Causal Relationship Among Bank Capitalization, Efficiency, and Risk-Taking in ASEAN Commercial Banks

Abstract

Purpose – This chapter explores the answer to the question of whether bank capital is sufficient to absorb risk while maintaining efficiency in ASEAN countries, a new emerging part of the globalized banking system. Design/methodology – This chapter focuses on three objectives: first, to investigate the contemporaneous interactions of capital, risk, efficiency; second, to determine directional Granger causality of the relationship; third, to adopt a new panel vector autoregression to track the explanatory power of causation through the impulse-response functions and variance decompositions. Results – This chapter contributes to literature through providing evidence on the causality of bank capital on cost efficiency and bidirectional causal interactions of bank capital and risk. Better capitalization induces the improvement in efficiency in ASEAN commercial banks even with different ownership, size, and across pre- and postcrisis period. Contribution – This chapter is perhaps the only study so far to investigate the dynamic causality among capital, risk, and efficiency taking into account the sensitivity of the interactions to influential factors of ownership, size, and crisis in ASEAN region—an emerging player in the global banking system.

1. Introduction

Banking sector over the world has altered significantly since the twenty-first century. As part of the globalization, the banking sector is transformed into stronger consolidation, increasing competition, stringent regulation, and innovation. The global financial crises coupled with complex financial transactions and extensive banking integration create pressure for banks in managing risks effectively but maintaining efficiency. The trade-off of bank risk, capital, and efficiency has been the interest of many studies [1, 2, 3, 4]. Some studies show that highly cost-efficient banks experience an increase in nonperforming loans [2, 3, 5, 6], while others indicate different results. As the theoretical literature does not suggest a conclusive relation and empirical evidence does provide diverse results for the nexus between bank risk and efficiency, there is not much comprehensive evidence for the causal, dynamic relationship between risk and efficiency. It is interesting to observe whether the relationship among risk and efficiency is causal and the impact of such causation is temporary or of long-term nature. The work of [7] provides further evidence on the dynamic interactions between risk and efficiency in a more comprehensive way of identifying shock of one variable on the variability of another variable for European banks.

Despite apparent interest of researchers on the interconnectedness of bank risk and efficiency, literature on directional causation with role of bank capital in the relation among risk and efficiency remains scant. Bidirectional causality of risk and efficiency is found in US banks [8], but in European banks, causality running only from risk to efficiency [6, 9] and efficiency causes improvement in capital [10]. Few studies look at the interrelationship of capital, risk, and efficiency in emerging market [11, 12, 13], but causality is not investigated. The recent study of [14] in 2021 also evaluates the contemporaneous causal relationship in Indian banks and finds out the causality from inefficiency to bank risk as well as capital to efficiency. However, majority of studies have not addressed the components explaining the causes of changes in the variables. In addition, no prior study investigated the sensitivity of the interactions to certain influential factors of ownership structure, size, or crisis.

The aim of this study is to fill the gap in the literature and to provide a thorough assessment with regard to the causal relationship between risk, capital, and bank efficiency for commercial banks in ASEAN using the panel Vector Autoregression (VAR) analysis to allow for dynamic changes among endogenous variables. The study focuses on the two questions: How do bank capital, risk, and efficiency respond dynamically to their own and other variables’ shocks? What are the primary shocks that explain the variability in each variable of capital, risk, and efficiency?

The paper contributes to literature in several ways. First, the study includes bank capital in the dynamic intertwining among risk, capital, and efficiency through a detailed assessment of a novel panel VAR techniques, which is not popularly researched in literature. Second, this is the first study to look at the causal relationship of the three variables for commercial banks in ASEAN by investigating the response of one variable upon the exogenous shock in another variable. Third, the study looks at ASEAN region, an increasingly important player in the global economy with distinctive feature. The ASEAN banking sector increases its integration into the global financial system through the formation of ASEAN banking integration framework (ABIF) to support economic growth and enhance financial inclusion. The integration process of ABIF results in more competitive banking sector, forcing banks to increase capital bases, controlling risk, and focusing on efficiency. Therefore, the region becomes an interesting place for empirical research, which can provide inference for the bank management, regulators, as well as the whole market.

