions.

Previous studies examining the relationship between CSR participation and banks’ core competencies have focused on developed countries such as China and the United Kingdom. Our study selects a sample of commercial banks in Vietnam because the banking sector is pivotal in economic development and financial reform. Furthermore, the institutional environment in Vietnam is still evolving, with many banks facing challenges related to transparency, risk management, and reputation building. Proactive CSR implementation enhances trust among customers, investors, and shareholders and strengthens internal capabilities such as human resource management, operational efficiency, and innovation capacity. Therefore, we propose the following hypothesis:

2.2. The relationship between bank complexity and core competencies

The conglomerate hypothesis argues that diversification of banking activities ensures that managerial efforts are optimized across different areas of banking operations [20]. This trend leads to economies of scope by sharing fixed costs across various products and reducing bank earnings volatility. Specifically, by offering a broad range of products and services, banks can utilize their existing infrastructure, resources, and customer base more efficiently. This approach can result in cost savings, improved operational efficiency, and enhanced profitability. In addition, Luu et al. [5] found that banks’ operational performance and income diversification are positively correlated. They argue that adopting this diversification strategy can generate substantial benefits for banks, including increased profitability, economies of scope, reduced monitoring costs through shared fixed expenses, better utilization of managerial capabilities, risk mitigation via expansion into multiple products and regions, enhanced franchise value and market resilience, and improved ability to build competitive advantage when entering new markets.

In contrast, agency theory [21,22] suggests that firms should concentrate on their core activities to minimize potential agency problems. It contends that diversification may lead managers to expand their firms beyond an optimal scale, undermining firm value. Moreover, diversification may not always be beneficial, as its associated costs can sometimes outweigh its advantages. When bank managers operate outside their core competencies and areas of expertise, it can lead to negative consequences such as increased information asymmetry, inefficient allocation of resources, higher agency costs, intensified competition, and a decline in competitive advantage [9]. Due to the inherent uncertainty of non-traditional banking activities, income diversification may increase the overall risk profile of banks. DeYoung and Roland [8] explained that weaker customer-bank relationships in non-lending sectors are a key reason why revenues from non-traditional businesses tend to be more volatile than those from conventional banking services. Additionally, engaging in non-interest activities may require substantial investments in technology and human capital, leading to significantly higher operational costs and increased earnings volatility.

Previous studies on the relationship between income diversification and banks’ core competencies have focused on developed countries like the United States. Our study selected a data sample from a developing country like Vietnam because its banking sector is currently undergoing a transitional phase and is deeply influenced by ongoing economic restructuring. In the context of rising competition and shrinking profit margins from traditional credit activities, banks are increasingly compelled to seek alternative revenue streams from services, investments, and retail finance. Therefore, banks must enhance their core competencies to mitigate operational risks and avoid losing competitive advantage, especially in an institutional environment that is still evolving, as is the case in Vietnam. Therefore, we propose the following hypothesis:

2.3. The buffering role of CSR on the relationship between bank complexity and core competencies

The conglomerate hypothesis demonstrates diversification’s benefits but highlights its risks, particularly when increasing complexity leads to ineffective management. The study by Luu et al. [5] expands upon the hypothesis by incorporating CSR as a buffering mechanism that helps mitigate the adverse effects of complexity on core competencies. Furthermore, Shao et al. [6] found that banks diversify their operations more extensively under the assumption that CSR functions as an implicit form of risk mitigation. They also found evidence that banks can reduce risk, enhance stability, and improve asset quality by broadening their activities and engaging in CSR initiatives. CSR activities may include investments in socially responsible projects, adoption of sustainable business practices, or initiatives that address social and environmental challenges. Through these efforts, banks can strengthen their reputation, foster stronger relationships with stakeholders, and lower their exposure to risks associated with non-compliance, reputational harm, or mismanagement. This approach, in turn, can contribute to improved asset quality as the bank’s overall risk profile becomes more balanced and sustainable [8].

