Human Capital and Bank Performance: Does Size Matter?

Abstract

This study was conducted to examine the moderating effect of size on the impact of human capital on bank performance, using data from 26 commercial banks in Vietnam from 2008 to 2023 through panel data regression methods. The results indicate that bank size and human capital are important resources for commercial banks to increase their performance, which is consistent with the resource-based view and economies of scale theory. However, bank size fails to exhibit a significant moderating effect on the impact of human capital on the bank performance in Vietnam. This phenomenon can be explained by the relatively limited influence of size effects on human capital, coupled with the fact that the majority of Vietnamese commercial banks place significant strategic emphasis on human capital development within their operational frameworks. In addition, this study highlights the impact of some internal factors and the macroeconomic conditions on bank performance. From these empirical findings, this paper recommends several critical policies.

1. Introduction

The national financial system plays an important role in the economy with its main function as an intermediary between capital providers and users (Filatie & Sharma, 2024). In the financial system, banks are the most important entities, considered the lifeblood of the economy (Vo & Tran, 2021), with many functions related to deposit management, payment intermediaries, and credit intermediaries (Mishkin, 2012). Although the banking sector does not directly create wealth for the economy, it plays a crucial role in promoting the financial system as well as the economy to develop sustainably. Currently, in Vietnam, according to T. K. O. Tran et al. (2017), the nation’s financial system is still developing at a limited level, primarily relying on the development of the banking system. At the same time, in developing countries like Vietnam, the economy primarily develops based on internal and external capital sources, through financial systems, and mainly emphasizes the role of the banking system. Therefore, to develop the economy sustainably, it is necessary for the banking system to operate efficiently and stably. This is because the efficiency and stability of commercial banks are important drivers for future GDP growth (Jokipii & Monnin, 2013).

Currently, commercial banks have many measures with which to increase their performance, such as those aimed at improving financial conditions and those related to the development of intangible and tangible assets of enterprises, such as intellectual capital. Pulic (1998, 2000) introduced the value-added intellectual coefficient (VAIC) model and pointed out that intellectual capital includes human capital efficiency, structural capital efficiency, and capital employed efficiency to measure the value added from the tangible to intangible assets of the enterprise. Among the three components of intellectual capital efficiency, human capital efficiency is the most important, and it is the component that helps organizations maintain a sustainable competitive advantage (Ur Rehman et al., 2022; Meles et al., 2016). At the same time, human capital plays a crucial role in the operations of any organization; it is the foundation for businesses to create differentiating factors, thereby determining the success or failure of a company compared to its competitors in the same industry (Meles et al., 2016). This is also reasonable as most organizations, in general, and commercial banks, in particular, always maintain a human resource management department to oversee and sustain performance.

Besides human capital, commercial banks are currently seeking to increase their size in order to enhance their performance. According to Berger and Mester (1997), increasing bank size helps commercial banks achieve efficiency in resource utilization and cost reduction. Laeven et al. (2016) and Demirgüç-Kunt and Huizinga (2010) stated that with a larger size, banks can better diversify revenue sources, allocate risks more effectively, and increase their resilience to risks. In Vietnam, the Government and the State Bank of Vietnam encourage banks to enhance their size, financial capacity, and competitiveness through increasing charter capital, mergers, consolidations, and acquisitions.

Regarding the research topics of human capital, size, and bank performance, most studies currently examine the individual impact of human capital or factors specific to commercial banks on their performance, such as the studies by Ozkan et al. (2017), Haris et al. (2019), Ur Rehman et al. (2022), Vo and Tran (2021), N. P. Tran and Vo (2020). As for examining the moderating relationship, most studies have focused on the moderating role of other factors (such as diversification) on the impact of human capital on bank performance, such as in the studies by Adesina (2021), Githaiga (2021, 2022a, 2022b), and D. T. Nguyen et al. (2023), or other factors in the relationship between bank size and the performance of commercial banks, such as in the study by Kouzez (2023). It appears that the moderating effect of size on the impact of human capital on bank performance has not yet been thoroughly examined by numerous studies. This is the gap that this research aims to fill.

To evaluate the moderating effect of size on the impact of human capital on the bank’s performance in a developing country like Vietnam, this study used data from 26 commercial banks in the period from 2008 to 2023 and employed panel data regression methods, including those that utilize p-values such as Pooled OLS, FEM, REM, GLS, and Bayesian frequency. The research results show that both human capital and bank size positively impact bank performance, consistent with the resource-based view and economies of scale theory. However, regarding the moderating relationship, the research results show an ambiguous moderating effect of bank size on the impact of human capital on bank performance. This can be readily explained by the fact that, in Vietnam, banks—regardless of size—consistently emphasize the development and role of their human resources teams.