The rest of the chapter is structured as follows: Section 2 presents a literature review, Section 3 covers the hypotheses and research methodology. Section 4 deals with data and data description. Analysis of result is explained in Section 5, and Section 6 concludes the paper.

2. Literature review

The theory governing the intertemporal causal relationship among bank capital, risk, and efficiency is initialized by Berger & De Young [8] with four hypotheses, namely “bad luck,” “bad management,” “skimping,” and “moral hazard.” Berger and De Young posit that the four hypotheses can occur concurrently, and they imply the different behavior of banks. Under “bad luck” hypothesis, external shock might cause non-performing loans to increase. Thereby banks react by incurring additional costs to monitor and work out with delinquent loans resulting in decrease in cost efficiency. Bad management hypothesis assumes that low cost efficiency is the result of poor management practices. Inadequate credit scoring, loan monitoring and controlling are caused by bad managers, leading to mounting problem loans. Both bad luck and bad management hypotheses predict negative association of non-performing loans and cost efficiency. Skimping hypothesis refers to the trade-off between short-term operating costs for long-term loan performance. A bank may increase cost efficiency through declining cost of credit appraisal, monitoring and controlling loans but at the expense of long-term problem loan portfolio. Skimping hypothesis predicts positive relationship between efficiency and problem loans. Moral hazard hypothesis does not imply the direct association of problem loans and efficiency. Bank managers, particularly weakly capitalized banks, can take excessive risk, given risk can be borne by creditors, thereby increasing the level of bad loans. Moral hazard can have impact on the above three hypotheses.

Berger & De Young [8] analyze US commercial bank data and find evidence that supports bad luck and bad management hypotheses. The interrelationship between efficiency and loan quality is two ways: increase in nonperforming loans followed by the decrease in cost efficiency and vice versa. Evidence for skimping hypothesis is observed for only a subset of banks, which are efficient over time. The studies from European banks of Fiordelisi et al. [6] and Williams [9] as well as the work of Prakash et al. [14] in India, support evidence of bad management behavior. Several other studies investigate the intertwinning of capital, risk, and efficiency and find mixed evidences. Kwan and Eisenbeis [3] find that less efficient US banks appear to have low risk, whereas Altunbas et al. [1] find the contrasting evidence in European banks where inefficient banks have higher capital and lower risk level. Negative association of risk and efficiency found in the studies of Bitar et al., Louati et al., Nguyen and Nghiem [11, 12, 15], and Isnurhadi in Islamic banks [16], whereas a positive relationship existed in the work of Tan and Floros [17], Tahir and Mongid [13], and Manta and Badircea [18]. Those studies do not explore in detail the direction of causation. Moral hazard behavior is found in various studies [8, 13, 19, 20, 21].

Another strand of literature pays attention on the causation in nexus between capital and efficiency. Berger and Bonaccorsi [10] examine the interactions between capital and efficiency and hypothesize that a reduction in capital ratio can improve efficiency because of the decrease in agency cost of external equity financing. This hypothesis is named “agency costs shareholders-managers” by Lešanovská and Weill [22]. Beside, another assumption is proposed that an increase in capital ratio can cause an improvement in efficiency because agency cost helps reduce the conflict between shareholders and debtholders. The assumption is called “agency cost shareholders-debtholders” hypothesis. Empirical evidences of Berger and Bonaccorsi [10], Skopljak and Luo [23], Pessarossi and Weill [24] support agency cost hypothesis where higher capital ratio is associated with improved efficiency.