Conversely, AlKhouri and Arouri [9] suggest that engaging in CSR amid growing bank complexity involves costs that may outweigh the benefits, primarily when executives must operate outside their core competencies and areas of expertise. Chiaramonte et al. [23] also provide evidence that diversification is not universally advantageous for banks, particularly when it involves unfamiliar industries or regions. Additionally, diversification can lead to greater risk and reduced stability if the associated costs surpass the potential gains [24]. Moreover, Lahouel et al. [25] demonstrate that overinvestment can negatively and significantly affect banks’ financial health, regardless of the income diversification measure used. Banks with high CSR performance and ambitious income diversification strategies may incur additional costs, resulting in reduced investment in training and development necessary to strengthen core competencies [24,26].

Previous studies exploring the relationship between income diversification, CSR participation, and the core competencies of banks have primarily been conducted in major economies such as the United States and China. Our study selected a data sample from a developing country like Vietnam because it is a transitional economy where the banking sector must expand its operations while simultaneously meeting sustainable development demands and improved transparency. In the context of shrinking profits from traditional credit activities, Vietnamese banks are increasingly diversifying their income sources to support growth, yet this also heightens complexity and operational risk. Effective risk control and maintaining core competencies remain significant challenges as the legal framework evolves. In such an environment, CSR is a complementary mechanism to build reputation, mitigate risks, and enhance internal cohesion. Therefore, we propose the following hypothesis:

3.1. Data

The study collected data from 26 commercial banks in Vietnam from 2010 to 2023. All the commercial banks’ financial statements are among the trustworthy databases. Having been established as a policy bank entirely funded by state funds, the analysis excludes Agribank following Duong et al. [3]. The study sampling period is from 2010 to guarantee no effects on the estimations from the global financial crisis of 2008. We winsorize the sample at the 1^st and 99^th percentiles to exclude outlier issues and follow Duong et al. [3] and Tran et al. [27] to remove observations for which the data is insufficient to calculate the necessary variables. The final sample is an unbalanced panel of 358 annual observations from 26 commercial banks in Vietnam.

3.2. Variable definitions

3.2.3. Bank complexity.

While the prior study by Duong et al. [3] only focused on CEO power and green credit, this paper follows Vuong and Nguyen [33] to introduce the concept of bank complexity. This proxy is measured through income diversification to represent the extent to which banks expand their operations beyond traditional sources of income. This approach emphasizes that bank complexity is an important mediating factor affecting the relationship between CSR and core competencies, which helps to expand the analysis compared to the previous papers that only focused on the direct impact of CSR. In addition, while the study by Le-Bao [34] mainly focused on the effects of diversification strategy on bank technical efficiency through a DEA model, this study inherits and extends it in several important directions. First, we shift the focus from technical efficiency to core competence, a more comprehensive indicator reflecting banks’ long-term performance and competitiveness. Second, instead of only considering the role of ownership and bank age, we include corporate social responsibility (CSR) as a strategic moderating mechanism in the relationship between income diversification and core competence. Overall, these extensions allow us to provide a more comprehensive and policy-relevant perspective on how CSR and diversification strategy jointly shape the long-term competitiveness of banks.

We use income diversification (ID) to measure bank complexity. We follow Nguyen et al. [2] and Luu et al. [5] to take one minus HHI. The Herfindahl-Hirschman index, or HHI of income specialization, is determined by splitting total operational income into its non-interest and net interest income components.

Net interest income is represented by the acronym NII, non-interest income by NONII, and net operational income by NOI. As the HHI increases, the bank gets less varied and more consolidated [35]. HHI has a range of 0–1. In the end, the study arrived at the following income diversification calculation:

This measurement method has more clearly reflected the bank’s income structure change and the potential risks when banks increase their diversification. The higher ID value indicates that banks have higher diversification banking activities. The advantage of this method is that it can be measured comprehensively by reflecting the bank’s income structure in Net Interest Income and Non-Interest Income activities.

