This section tries to examine the empirical findings derived from investigating the impact of intellectual capital (VAIC) and its components—human capital efficiency (HCE), structural capital efficiency (SCE), and capital employed efficiency (CEE)—on the stability of banks in Saudi Arabia. Leveraging a robust analytical framework, this section delineates the outcomes of the statistical tests conducted to validate the proposed hypotheses, offering a nuanced exploration of how VAIC influences bank stability amid varying economic conditions and the unforeseen challenges posed by the COVID-19 pandemic.
Through a detailed examination of the data collected from financial statements, annual reports, and relevant macroeconomic indicators, this segment elucidates the direct and moderating effects of VAIC components and control variables, such as leverage, size, and operational efficiency, on banks’ resilience and risk management capacities. The subsequent analysis sheds light on the theoretical propositions posited in the hypotheses. It provides practical insights into the strategic value of VAIC in enhancing the stability and sustainability of banking institutions within the volatile financial landscape of the Saudi banking sectors.
4.3. Panel Least Squares Estimation Results
The results of the panel least squares estimation, encompassing cross-sectional random effects and fixed effects, as well as estimations excluding these effects, are presented in
Table 6.
Based on
Table 6 and according to the results of the model without fixed or random effects, it is observed that the VAIC stands out with a coefficient of 1.7065, a t-statistic of 2.3033, and a probability of 0.0233, indicating a statistically significant positive impact on bank stability. Conversely, the LEV carries a negative coefficient of −27.36, a t-statistic of −2.2431, and a probability of 0.0271, showing a significant adverse effect. The OPEF, though positive, is not significant, with a high
p-value implying no clear influence. The SIZE and the GGDP also appear insignificant, suggesting they do not play a critical role in this context. The INF has a positive coefficient, while the COV has a negative coefficient. However, both are not statistically significant, implying that within the scope of this model, their impacts on bank stability are not discernible. The R-squared and adjusted R-squared are relatively low at 0.1754 and 0.1188, respectively, suggesting that the model without fixed or random effects explains only a modest portion of the variance in bank stability.
The Durbin–Watson statistic is low at 0.0602, indicating potential autocorrelation issues, and the F-statistic, while significant at 3.0988 with a p-value of 0.0053, suggests that the model is better than none but may not be optimal.
Moving to the fixed-effects model, we notice a stark improvement in the fit, as evidenced by an adjusted R-squared of 0.9867. Here, VAIC’s significance is even more pronounced, with a coefficient of 0.3737, a t-statistic of 3.7407, and a probability nearing zero, reinforcing the strong positive relationship with bank stability. The LEV variable remains negative and is highly significant, which underscores the consistent negative influence of LEV on stability. Interestingly, SIZE and OPEF show negative coefficients of −1.4564 and −4.1848, respectively, a departure from the previous model, indicating that when accounting for individual bank effects, both SIZE and OPEF have a noticeable negative impact on stability. The GGDP continues to show insignificance, whereas the INF and COV variables maintain positive and negative coefficients but still lack statistical significance. The model’s robustness is confirmed by a substantial F-statistic of 507.4014 and a probability of zero, alongside a satisfactory Durbin–Watson statistic of 1.564238, suggesting less concern over autocorrelation.
In the model with cross-section random effects, we observe consistency with the fixed-effects model regarding the significance and direction of all variables. The VAIC is still positive and significant. The SIZE, OPEF, and LEV variables remain significant negative relationships, while GGDP, INF, and COV remain insignificant.
This model’s F-statistic of 34.4629, with a zero probability indicating a good model fit. The Durbin–Watson statistic is moderate at 1.454395, which may indicate minimal autocorrelation concerns.
The result of the Husman test in
Table 7 shows that the preference would be towards a random-effects model rather than a fixed-effects model because the
p-value associated with the Chi-Square test in the cross-section random-effects model is indeed 1, which is much greater than the conventional significance level of 0.05. This high
p-value indicates no evidence to reject the null hypothesis, which states that the individual bank effects are not correlated with the other explanatory variables in the model. In other words, the test suggests that the variation across banks does not systematically affect the dependent variable, which, in this case, represents bank stability. Generally, the random effects effectively capture the individual differences among banks, preventing any correlation with the predictors.
Analyzing the random effect results in
Table 6, we find that the overall impact of VAIC on financial stability is positive and statistically significant, supporting the main hypothesis (H1) that intellectual capital positively influences the stability of banks in Saudi Arabia. This aligns with resource-based view (RBV) theory, which emphasizes the strategic management of a firm’s internal capabilities. It suggests that the key to achieving sustained competitive advantage lies in exploiting internal strengths that are valuable, rare, inimitable, and non-substitutable [
5]. In the context of banking, CEE represents such a resource; it is not only a measure of how effectively a bank uses its financial and physical assets but also reflects its ability to generate superior returns and maintain stability even during economic downturns.