Compared to previous studies, this research has the following contributions: First, this study examines the impact of human capital—the most important component of intellectual capital—on bank performance. Although this topic has been extensively examined, most studies have been conducted in developed countries and often focus on the overall impact of intellectual capital on bank performance. This study delves into the most important component of intellectual capital, which is human capital, in relation to bank performance in a developing country like Vietnam. Second, this study specifically examines the moderating effect of size on the impact of human capital on bank performance. It seems no in-depth analysis has been conducted yet on this topic, but rather evaluations of the individual relationships of each factor with the operational efficiency of commercial banks. Third, in addition to examining traditional estimation methods, this study also employs the Bayesian method to test the robustness of the model. According to N. T. Nguyen (2025), the Bayesian method has several advantages over traditional methods, thanks to a different approach in statistical inference, making Bayesian estimation more flexible in handling multicollinearity. At the same time, the use of informative prior distributions helps enhance the reliability of the regression coefficients.

Based on the above content, the remainder of this paper is organized as follows: Section 2 presents the foundational theories, relevant literature, and development of research hypotheses. Section 3 describes the methodology, data, and research model. Section 4 analyzes the research findings, and Section 5 concludes the study.

2. Related Literature and Hypothesis Development

Stewart (1997) and Luthy (1998) argue that human capital is the most fundamental component among the three basic components of intellectual capital, alongside structural capital and capital utilization efficiency. Accordingly, human capital represents the knowledge, skills, experience, and educational background of employees (Luthy, 1998). Additionally, human capital defines the ability of individuals within an organization to combine their efforts to solve complex business problems in order for the enterprise to achieve its goals. Meles et al. (2016) cited the research of Roos et al. (1997) and pointed out that, within an organization, employees can create intellectual capital through their competence, agility, and abilities. Therefore, human capital can be understood as a measure of the capabilities, professional skills, and experience of employees, and this is considered a prerequisite for a business to achieve sustainable development and is the most important component in the development of intellectual capital. Regarding scale, scale is typically measured by the total assets that a commercial bank has, reflecting the total resources that the bank currently possesses.

The relationship between human capital, size, and bank performance is discussed by several theories. First, the economies of scale theory explain why large banks can achieve better efficiency. Berger and Mester (1997) argue that large banks can leverage fixed costs from significant investments in information technology, risk management systems, and digital platforms, increase capital mobilization efficiency due to their good reputation and low-cost market access, provide product diversification, and enhance management efficiency, thereby improving operational performance. Therefore, it can be seen that due to economies of scale, input costs can be reduced, including labor costs, which helps businesses increase their human capital.

Additionally, resource-based view (RBV) theory proposed by Wernerfelt (1984) and Barney (1991) states that each organization possesses a unique set of resources and capabilities, which serve as the foundation for businesses to create sustainable competitive advantages. Among them, the assets and human capital are the resources that this theory refers to, with human capital being considered a valuable, rare, and unique asset (Kraaijenbrink, 2011). According to Wright et al. (1994), high-quality human resources and good assets are prerequisites for bringing benefits to the enterprise and allowing the business to develop and succeed in new markets.

• Literature Review

Regarding the research topic on the impact of human capital on the performance of commercial banks and the role of bank size, current studies mainly examine the individual impact of human capital or the overall impact of intellectual capital on the performance of commercial banks. Meles et al. (2016) used a sample of 5749 commercial banks in the U.S. and indicated that higher intellectual capital positively impacts the operational efficiency of U.S. commercial banks. In addition, the authors also showed that the effectiveness of human capital has a stronger impact compared to other components of intellectual capital on operational performance. Similarly, Mondal and Ghosh (2012) used data from 65 banks in India to show that intellectual capital is an important factor determining the profitability and productivity of banks. However, by breaking down the topic into smaller components, Mondal and Ghosh (2012) also pointed out that the effectiveness of human capital is a key component in enhancing the profitability of banks. Also, on the topic of research related to the relationship between intellectual capital and bank performance, Alhassan and Asare (2016), Haris et al. (2019), and Ozkan et al. (2017) also showed that the effectiveness of human capital, as a component of intellectual capital, has the most significant positive impact on the performance of banks.

In addition to these studies, when examining the impact of human capital efficiency and corporate value, Sisodia et al. (2021) provide further empirical evidence of the positive relationship between human capital efficiency and corporate value. Accordingly, the research argues that human capital enhances the value of a business through several aspects, the first being that businesses are better able to seize growth opportunities in the present when they improve their human capital. At the same time, increasing the efficiency of human capital helps businesses create more growth opportunities in the future and reduce fluctuations related to the growth rate of the business. Adesina (2019) analyzed the impact of intellectual capital with its basic components including structural capital, human capital, and physical capital on technical efficiency, allocation, and banking costs. The author used data from 339 commercial banks operating in 31 African countries during the period from 2005 to 2015 with Tobit and GMM regression estimation methods. The research results indicate that among the three components of intellectual capital, only human capital positively affects the allocation efficiency, cost, and technical efficiency of banks across all regions of Africa.