3.3. Model construction

According to the stakeholder theory, Belasri et al. [16] and Wu and Shen [4] discovered a favorable correlation between banks’ performance and their implementation of CSR. In contrast, Nguyen and Nguyen [2] and Zhou et al. [7] argued that CSR endeavors can harm banks’ performance in line with neoclassical economic theory. Therefore, this study builds Model (1) to test hypothesis 1 to see how CSR effectiveness affects the core competencies of banks in Vietnam and test stakeholder theory and neoclassical economic theory.

(1)

Regarding bank complexity, Luu et al. [5] found evidence that the conglomeration hypothesis explains why income diversity enhances bank performance. However, due to the agency theory, Ammar and Boughrara [22] and Nguyen [36] argued that complexity negatively influences bank performance. Thus, this study constructed Model (2) to examine how income diversification (ID) affects core competence.

(2)

In addition, this study constructed the 3^rd Model to determine whether banks’ core competencies are impacted by CSR performance and bank complexity in the same Model.

(3)

To further test for stakeholder theory and the buffering role of CSR on the relationship between income diversification and core competence of Vietnamese commercial banks, we add the interaction variable (CSR*ID) to the regression model (4), following the results of Harjoto and Laksmana [24], Shao et al. [6], AlKhouri and Arouri [9], and Luu et al. [5].

(4)

Where “i” is cross-sections, “t” is time, and “α” is the intercept. In addition, CSR represents the corporate social responsibility index; ID is income diversification; CC represents core competence. Control variables include the non-performing loans (NPL), bank size (SIZE), bank tangible assets (TANG), and return on assets ratio (ROA). α_i is the firm fixed effect, and α_t is the year fixed effect. _it is the residual value. S2 Appendix has definitions for all variables.

3.4. Estimation methodology

The study used the Hausman and Redundant tests to determine which of the three estimate techniques – Pooled Ordinary Least Squares (OLS), Fixed Effects Model (FEM), and Random Effects Model (REM) – was the most suitable. Ordinary Least Squares (OLS) estimates linear regression parameters by minimizing squared prediction errors. It is efficient and straightforward under classical assumptions but can be violated by multicollinearity, heteroscedasticity, or endogeneity problems. FEM controls unobserved heterogeneity by allowing individual-specific intercepts, effectively reducing omitted variable bias, but it can be less efficient and cannot estimate time-invariant variables. The REM postulates that each effect is independent of the regressors and random, providing more efficient estimates if assumptions hold, but can be biased if the effects are correlated with any regressors. However, according to Greene [37], endogeneity issues are violated by OLS, FEM, and REM. As a result, we employ the Durbin-Wu-Hausman test by Duong et al. [3] to check for endogenous problems. Lastly, the study applies the Two-Stage Least Squares (2SLS) approach following Fan et al. [10] and Bawono and Handika [11] to address the endogenous variables. According to Wang et al. [38], the Two-Stage Least Squares (2SLS) estimate approach handles endogeneity in regression models by using instrumental variables. In parallel with using the primary method of 2SLS to overcome the endogeneity problem, we have combined it with fixed effects to eliminate estimation bias due to unobserved factors related to the independent and dependent variables [39]. Combining GLS with cross-section weights ensures that units with high variability do not distort the regression results, improving heteroscedasticity [40]. On the other hand, in panel data, changes across periods can cause serial correlation or time-varying variance. Period covariance weights help mitigate this phenomenon, ensuring more accurate estimates of standard errors [41].

Furthermore, the research carried out two robustness tests to validate the consistency of our primary empirical outcomes. First, we perform a robustness test by splitting our data into two subgroups: small banks and major banks, based on the State Bank of Vietnam’s Circular 52/2018/TT-NHNN. The second robustness test re-estimates the main regression result using the Panel-Corrected Standard Errors (PCSE) and the Robust Least Squares (RLS) method.