This also aligns with the literature, which generally contends that intellectual capital enhances organizational performance, a critical factor in banking stability. Refs. [
10,
56] found evidence supporting intellectual capital’s positive role in banking performance and stability, suggesting a consistency with the positive coefficient of VAIC in the provided table. Refs. [
2,
4] also support the idea that intellectual capital contributes to bank stability and risk management, which aligns with the significance of VAIC in the table’s model.
Studies like [
16,
67] explore the impact of intellectual capital on bank risk, providing a broader context for understanding the implications of VAIC on financial stability. Refs. [
1,
68] examine the relationship between intellectual capital and stability, with findings that could provide insights into the broader implications of the positive VAIC coefficient. Refs. [
17,
22] offer insights into the efficiency of intellectual capital and its contribution to financial stability, potentially corroborating the positive impact of VAIC.
Size is significantly and negatively associated with bank stability; larger banks may confront specific challenges and risks, such as regulatory and systemic risks, that can impair their financial stability. This result aligns with the findings [
50]. Conversely, ref. [
2] found that bank size might positively correlate with stability, attributing this to the efficiency benefits of economies of scale. The OPEF, which exhibits both significance and a negative association, tells us that banks spending more on their operations than they make can lead to less stability. More spending can help a bank’s stability and profit. This concept supports the cost–benefit theory, which argues that businesses aim to balance their spending and earnings to maximize profits. Similar findings were observed in the research conducted by [
34,
35,
69].
LEV is significantly negative, indicating that higher leverage is associated with decreased stability. Theories related to the agency problem suggest that higher leverage may encourage riskier behavior by managers due to the pressure to deliver returns on borrowed capital. This can lead to underinvestment in safety and excessive risk-taking, further undermining the stability of financial institutions. Thus, the empirical findings reported in the literature, such as those of [
70], are consistent with these theoretical frameworks, indicating that high leverage can be detrimental to bank stability. These studies underscore the importance of maintaining a balanced capital structure to mitigate risks associated with high leverage.
Excessive leverage increases the risk of bank instability, as evidenced by studies across various regions. For instance, ref. [
39] shows that in Saudi Arabia, banks with higher leverage are more vulnerable during economic downturns. Ref. [
62] finds that high leverage correlates with financial distress internationally, while [
70] illustrates that Islamic banks’ lower leverage contributes to greater stability. These studies underscore the need for prudent leverage management to ensure bank stability, especially during economic shocks.
GGDP and INF are not statistically significant, suggesting that these macroeconomic variables do not have a discernible impact on bank stability within this model. This is interesting as macroeconomic factors are often seen as significant in literature, such as [
33,
71], who discuss macro-financial stability in crises. On the other hand, the banks’ stability is more closely tied to its internal management and the quality of its loans than to external economic conditions. This idea is supported by broader research into intellectual capital, which shows that human, structural, and relational capital directly affect bank stability more than external factors do [
48].
This research compares the non-significant impact of COV on bank stability, with studies such as [
33,
34] discussing the impact of the COVID-19 pandemic on bank stability, with the implication that the pandemic had a noticeable effect on the financial sector. The non-significant impact in our results suggests that banks in this research had stronger resilience or better management practices to buffer against the pandemic’s effects. Refs. [
71,
72] reflect on macro-financial stability during COVID-19, and they likely predict a significant impact of the pandemic on bank stability. The discrepancy with the provided table’s results might indicate that the banks in the research were insulated from some of the broader economic shocks, possibly due to specific regulatory environments, risk management strategies, or government interventions.
Table 8 displays the second model, which analyses the regression results for the three models in the research on the Saudi banking sector; we begin with the model without fixed and random effects. In this model, intellectual capital components such as HCE and CEE display positive coefficients, with CEE nearing significance with a
p-value of 0.069. However, the SCE is not significant. LEV presents a significant negative coefficient, implying an inverse relationship with bank stability. The other control variables, including SIZE, OPEF, GGDP, INF, and the impact of COV, are not statistically significant. The model’s explanatory power, as indicated by the R-squared, is relatively low at 25.92%, and the adjusted R-squared is at 19.25%, suggesting that other unobserved factors may be influencing bank stability. The Durbin–Watson statistic is at 0.110087, suggesting potential autocorrelation issues, and the F-statistic is significant, suggesting that the variables are jointly significant.