In Vietnam, N. P. Tran and Vo (2020) used a dataset of 227 listed companies on the stock market during the period from 2011 to 2018. The research results show that the effectiveness of human capital varies across industries in Vietnam; however, it has a positive impact on the performance of businesses. Therefore, the study proposed policy implications to encourage the development and more efficient use of human capital. In addition to this study, in Vietnam, several other authors have also analyzed the impact of human capital as a component of intellectual capital on the operational efficiency of commercial banks, such N. T. Nguyen and Le (2021), D. T. Nguyen et al. (2021), and Phan and Nguyen (2023). These studies all show a positive relationship between the effectiveness of human capital and the operational efficiency of commercial banks.

Regarding the interactive relationship, previous studies mainly examined the interactive impact of diversification on the effect of human capital on bank performance, such as those by Githaiga (2021, 2022a, 2022b). In Vietnam, D. T. Nguyen et al. (2023) examined the moderating effect of income diversification on the relationship between intellectual capital and bank performance. Their study used data from Vietnamese commercial banks during the period from 2007 to 2020 through GMM estimation, and the research results showed that VAIC and its components positively impact the operational efficiency of banks, while income diversification has a negative and significant impact on bank performance. Finally, their study showed that income diversification is a moderating factor, reducing the overall impact of intellectual capital efficiency on banking performance. Thus, revenue diversification improves the impact of structural capital efficiency while reducing the impact of human capital efficiency on banking performance. At the same time, income diversification has a lower impact on the operational efficiency of banks in mitigating the effects of capital utilization efficiency.

Thus, it can be found that studies examining the moderating role of size on the impact of human capital on bank performance have not yet received much attention. Most studies only examine the individual impact or the moderating effect of other factors such as diversification on the impact of human capital on bank performance. Therefore, this study will fill the research gap regarding the moderating role of scale in the impact of human capital on the performance of commercial banks. In fact, size and human capital are both resources that commercial banks use in their business operations. Commercial banks with large capital scales often have more opportunities to invest in technology and process management to enhance operational efficiency (Berger & Mester, 1997), while human capital is a component that reflects knowledge, skills, experience, health, and motivation to enhance the operational efficiency of commercial banks (Meles et al., 2016). At the same time, based on the economies of scale theory, large-scale banks often have better conditions to invest in training, retraining, skill development, and building modern human resource management systems, thereby helping to increase operational efficiency. Based on the results of empirical studies and foundational theories, the author proposes the following hypotheses:

Hypothesis H1. 

Human capital has a positive impact on bank performance in Vietnam.

Hypothesis H2. 

Bank size has a positive impact on bank performance in Vietnam.

Hypothesis H3. 

Size has a positive impact on the impact of human capital on bank performance in Vietnam.

3. Methodology, Data, and Model

• Model and Data

Based on the research models of Adesina (2021), Meles et al. (2016), and Ur Rehman et al. (2022), the research model is as follows:

$$bank performance=f\left(human capital, human capital\times size,size, Controls\right)$$

Specifically, the research model is as follows:

$$Y_{i,t}={\alpha }_0+{\delta }_i{{HCE}_{i,t}+{\beta }_i{HCE}_{i,t}\times {Size}_{i,t}+\alpha }_i{Size}_{i,t}+{\gamma }_k{BSC}_{i,t}+{\mu }_j{MC}_t+{\epsilon }_{i,t}$$

(1)

In model (1), i and t represent bank i in year t, Y denotes the dependent variables, measured by the return on assets (ROA) and risk-adjusted return on assets (RAROA), human capital is measured through the HCE indicator, Size is the bank size, HCE × Size is the interaction variable between human capital and the bank size, and BSC denotes control variables that reflect some intrinsic characteristics of commercial banks, including the bank capitalization ratio, loan-to-assets ratio, and customer deposit-to-assets ratio. MC denotes control variables, representing the macroeconomic factors affecting bank performance, including GDP growth and the inflation rate. Finally, ${\alpha }_0$ is the intercept, coefficients δ, β, α, γ, and µ are the regression coefficients, and ${\epsilon }_{i,t}$ is the residual of the model.

• Definitions of the Variables in the Research Models
• Variables Representing Bank Performance

In this study, the author uses the return on assets (ROA) and the risk-adjusted return on assets (RAROA) to measure bank performance. The ROA indicates how much profit the bank generates from every 100 units of assets it invests. Generating a larger profit for commercial banks per 100 units of assets indicates an increase in the bank’s performance. In fact, previous studies such as that of D. T. Nguyen et al. (2023), N. P. Tran and Vo (2020), and Githaiga (2022b) used indicators including both the ROA and ROE (return on equity) to represent bank performance. However, the ROE indicator shows how much profit the bank generates from every 100 units of the owner’s investment. Because this indicator is calculated based on the invested owner’s equity, it depends on the bank’s leverage ratio; at the same time, the nature of commercial banks’ operations is financial intermediation, so the capital used by the bank mainly comes from deposit activities (Vo & Tran, 2021). Therefore, ROA is a more objective indicator of the bank’s performance. It has been widely used in previous studies such as in that of Githaiga (2022b), Ozkan et al. (2017), and Vo and Tran (2021). According to D. T. Nguyen et al. (2023), N. P. Tran and Vo (2020), and Githaiga (2022b), the ROA indicator can be calculated using the following formula:

$$ROA={\displaystyle \frac{Net income}{Average of total assets}}$$

(2)

