The main objective of this research is to investigate the impact of intellectual capital efficiency (ICE) and its components – human capital efficiency (HCE) and structural capital efficiency (SCE) – on bank credit and insolvency risks in the Saudi banking industry. Credit risk is considered one of the oldest risks associated with lending activities, which is the core business of the banking industry. Most of the studies in the literature such as Gigante (2013), Ozkan et al. 2018) address the impact of IC and its components on banks' financial performance.
Data
Dependent Variables
Insolvency Risk: Insolvency is another type of risk that threatens banks and arises when a bank struggles to meet its obligations. In this document, a bank's solvency ratio (SOL) is used as a measure of a bank's insolvency risk in an inverse manner: the higher the ratio, the lower the insolvency risk.
Independent Variables
Where NI stands for the net income of the firm, T stands for the corporate tax, I stands for the company's interest expense, D is depreciation, A stands for amortization, and EC refers to the employee expenses. Human Capital Effectiveness (HCE): According to Edvinsson and Malone (1997), HC is the most critical component of ICE. In fact, HR refers to workers' knowledge, skills, competencies, training, education, experience and expertise that an employee accumulates during his/her time with the organization.
However, in the VAIC model, HC is defined as the wages and salaries of employees in a given period of time (Pulic, 2000); in fact, it is considered an investment of the company (Tan et al calculated HCE as how much added value is generated by one monetary unit invested in human capital. Thus, a higher HCE ratio results from a higher level of VA for a given level of wages and salaries (HC).Structural Capital Efficiency (SCE): Structural Capital (SC) is considered the non-physical supporting infrastructure of an organization that allows human capital to function (Bollen et al., 2005).
In the VAIC model, the structural capital is the difference between the VA and HC, i.e.
Control Variables: For the evaluation of intellectual capital (IC) and its influence on bank risks, two widely used control variables are taken into
- Bank Size: In this paper, the bank size is measured by the natural logarithm of bank total assets (lnTA), which is considered to be the most
- Net Interest Margin to Total Assets (NIM_TA): The second control variable is the net interest margin (𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡 𝐸𝑎𝑟𝑛𝑒𝑑 − 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡 𝐸𝑥𝑝𝑒𝑛𝑠𝑒)
The higher the ratio, the higher the bank's efficiency and the lower the risk exposure for the bank.
Research Methodology and Empirical Results
Unit Root Test
From Table 1, it can be understood that all variables are stationary in their levels, except the net interest margin to total assets (NIM_TA), which became stationary after taking the first difference. Thus, panel techniques are applied using the levels of all variables except net interest margin as a percentage of total assets, which is obtained using its first difference.
Models
- Bank Credit Risk Section
Pooled Model
In this article, econometric panel techniques are used, starting with the pooled model and through random effect and fixed effect models. In the same context, statistical evidence of a negative relationship between the CAR and bank credit risk is observed in the pooled model results above, such that if the CAR increases by one unit, the CR will decrease by 0.001 on average. However, in these types of models we assumed that all banks are the same, which is in fact not true.
To achieve this, we have two options – first, random effect and second, fixed effect models. We do not have the freedom to choose between the two models at will; but there is a test to determine which model is the right one to use.
Random Effect Model
But before we can accept these results, we need to test whether the random effects model is the right model to use over the fixed effects model. The Hausman test has a null hypothesis that states that the random effects model is an appropriate model to use, while the alternative hypothesis. From Table 4, since the p-value is less than 5 percent, we thus determine that the fixed effect model is the correct model among the three models to be used.
Fixed Effect Model
This conclusion is compatible with the findings seen in India (Ghosh and Maji, 2014) and Europe (Kaupelyte and Kairyte, 2016). As done before, a pooled model is used to see if the IC's components, together with other significant variables, have a significant influence on bank credit risk. In the same context, there is a significant negative relationship between the CAR and CR; for example, if the CAR increases by one unit, the CR will decrease by 0.0004.
However, the sign of the structural capital efficiency (SCE) is not compatible with the theory, which states that there is a negative relationship between the SCE and CR. Yet one cannot consider this model as the right model because not all banks have the same structure and their individuality cannot be denied and ignored.