The fixed-effects model adjusts for unobserved heterogeneity within entities (banks). Here, HCE becomes significant with a negative coefficient, and CEE remains significant with a positive coefficient, reinforcing the importance of efficiently managing capital employed for bank stability. LEV’s negative influence on stability is pronounced and highly significant, a consistent theme from the previous model. SIZE and OPEF show different coefficients in signs from the first model. However, the OPEF is significant while SIZE is still insignificant, indicating their relationship to stability may be influenced by entity-specific characteristics. The adjusted R-squared value increases dramatically to 99.05%, implying that much of the variability in bank stability is explained when controlling for individual bank effects. The Durbin–Watson statistic improves to 1.689428, and the F-statistic remains significant, underscoring the model’s robustness.
CEE retains its significance and positive relationship with bank stability in the model with cross-section random effects. Interestingly, LEV’s negative coefficient is consistent across all models, affirming the adverse impact of higher leverage on stability. SCE becomes significant, suggesting that when random effects are considered, the efficiency of structural capital plays a role in bank stability. SIZE is still significant, with a negative coefficient, while OPEF is insignificant, like the fixed-effects model. The adjusted R-squared is not provided, but the R-squared remains high, and the F-statistic is significant, indicating the model’s goodness of fit. The Durbin–Watson statistic indicates a potential for autocorrelation, like the fixed-effects model.
The output of the Husman test in
Table 9 demonstrates that the preference would be towards a random-effects model rather than a fixed-effects model, and the null hypothesis that the individual bank effects are not correlated with the other explanatory variables in the model is not rejected. The random effects effectively capture the individual differences among banks, preventing any correlation with the predictors.
Analyzing the random effects results presented in
Table 8, we observe that the coefficient for HCE is a negative sign and significant (Prob. = 0.0223), indicating that higher human capital efficiency is associated with lower financial stability. This result is counterintuitive, as H1 hypothesizes a positive relationship, indicating that human capital in the banking sector of Saudi Arabia may need help with challenges that reduce stability, such as skill mismatches or inefficient use of talent. The negative impact of SC and HCE contradicts findings from several studies, such as those by [
10,
56], which generally find positive associations between various components of intellectual capital and bank stability.
SCE is negative and statistically significant (Prob. = 0.0049), indicating that an increase in structural capital efficiency negatively impacts the financial stability of banks. This finding is unexpected and contradicts hypothesis H1b, which anticipated a positive influence. This may imply that within the context of this research, increases in structural capital could be more effectively utilized or may reflect an overemphasis on the non-human elements of intellectual capital. This aligns with the findings of [
1], but contradicts those of most studies, including those by [
3,
73].
CEE shows a positive coefficient with high statistical significance (Prob. = 0.0000), which supports H1c, indicating that efficient use of financial and physical resources contributes positively to financial stability. The positive relationship between CEE and bank stability aligns with [
17,
22], suggesting that efficient capital deployment contributes to stability and is consistent across different geographies and banking systems.
SIZE is slightly positive but not statistically significant (Prob. = 0.1135), indicating that the size of the bank does not clearly impact financial stability in this model. OPEF’s negative coefficient is significant (Prob. = 0.0001), suggesting that higher operational costs are negatively associated with financial stability. The significant negative relationship between OPEF and bank stability aligns with conventional financial theory, which suggests that inefficiencies in [
21,
37].
LEV is negatively associated with financial stability and is highly significant (Prob. = 0.0000), indicating that more leveraged banks are less stable. The negative impact of leverage on stability is a well-established finding in the literature on financial stability, supported by studies like [
39,
70], which discuss the risks associated with high leverage.
Both GGDP and INF are not statistically significant (Prob. > 0.05), suggesting they do not significantly impact bank stability in this model. The non-significant effect of macroeconomic variables (GGDP and INF) contrasts with studies like [
33,
34], which reflects a disconnect between broader economic trends and the financial stability of banks in the context.
COV is not significant (Prob. = 0.1161), suggesting that the model does not find a direct impact of the COVID-19 pandemic on the financial stability of banks within the scope of this research. The non-significant impact of COV is notably different from the findings of several studies, such as [
71,
72], which discuss the profound impact of the pandemic on financial systems. This might suggest that the specific banks in this research were insulated from the pandemic’s worst effects or that other variables are capturing the pandemic’s impact.
The empirical analysis of the data in
Table 6 and
Table 8 highlights the complex role of intellectual capital in banking stability. While intellectual capital bolsters stability, the structural and human capital components may have intricate effects that warrant further investigation. Operational efficiency and leverage are critical, with inefficiencies and high leverage posing risks to stability. The influence of bank size and broader economic conditions, including the impact of the COVID-19 pandemic, appears to be less pronounced than expected. This nuanced understanding underscores the multifaceted nature of factors contributing to banks’ financial resilience and suggests areas for future research to unpack these dynamics more fully.