Besides the ROA, the author also uses the RAROA indicator to measure bank performance. RAROA measures the volatility of ROA, indicating the level of risk of the commercial banks. This indicator was used and calculated by Saghi-Zedek (2016) according to the following formula:

$$RAROA={\displaystyle \frac{ROA}{\sigma \left(ROA\right)}}$$

(3)

In formula (3), ROA and σ(ROA), respectively, represent the return on assets and the standard deviation of the return on assets of commercial banks. When ROA fluctuates, causing σ(ROA) to increase, it results in a decrease in RAROA and vice versa. From this approach, a low RAROA means that commercial banks face high profit volatility risk; conversely, a high RAROA means that commercial banks do not have significant fluctuations in profit margins, indicating low profit volatility risk and increased operational efficiency.

• Variable Representing Human Capital

Human capital efficiency (HCE) is the main explanatory variable in the research models, and it refers to the effectiveness of human capital in utilizing intellectual resources to create value (Adesina, 2021). Building on previous studies by N. P. Tran and Vo (2020), Sisodia et al. (2021), and Adesina (2021), the author measures the human capital efficiency of Vietnamese commercial banks using the human capital component in the value-added intellectual coefficient (VAIC) model proposed and developed by Pulic (1998). In the VAIC model, HCE indicates the amount of value added that humans are capable of creating for the bank (Adesina, 2021). Building on the research of Adesina (2021), Ur Rehman et al. (2022), Vo and Tran (2021), and Meles et al. (2016), HCE is calculated based on the marginal contribution per unit of money invested in employees using the following formula:

$$HCE={\displaystyle \frac{VA}{HC}}$$

(4)

In Formula (4), VA is the added value of the bank and is calculated using the formula

$$VA=OP+HC$$

where OP is the operating profit of the bank (the bank’s profit before tax—general expenses (including labor costs)—provisions for loan risks), and HC is employee expenses (including salaries, bonuses, and other costs) (Vo & Tran, 2021).

• Variable Representing Size

Bank size is measured using the logarithm of its total assets. When a commercial bank has a larger total asset size, it will gain economies of size and a better ability to increase market share; thus, the bank’s performance is expected to improve.

• Other Control Variables

Besides variables related to human capital and size, bank performance also depends on several other factors, including those characteristics of the banks and those related to the macroeconomic condition. Among them, the characteristic factors of commercial banks include the bank capitalization ratio, the loan-to-assets ratio, and the deposit ratio. Macroeconomic factors include GDP economic growth and inflation rate. Among them, the capitalization ratio of commercial banks is calculated using the ratio of equity to total assets. According to Githaiga (2021), increasing the capital of commercial banks helps reduce financial distress costs and bankruptcy risks, thereby lowering the capital costs of commercial banks. When the equity-to-assets ratio increases, lenders view the commercial bank as having sufficient capital to address adverse conditions in the macroeconomy (Githaiga, 2021). Therefore, we expect the equity-to-total-assets ratio to have a positive impact on bank performance. The loan-to-total-assets ratio indicates the efficiency of capital utilization by commercial banks. When the loan-to-asset ratio increases, interest income rises, which provides a basis for commercial banks to increase profits, thereby enhancing operational efficiency. The deposit ratio is calculated using the ratio deposits to total assets, reflecting the liquidity of commercial banks (Abuzayed et al., 2018), and it should have a positive impact on bank performance. GDP and CPI reflect the “health” of the economy. According to Abuzayed et al. (2018), increases in GDP growth increases and decreases in the CPI inflation rate will motivate commercial banks to generate more profits. Therefore, with regard to bank performance, in this article, we expect a positive relationship with GDP economic growth and a negative relationship with inflation (

Table 1. Description of variables in the research models.