Random effect model
Bank Solvency Section
Pooled model
Similar to the bank credit risk section, this section begins the examination with the overall level of intellectual capital efficiency and uses a pooled model to measure the influence of ICE and other significant explanatory variables on bank solvency. ICE has a significant positive relationship with the solvency of Saudi banks, so that if the ICE increases by one unit, the bank's solvency will increase by an average of 0.14 percent. In addition, an increase in the CAR by one unit will result in an increase in the bank's solvency of 0.014.
Moreover, the net interest margin also has a positive relationship with the solvency of the banks, so that if the net interest margin increases by one unit, the solvency of the banks increases by 0.59 units. Two sides of the same coin, these variables will have a negative impact on banks' insolvency risk. However, as was the case before, we cannot consider this result to be definitive because we assumed that all banks have the same structure, ignoring individuality, which is not always the case.
Random Effect Model in Bank Solvency
Thus, the results from the fixed effect model provide statistical evidence for a positive correlation between all variables except lnTA and bank solvency. Statistically, if IC increases by one unit, the bank's solvency will increase by 0.23 percent. Additionally, if there is an increase in CAR by one unit, there will be an increase.
Furthermore, if there is an increase in the NIM_TA by one unit, there will be an increase in bank solvency by 0.69. Statistically, there is a positive relationship between human capital efficiency (as one of the intellectual capital components) and Saudi bank solvency, so that if there is an increase in the HCE by one unit, there will be an increase in the bank solvency by 0.13 . average. Furthermore, if there is an increase in the structural capital efficiency, there will be an increase in the bank solvency by an average of 0.35.
Moreover, if the CAR increases by one unit, the banks' solvency increases by 0.015 on average. Finally, if the net interest margin as a percentage of the balance sheet total increases by one unit, the banks' solvency increases by an average of 0.58.
Random Effect Model for Banks’ Solvency ( IC’s Components) This model allows all banks to have a common mean value at the
In fact, we cannot take these results as conclusive, since we assumed that all banks have the same human capital efficiency and structural capital efficiency, which is not always the case. Therefore, we should examine the influence of IC components, along with other explanatory variables, using either a random effect or a fixed effect model. Random Effects Model for Bank Solvency (Components of IC) This model allows all banks to have a common average value at.
However, as before, these results cannot be considered definitive because we need to test which model is most appropriate to measure the impact of IC components along with other variables on bank solvency. Thus, the Hausman test is performed again so that we can draw consistent conclusions about the correct model that should be used for this work. From the Hausman test, we reject the null hypothesis that the random-effect model is the correct model to use and accept the alternative hypothesis that the fixed-effect model is the appropriate model for the task.
Fixed Effect Model for Banks’ Solvency ( IC’s Components)
Conclusion
More specifically, we examine how intellectual capital (IQ) and its components (human capital and structural capital) affect both bank credit and insolvency risks. However, most of those studies only investigated the impact of intellectual capital (IC) and its components on a firm's financial performance (profitability), and very few studies attempted to investigate the IC impact on the firm's risk management . Using the VAIC model developed by Pulic (1998) to obtain the effectiveness of IC and its components, and implementing various panel data techniques, the findings of this study conclude that there is a negative relationship between the intellectual capital efficiency and Saudi bank credit risk .
In context, a fixed effect model shows that there is only a negative relationship between the human capital efficiency, as one of the intellectual capital efficiency factors, and Saudi bank credit risk. The other intellectual capital efficiency factor, structural capital, is shown to have a positive sign of the coefficient, which is not compatible with the theory and the literature. In this respect, the model showed that there is a positive relationship between intellectual capital efficiency and the Saudi bank solvency, indicating a negative impact on bank insolvency risk.
The Impact of Intellectual Capital on Organizational Performance: An Empirical Study of Hang Seng Index Companies (Part 1). In 1998, the Austrian Intellectual Potential Team presented at the 2nd McMaster World Congress on the Measurement and Management of Intellectual Capital.