4.1. Descriptive statistics

Descriptive statistics help researchers understand the general characteristics of the data, thereby identifying the appropriate direction to address the research question. Descriptive statistics not only assist in verifying the reliability of the data but also provide a solid foundation for subsequent quantitative analyses, ensuring coherence and logical consistency throughout the research process. Table 1 illustrates the descriptive statistics of the sample. It shows a standard deviation of roughly 1.0613 and the 0.8763 mean value of CC. The average CSR is 3.1704, similar to findings from the study [29] in Vietnam. The ID’s average value of approximately 0.4053 aligns with the average ID of Vietnamese commercial banks, which was 0.201 for 11 years from 2007 to 2017 [5]. This result means that commercial banks in Vietnam have increased their income diversification in recent years. Additionally, compared to the countries of the Gulf Cooperation Council, Vietnamese commercial banks have a lower ID than [9,42] and higher than that of all 14 economies in the Asia Pacific [43]. Table 1 shows that the non-performing loans (NPL) have a 0.0201 mean value and a 0.0135 standard deviation value. The descriptive statistics of other variables, including the bank size (SIZE), tangible assets (TANG), and return on assets (ROA) of the banks, are also shown in this table. Figs 1–3 report the CC, CSR, and ID trends of commercial banks in Vietnam during the sampling period.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Descriptive statistics.

https://doi.org/10.1371/journal.pone.0330723.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. The Mean and ±1 Standard Deviation of CC of commercial banks in Vietnam (source: author’s calculations).

https://doi.org/10.1371/journal.pone.0330723.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. The Mean and ±1 Standard Deviation of CSR of commercial banks in Vietnam (source: author’s calculations).

https://doi.org/10.1371/journal.pone.0330723.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. The Mean and ±1 Standard Deviation of ID of commercial banks in Vietnam (source: author’s calculations).

https://doi.org/10.1371/journal.pone.0330723.g003

4.2. Pearson correlation matrix

The independent variables’ correlations are displayed in Table 2. Notably, CSR exhibits a strong positive correlation with TANG, presented by a coefficient equal to 0.2902, and ROA, which is presented by a coefficient equal to 0.2401, indicating a significant relationship with these factors. Consequently, we computed the variance inflation coefficient (VIF) and discovered that multicollinearity issues do not exist because all VIF values are less than five [27].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Pearson correlation matrix.

https://doi.org/10.1371/journal.pone.0330723.t002

4.3. Estimation results

Based on the results in Table 3, the REM method was selected for Model (1) based on the Hausman test results, with Prob (Hausman Test) equal to 0.471 (less than 0.05), not enough basis to reject the hypothesis H0, showing that REM is more suitable than FEM. On the contrary, Models (2), (3), and (4) have Prob (Hausman Test) of 0.028, 0.001, and 0.001 (less than 0.05), respectively, rejecting the hypothesis H0, so FEM was selected. At the same time, the Redundant Fixed Effect Test in all models has Prob less than 0.001, confirming that FEM is more suitable than the pooled Model.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Regression results from the REM and FEM methods.

https://doi.org/10.1371/journal.pone.0330723.t003

In terms of the Model’s explanatory power, R-squared shows that FEM is superior to REM. Specifically, the R-squared increased from 0.6352 in Model (2) to 0.7405 in Model (4), and the Adjusted R-squared also improved from 0.5854 to 0.7030, demonstrating the effectiveness of adding the interaction variable CSR*ID. The F-statistic values of all models have Prob (F-statistic) less than 0.001, indicating that the models are overall statistically significant. FEM, especially Model (4), is not only selected through appropriate tests but also has superior explanatory power through the R-squared and F-statistic indices.

After applying FEM and REM method to analyze the effect of ID and CSR on CC, the study uses the Durbin-Wu-Hausman test, as described by Duong et al. [29], to determine whether endogenous variables exist. Model (5) is initially built as the endogenous examination test model:

(5)

As an alternative variable of CSR, the CSR residuals (Res-CSR) obtained in the first step are included in the model (3) [3]. An endogeneity issue exists if the residual’s coefficient is a statistically significant variable. We also test for ID, NPL, SIZE, TANG, and ROA.