Symbol	Variable	Measure	Expected of Sign	Previous Research
Dependent variables
ROA	Return on assets	$ROA={\displaystyle \frac{EAT}{Average total of assets}}$		Adesina (2021), Ur Rehman et al. (2022), Vo and Tran (2021), Meles et al. (2016)
RAROA	Risk-adjusted return on assets	$RAROA={\displaystyle \frac{ROA}{\sigma \left(ROA\right)}}$		Saghi-Zedek (2016), Adesina (2021), Radojičić and Marinkovic (2023)
Independent variables
Main regressors
HCE	Human capital efficiency	Formula (4)	+	Adesina (2021), Meles et al. (2016), Phan and Nguyen (2023), Ur Rehman et al. (2022), Vo and Tran (2021)
SIZE	Bank size	Logarithm of total assets	+	Sisodia et al. (2021), N. P. Tran and Vo (2020), Githaiga (2021), Radojičić and Marinkovic (2023)
Control variables
ETA	Bank capitalization ratio	${\displaystyle \frac{Owne{r}^{\prime }s equity}{Total of assets}}$	+	Adesina (2021); Stiroh (2004), Githaiga (2021), Phan and Nguyen (2023)
LTA	Loan-to-assets ratio	${\displaystyle \frac{Loan}{Total of assets}}$	+/−	Phan and Nguyen (2023), Adesina (2021)
DEP	Deposit ratio	${\displaystyle \frac{Deposit}{Total of assets}}$	+	Sisodia et al. (2021), Adesina (2021), Phan and Nguyen (2023)
GDP	GDP growth	${\displaystyle \frac{{GDP}_t-{GDP}_{t-1}}{{GDP}_{t-1}}}$	+	Abuzayed et al. (2018), Ur Rehman et al. (2022)
INF	Inflation rate	${\displaystyle \frac{{CPI}_t-{CPI}_{t-1}}{{CPI}_{t-1}}}$	-	Abuzayed et al. (2018), Ur Rehman et al. (2022)

Source: Compiled by the author.

).

Regarding the research data, the model variables are calculated based on the data provided by commercial banks from audited financial statements and annual reports. At the same time, to ensure data accuracy, the research was rechecked through the Fiinpro X database. GDP and CPI data were obtained from the World Bank database. Thus, it can be seen that the data sources used in this article are highly reliable, ensuring quality for estimating the regression coefficients.

• Methodology

This study used data from 26 Vietnamese commercial banks from 2008 to 2023 to examine the moderating effect of size on the impact of human capital on bank performance. To estimate the research models, the author used panel data regression techniques including Pooled OLS, a Fixed Effects Model (FEM), and a Random Effects Model (REM), followed by the selection of the most appropriate specifications. However, diagnostic tests revealed the presence of heteroscedasticity and autocorrelation in the selected models. To address these issues and ensure robust estimation, the author adopted the Feasible Generalized Least Squares (FGLS) method.

4. Results

4.1. Baseline Results

This study was conducted with a sample of data from 26 Vietnamese commercial banks during the period from 2008 to 2023. However, some observations missed data and some observations with outliers were removed; thus, in the end, 367 observations with complete data were included in the model to estimate the regression coefficients.

Table 2. Descriptive statistics.

Variables	Observations	Mean	SD	Min	Max	Skewness	IQR (Q3–Q1)
ROA	367	0.0101	0.0073	0.0000	0.0365	0.7823	0.0098
RAROA	367	6.7177	9.6240	0.0107	92.9620	4.9423	5.2154
HCE	367	2.4887	1.0233	1.0021	6.5209	0.8142	1.4098
SIZE	367	14.1282	0.5215	12.8738	15.3619	0.5027	0.7200
ETA	367	0.0907	0.0382	0.0262	0.2662	1.5458	0.0407
LTA	367	0.5660	0.1182	0.1910	0.7881	−0.7176	0.1776
DEP	367	0.6496	0.1179	0.2881	0.8937	−0.3720	0.1746
GDP	367	0.0596	0.0161	0.0255	0.0812	−0.9539	0.0159
CPI	367	0.0522	0.0487	0.0063	0.2312	2.2575	0.0380

Source: Author’s calculations.

presents the descriptive statistics of the variables in the research models.

From the results in Table 2, it can be observed that the bank’s performance differs through the ROA indicator, with a standard deviation of 0.0101, a minimum value of 0, and a maximum value of 3.65%. with a relatively low standard deviation (0.0073). This indicates modest but consistent profitability among Vietnamese commercial banks. The positive skewness (0.7823) and interquartile range (IQR = 0.0098) suggest that while most banks maintain low to moderate profitability, a few achieve significantly higher returns, pointing to the presence of mild outliers. Similarly, the risk-adjusted return on assets (RAROA) also shows differences among Vietnamese commercial banks, with a standard deviation of 9.62, where the minimum value is 0.0107 and the maximum value is 92.96, and the distribution is highly positively skewed (4.9423). The wide interquartile range (IQR = 5.2154) reflects a substantial dispersion in risk-adjusted performance, implying that some banks may be taking on higher risk or possess superior risk management capabilities.

Regarding the human capital indicator, human capital in commercial banks seems to not differ much, with an average HCE indicator of 2.4887, ranging from 1.0021 to 6.0259 with a standard deviation of 1.0233. This indicates a notable variation in how banks utilize their human resources. The positive skewness (0.8142) and IQR of 1.4098 suggest that while most banks cluster around the average level of efficiency, a subset of institutions demonstrate substantially higher efficiency in translating human capital into value. Bank size (SIZE), measured as the natural logarithm of total assets, is more symmetrically distributed with a mean of 14.13 and standard deviation of 0.52. Its skewness (0.5027) and modest IQR (0.7200) suggest that while size varies across banks, extreme deviations are limited. This relative consistency is expected in a market dominated by a mix of mid- to large-sized banks.

Regarding the control variables, during the period from 2008 to 2023, commercial banks had an average equity-to-total-assets ratio, loan-to-total-assets ratio, and average deposit-to-total-assets ratio of 0.0927, 0.5660, and 0.6496, respectively, fluctuating in the ranges of 0.0262 to 0.2662, 0.1910 to 0.7881, and 0.2881 to 0.8937. Regarding the macroeconomic variables, during the study period, Vietnam’s average macroeconomic growth reached 5.96% with an inflation rate of 5.22%.

To provide an additional basis for including variables in the research models, this article examines the correlation coefficients between the variables in the models.

Table 3. Correlation coefficients.

	ROA	HCE	SIZE	ETA	LTA	DEP	GDP	CPI	VIF
ROA	1
HCE	0.828	1							1.56
SIZE	0.2552	0.1858	1						1.96
ETA	0.3475	0.2201	−0.4826	1					1.66
LTA	0.0882	−0.0743	0.3403	−0.0086	1				1.77
DEP	−0.3503	−0.3806	0.2241	−0.25	0.5463	1			1.82
GDP	−0.0829	−0.0686	−0.0741	−0.0097	−0.0195	−0.0009	1		1.01
CPI	0.1441	0.297	−0.3361	0.2491	−0.371	−0.3902	0.0267	1	1.43
Average of VIF									1.60

Source: Author’s calculations.

presents the correlation coefficients between the variables in the research model and the variance inflation factor (VIF). The research results show that the independent variables have relatively low correlation levels with VIF coefficients, all below 5. This indicates that the independent variables do not exhibit serious multicollinearity, allowing for further estimations to be conducted.

To estimate the parameters in the regression models, the author used Pooled OLS, FEM, and REM estimators to conduct the regressions, and then selected the appropriate model. However, the appropriate model exhibited heteroscedasticity and autocorrelation, so the author proceeded with FGLS estimation.

Table 4. Estimation results of the FGLS model.

Variables	Model with Dependent Variable of ROA	Model with Dependent Variable of RAROA
HCE	0.0054 *** (0.0002)	0.9545 *** (0.3257)
SIZE	0.0022 *** (0.0005)	4.3875 *** (0.7219)
HCE × SIZE 1	−0.0001 (0.0003)	−0.3280 (0.4551)
ETA	0.0292 *** (0.0050)	14.7668 ** (7.4428)
LTA	0.0042 ** (0.0017)	4.7998 * (2.4930)
DEP	−0.0015 (0.0015)	−1.6715 (2.6103)
GDP	0.0037 (0.0052)	−4.2640 (9.1989)
CPI	0.0060 ** (0.0026)	4.0178 (4.4563)
_cons	−0.0395 (0.0069)	−62.0023 (10.3272)

Source: Author’s calculations. ***, **, and * indicate that the regression coefficients are statistically significant at the 1%, 5%, and 10% levels. ¹ To avoid multicollinearity between the interaction term and the original independent variables, the author centered the variables before creating the interaction term.

presents the FGLS estimation results.1

The research results presented in Table 4 can be explained as follows:

First, human capital has a positive impact on bank performance at a significance level of 1%. Based on these results, when commercial banks invest in human capital, their performance increases. This result aligns with the research hypothesis and most previous studies, such as those of N. P. Tran and Vo (2020), Sisodia et al. (2021), Rahman and Akhter (2021), Le and Nguyen (2020), N. T. Nguyen and Le (2021), and Phan and Nguyen (2023), as well as with the resource dependence theory. In fact, human capital efficiency reflects the wealth of knowledge that individuals possess, and its effective utilization will contribute to enhancing the operational efficiency of commercial banks (Mondal & Ghosh, 2012). Specifically, increasing human capital will help improve labor productivity, thereby enhancing the operational efficiency of commercial banks. The investment of commercial banks in their workforce through training activities, technology support, and work environment improvements helps bank employees serve customers better and save time, increases the cross-selling rate of other products, and reduces errors in operations and transactions. At the same time, in high-risk operations such as credit, investment, and asset management, it helps commercial banks enhance decision-making quality, thereby improving capital utilization efficiency and reducing risks during the process, such as in non-performance loan situations. This helps banks expand their portfolio of banking services and financial products to meet diverse customer needs, increasingly meeting the comprehensive financial needs of customers, including both individual and institutional clients. From there, it helps increase the bank’s market share and serves as a foundation for the commercial bank to enhance its reputation, brand, and operational efficiency. From these analyses, this article accepts hypothesis H1: The human capital efficiency has a positive impact on bank performance in Vietnam.