Table 3 shows that the probability of the Wald test is less than 0.05, indicating a heteroscedasticity problem. Table 4 shows that endogenous variables exist in the models, so we follow Fan et al. [10], Bawono and Handika [11] to perform the 2SLS combined with cross-sectional weights to ensure that highly volatile units do not bias the regression results, improving the problem of heteroscedasticity [40]. At the same time, Table 3 detected autocorrelation issues, so we added period covariance weights to help adjust the Model to address this phenomenon, ensuring more accurate estimates of standard errors [41]. The 2SLS estimations are reported in Table 5.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Coefficient of residual variables.

https://doi.org/10.1371/journal.pone.0330723.t004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Regression results from the 2SLS.

https://doi.org/10.1371/journal.pone.0330723.t005

4.4. Robustness tests

Nguyen and Nguyen [1] argued that firms with limited resources allocate their funds more prudently, investing just in CSR initiatives that are deemed vital and serving as conflict-reduction tools. At the same time, financially unconstrained banks may overinvest in CSR initiatives, harming the bank’s financial performance. In addition, Ammar and Boughrara [22] stated that more prominent banks frequently take an aggressive approach to diversification, which causes problems with their managers’ agencies. Therefore, Model (4) was used in the research to investigate the robustness of our primary findings using subsamples of varying bank sizes found in Table 6. According to Circular 52/2018/TT-NHNN issued by the State Bank of Vietnam, commercial banks are classified into two primary categories based on their total assets. These categories encompass large banks with total assets exceeding VND 100,000 billion and small banks with total assets of VND 100,000 billion or less.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Regression results from across bank size subsamples.

https://doi.org/10.1371/journal.pone.0330723.t006

Then, the study implemented the Two-Stage Least Squares (2SLS) method with different bank sizes to verify the stability of our main findings through model 4. When comparing the factors affecting a bank’s core competence, one can see that its scale affects its financial capacity, operational network, and business strategy. Large banks often have more substantial resources and investment capabilities, helping them develop core competencies more effectively. Classifying by size helps us know how banks of different sizes will be affected by various factors [12].

Finally, we estimate Model 4 results by using three different regression methods, including the Two-Stage Least Squares (2SLS), the Panel-Corrected Standard Errors (PCSE), and the Robust Least Squares (RLS) to affirm the consistency of our main findings. Alfadli et al. [44] suggested that the PCSE method is used to analyze models with balanced panel data, helping to control serial correlation and cross-sectional dependence between variables. Thus, it generates more accurate results on the impact of industry-specific, macroeconomic, and bank-specific elements on bank performance. Costa et al. [45] suggested that the RLS method solves the problem of data uncertainty and lack of reliability in prediction and improves the accuracy and robustness of semi-supervised image classification. Furthermore, Wang et al. [38] claim that this method simultaneously minimizes the influence of noise and uneven data structure. It also optimizes the parameter estimation process and enhances classification capabilities.

6.1. Conclusion

This study investigates how CSR and bank complexity develop the core competence of commercial banks in Vietnam. The study employs the 2SLS approach to examine a sample from 26 commercial banks in Vietnam from 2010 to 2023. This study yields some noteworthy findings. The findings indicate that the bank’s core competency is improved by strengthening its CSR efforts. Secondly, commercial banks that are more complex in income diversification increase their core competence. Third, by developing CSR activities and implementing an income diversification strategy, commercial banks can strengthen their core competence by reducing the potential risks of aggressively engaging in diversification strategies, such as agency problems, since CSR is an implicit hedging strategy and a control mechanism. These findings are robust across size subsamples. The findings remain robust even if we employ alternative estimation methods such as PCSE and RLS. Stakeholder theory, the conglomeration hypothesis, and earlier research align with the study, even if it is inconsistent with agency and neoclassical economic theories.

6.2. Implication

The empirical findings support the bank managers and policymakers in emerging markets in developing sustainable banking performance in the following ways. First, we will make recommendations to stakeholders based on our research results. Shareholders should consider income diversification and CSR initiatives as value-added strategies because income diversification helps stabilize customer profits, while CSR helps enhance reputation and long-term sustainability, benefiting shareholders. Customers will often expect banks to participate in socially conscious activities. Banks providing more diverse and stable services through income diversification and CSR will give customers a positive view of more stable and customer-oriented services. Employees should see banks’ CSR efforts as a positive signal of ethical and responsible business because income diversification helps build a more stable and safe working environment for employees.