Second, bank size also has a positive impact on bank performance at a significance level of 1%. This result is also consistent with resource dependence theory and the economies of scale theory, as well as with previous studies, such as those of Berger and Mester (1997) and Adesina (2021). This result demonstrates that larger banks tend to be more efficient in pricing and utilizing inputs for certain outputs (Adesina, 2021). This is also true for Vietnamese commercial banks. Typically, banks with larger capital can often distribute fixed costs such as technology expenses, senior personnel, and risk management systems across more operating units. As a result, average costs will decrease, helping to increase profit margins, which is also consistent with the views of Berger and Mester (1997). Additionally, when commercial banks have large capital, they often have stronger market positions, which makes it easier for them to negotiate and bargain when raising funds to obtain lower fees and interest rates. In addition, a large size helps commercial banks more easily expand their product portfolios and branch networks, thereby increasing the number of customers. This contributes to reducing risk and stabilizing income streams, thereby enhancing performance. So, this study accepts hypothesis H2: Bank size has a positive impact on bank performance in Vietnam.

Third, although the results indicate that both size and human capital positively impact the performance of commercial banks, when examining the moderating role of bank size, the results show that the regression coefficient of the interaction variable between bank size and human capital efficiency is not statistically significant. That is, the moderating role of size on the impact of human capital on the performance of commercial banks was not found. This result differs when considering resource-based theories and economies of scale theories, as well as the hypotheses set for the economy. However, this result seems to align with the business operations of Vietnamese commercial banks. Unlike developed economies where larger banks often exhibit more structured human resource (HR) systems and strategic investments in human capital, the HR practices in Vietnam may be more homogeneous across institutions regardless of size. Additionally, regulatory frameworks and labor market conditions may constrain the extent to which large banks can leverage their scale to enhance human capital efficiency. Specifically, because human capital reflects the productivity of the workforce in creating added value for commercial banks, this indicator measures the quality of personnel utilization rather than simply depending on the size of the organization or the number of employees. Moreover, in Vietnam, almost all banks (including both small and large banks) consider the quality of their workforce. Small commercial banks are still willing to offer high salaries to experienced senior staff with good management skills. Additionally, personnel management, in some respects, does not necessarily depend on asset size but mainly relies on training policies, compensation levels, organizational culture, and the ability to retain talent. This is the reason why this study did not find a moderating relationship of scale on the impact of human capital on the performance of commercial banks.

Fourth, regarding the results of the control variables, the research findings indicate that the equity ratio has a positive impact on bank performance. This demonstrates that when commercial banks possess a higher level of equity capital, their operational efficiency improves, which is considered a very good buffer when commercial banks encounter shocks. In addition, the loan-to-total-assets ratio also positively impacts the operational efficiency of commercial banks, demonstrating the role of their credit activities. Although currently, commercial banks are still focused on developing non-interest activities and services, credit activities remain the main source of income for commercial banks, accurately reflecting the intermediary nature of commercial banks when they receive deposits from surplus funders to lend to those in need of borrowing. However, studies have not yet found a relationship between the deposit rate and the operational efficiency of commercial banks. An explanation for this is that currently in Vietnam, commercial banks have to compete fiercely; banks not only compete with each other but also with foreign commercial banks. Therefore, banks are forced to raise deposit interest rates to retain customers, which does not improve a bank’s efficiency despite the large amount of mobilized capital. Additionally, this study did not find a relationship between GDP economic growth and the operational efficiency of commercial banks. This is because, during the research period, the Vietnamese economy went through many ups and downs. The 2008–2009 period was marked by a global economic crisis, followed by a phase of economic recovery. In 2015, Vietnam faced the restructuring of the banking sector; then, in the 2020 phase, it faced the COVID-19 pandemic. However, during these difficult periods, Vietnam’s inflation remained low, which had a stimulating effect on the economy.

4.2. Robustness Check

To check the robustness of the models, this study used Bayesian estimation to evaluate the impact of human capital on the performance of commercial banks considering the role of human capital. According to T. N. Q. Nguyen and Nguyen (2024), Bayesian estimation has several advantages over traditional inferential estimation. First, traditional estimation methods mainly rely on research data without prior information, while Bayesian estimation is not only based on research data but also on prior information. Second, Bayesian estimation can address some data deficiencies such as autocorrelation, heteroscedasticity, and endogeneity (N. T. Nguyen, 2025). Third, traditional estimation methods may overlook insignificant variables despite their potential impact on the analysis, whereas the Bayesian approach considers the influence of all variables (T. N. Q. Nguyen & Nguyen, 2024).

Table 5. Estimation results using Bayesian estimation.