Second, we will make some recommendations to policymakers. In encouraging and monitoring CSR and income diversification, they should develop and strengthen the legal framework to promote and monitor the implementation of CSR and income diversification of banks, which will help create a more sustainable and responsible banking industry. In the current context in Vietnam, CSR regulations are only general principles, and further assessment of legal feasibility is needed. Coordination between ministries and sectors to build a specific legal framework on CSR in the banking sector is extremely necessary. This policy relates to the balance of risks in income diversification; the authorities should issue guidelines to ensure the balance between income diversification and risk control and equip banks with prudent risk management measures to avoid excessive risk-taking. They can issue policies to encourage banks to implement CSR, develop preferential policies such as tax exemptions and reductions for banks that effectively implement CSR projects, or provide financial support or preferential loan programs to promote CSR activities, create a healthy competitive environment, and encourage responsible and diversified banking operations. Policymakers need to improve the legal framework on CSR in Vietnam further, develop specific regulations on the minimum budget ratio for CSR, and develop clear criteria to evaluate CSR effectiveness in the banking industry. In addition, banks need to establish clear risk management plans when expanding into new business areas, especially for high-risk industries such as financial investment and insurance. The role of the State Bank is also vital in implementing this process; they need to play a coordinating and technical support role for banks to implement CSR effectively and sustainably, as well as coordinate with ministries and branches to ensure the legality and feasibility of applying CSR.

Finally, the responsibility belongs to the leading players, such as commercial banks; we suggest that they adjust policies based on the size and type of bank. Applying CSR at the strategic level is essential for large banks to manage risks in diverse and complex operations while creating long-term competitive advantages. These banks need to integrate CSR into their overall business strategy, focusing on projects with significant impact, such as environmental protection, community development, and green technology investment. CSR should be used as a risk management tool, especially in complex areas such as financial investment, insurance, and international banking services. Meanwhile, small banks should prioritize CSR at the community level to improve their image and build trust with customers and partners. They can focus on local support activities such as education, health, infrastructure development, and charity programs to develop sustainable relationships with the community. CSR can also become an effective marketing tool, helping to enhance its reputation and increase competitiveness in the domestic market. The most essential thing for banks is human capital; they need to train and improve the capacity of members of their organization. Commercial banks should regularly develop training programs to enhance the capacity to implement CSR for senior managers. Bank managers should implement intensive training programs to improve CSR strategy planning skills, including impact assessment, risk management, and measuring the effectiveness of CSR initiatives. Banks must also disseminate international standards and organize regular workshops to update global CSR trends and relevant legal changes. In short, both policymakers and stakeholders play an essential role in promoting these activities, contributing to building a sustainable, stable, and socially responsible banking system. In addition, integrating CSR enables banks to build a reputable image and enhance trust among customers and the broader community, strengthening core competencies such as brand reputation and service quality. A diversification strategy allows banks to reduce concentration risk, expand growth opportunities, and develop new capabilities. When CSR and diversification are effectively integrated, banks achieve short-term performance gains and establish a strong foundation for long-term competitive advantage through enhanced internal capabilities and greater adaptability.

6.3. Limitation

Although this study extends the prior literature on CSR and diversification in the banking industry, it has the following drawbacks. Firstly, the study is constrained by a relatively modest sample size of only 358 annual observations. This limitation is attributed to data restrictions and may require more robust statistical inferences. Some Vietnamese commercial banks need the data to create CSR and core competence indices covering 2010–2023. As a result, observations with incomplete data were not included in the research analysis. Second, given that Vietnam is a frontier and transitional market, the study’s conclusions have limited relevance to rising and industrialized nations. With these limitations in mind, the study emphasizes how important it will be to broaden the scope of its data to enable a comparative evaluation of CSR performance and the diversified income influence on banks’ core competence in developed and developing countries. This strategy will produce more thorough insights.