	Mean	Std. Dev.	MCSE	Median	Probability of Coefficient Mean > 0	Equal-Tailed [95% Cred. Interval]
ROA
HCE	0.0047	0.0062	0.0001	0.0047	0.8950	[−0.0097, 0.0187]
SIZE	0.0032	0.0146	0.0003	0.0032	0.6117	[−0.0258, 0.0327]
HCE × SIZE	−0.0004	0.0107	0.0002	−0.0003	0.4750	[−0.0256, 0.0234]
ETA	0.0441	0.1579	0.0029	0.0450	0.6887	[−0.3187, 0.4119]
LTA	0.0135	0.0546	0.0011	0.0133	0.6717	[−0.1121, 0.1276]
DEP	−0.0120	0.0581	0.0011	−0.0099	0.3717	[−0.1516, 0.1086]
GDP	−0.0062	0.2525	0.0046	−0.0080	0.4627	[−0.5728, 0.6171]
CPI	−0.0021	0.1146	0.0021	−0.0044	0.4717	[−0.2576, 0.2755]
_cons	−0.0506	0.2066	0.0041	−0.0526		[−0.4696, 0.3728]
RAROA
HCE	1.3698	0.4699	0.0086	1.3666	0.9980	[0.4450, 2.3132]
SIZE	0.2387	0.1366	0.0024	0.2382	0.9610	[−0.0300, 0.5081]
HCE × SIZE	0.0835	0.7172	0.0126	0.0820	0.5490	[−1.3992, 1.4576]
ETA	0.0235	1.0163	0.0186	0.0288	0.5133	[−2.0551, 2.0130]
LTA	0.3623	0.9629	0.0182	0.3653	0.6520	[−1.5278, 2.2691]
DEP	0.0993	0.9850	0.0180	0.1301	0.5490	[−1.8512, 2.0201]
GDP	−0.0274	1.0304	0.0183	−0.0307	0.4883	[−2.0925, 2.0456]
CPI	−0.1493	0.9929	0.0182	−0.1562	0.4360	[−2.0820, 1.7972]
_cons	−0.2319	0.9976	0.0182	−0.2024

Source: Author’s calculations.

presents the estimation results of the research models according to Bayesian estimation.

According to Bayesian estimation (Table 5), human capital and scale have a very clear positive impact on bank performance (with the posterior probability of this regression coefficient having a positive impact on performance being greater than 60%). Among them, the posterior probabilities of the HCE and SIZE variables in the ROA-dependent variable model are 89.50% and 61.17%, while the posterior probabilities of these two variables in the RAROA-dependent variable model are 99.8% and 96.10%. This means that the impact of the SIZE and HCE variables on bank performance is indeed positive, and the results are completely consistent with the FGLS estimation results. This indicates the robustness of the impact of human capital efficiency and scale on the operations of Vietnamese commercial banks. Meanwhile, the impact of interaction variable HCE × Size is ambiguous, with the posterior probability of this regression coefficient having a positive effect on ROA (47.5%) and RAROA (54.9%). These results are also almost identical to the research findings based on FGLS estimation, meaning that no clear moderating relationship was found regarding size on the impact of human capital efficiency on bank performance in Vietnam. For the control variables, the impact of these variables on the operational efficiency of commercial banks also does not seem to be very clear.

5. Conclusions

This study used data from 26 Vietnamese commercial banks during the period from 2008 to 2023, employing panel data regression methods. The research results show that both human capital and bank size drive the bank’s performance. However, this study found no evidence of a moderating effect of size on the impact of human capital on bank performance. The explanation for this is that in Vietnam, both small- and large-scale commercial banks consider and prioritize the development of human resources. Additionally, this study indicates a positive correlation between the ratio of the owner’s equity, loan-to-total assets ratio, and inflation rate with bank performance. However, this study did not find a clear relationship between the deposit rate and GDP economic growth with bank performance.

From these results, to enhance the performance, this study proposes the following policy implications: First, Vietnamese commercial banks need to continue focusing on improving human capital. Policymakers and regulatory bodies can consider establishing a national human capital benchmarking system for the banking sector. This system could track and compare human capital efficiency indicators across banks, thereby creating transparency and motivating continuous improvement. Additionally, regulators could encourage banks to allocate a minimum percentage of their operational budget (e.g., 3–5%) toward talent development, including training programs, leadership development, and digital skill building. At the same time, we recommend that banks integrate regular performance evaluations tied to human capital metrics, such as return on training investment or employee productivity ratios, to ensure alignment between workforce capabilities and strategic goals. Second, they also need a path with which to scale up, by increasing the owners’ equity and seeking low-cost capital sources to implement loans. Third, they should continue to invest in the lending activities of commercial banks through control and risk prevention systems.

Thus, fundamentally, this research achieved its goal of evaluating the moderating role of size on the impact of human capital on bank performance in Vietnam. However, this study still has some limitations. One limitation is that the size of a bank can be measured through various indicators, such as revenue growth, number of employees, revenue, or market share; however, in this study, the most commonly used measure to assess size was total assets. Therefore, in future studies, the author will use additional measures to provide a more comprehensive picture. Secondly, bank performance can also be measured using various indicators; however, this study only uses financial indicators. Therefore, in future research, the author will incorporate additional measures. Additionally, this study primarily focuses on Vietnamese commercial banks, but there are many other types of banks operating in Vietnam, such as joint venture banks, 100% foreign-owned banks, and other banking models. Therefore, in subsequent studies, the author will include more types of banks to reassess the regulatory role of size on the impact of human capital on the performance of commercial banks.