Berger and Bonaccorsi [10] also look at reverse causality running from efficiency to risk. They suggest the two hypotheses that are “efficiency risk hypothesis” where more efficient banks cause lower level of capital because higher returns can substitute for financial distress risk and “franchise value hypothesis,” where efficient banks maintain high capital to preserve their economic rents. The results from studies of Berger and Bonaccorsi [10], Williams [9], and Kwan and Eisenbeis [3] show dominant substitution effect of the efficiency risk hypothesis, whereas the franchise value hypothesis is found in small banks. The result of Bagntarasian and Mamatzakis [25] finds evidence of structural breakpoint in the relationship between capital and efficiency. Evidence that supports franchise value hypothesis is found in low-efficiency banks, whereas evidence from high-efficiency banks supports the efficiency risk hypothesis.

Koutsomanoli and Filippaki [7] adopt a panel VAR framework to closely investigate the complex causal relationship of risk and efficiency. For a sample of European banks, they observe that the impact of inefficiency on risk is small and short-lived, whereas the reverse effect is negative and significant. Other study by Jouida [26] also adopts the same panel VAR framework to examine the causality of bank capital and systemic risk in French market. The study finds a negative bidirectional relationship of capital and systemic risk. Further study of Bagntarasian and Mamatzakis [25] explores the nexus of capital buffer, Zscore, and performance using dynamic panel threshold analysis. They focus on testing the efficiency risk hypothesis and franchise value hypothesis and find the evidence of efficiency’s impact on capital buffer and risk.

Yet, no study so far has addressed the combined causality and directional interactions among the three factors of capital, credit risk, and efficiency and explained the different behavior among banks within this nexus. Therefore, further investigation can provide evidence on the underlying nature of the directional impact of one component on another. This chapter reveals the dynamic interactions among risk, efficiency, and capital through investigating primary shocks that cause the variability in each of the three variables. This type of comprehensive assessment of dynamic relationships can provide evidence that explains the different results in literature, particularly in ASEAN region where behavior of banks is not thoroughly studied in order to provide useful implication for management and regulators.

3. Methodology

Literature on the nexus of bank capital, risk, and efficiency has considerable growth. However, there is paucity in the studies that investigate the directional causality of triumvariate relationship [6, 8, 10, 22, 24]. Yet the focus of those studies is still on developed economies. Few studies address the dynamic intertwinning of the three variables through the use of novel vector autoregression method for panel data (PVAR) [7, 25]. However, there has not been any study looking at the comprehensive causality of the three contemporaneous factors of bank capital, risk, and efficiency using the same PVAR framework. This method combines the traditional vector autoregression with panel data technique to encounter the issue of endogeneity while allowing for inclusion of fixed effects in the model [27]. Apart from analyzing causality, the method also helps identify the response of a factor from the change in another and variance is decomposed to analyze the percentage of explanation by each component.

3.1 The hypotheses

We hypothesize the dynamic interdependencies among risk, efficiency, and capital. The relations between bank efficiency and risk can be either positive or negative and causal direction can be two-way.

Hypothesis 1a. Bank risk causes a change in bank efficiency.

Hypothesis 1b. Bank efficiency causes a change in bank risk.

Hypothesis 1a with a negative sign is the bad luck hypothesis. Hypothesis 1b with a negative sign is the bad management hypothesis. Hypothesis 1b with a positive sign is the skimping hypothesis. [8].

The level of capital and hence bank leverage affects efficiency because of agency costs, which arise from conflicts of interest between shareholders and managers or between shareholders and debtholders.

Hypothesis 2a. Bank capital causes a change in bank efficiency.

Hypothesis 2b. Bank efficiency causes a change in bank capital.

Hypothesis 2a with a negative sign is the agency costs shareholders-managers hypothesis: high equity capital and less pressure of debt give managers more cash to invest and may lead to wasteful investment, lowering efficiency. Hypothesis 2a with a positive sign is the agency costs shareholders-debtholders hypothesis: if shareholders are tempted to maximize their value at the expense of debtholders, a higher capital ratio will reduce agency costs and thus increase efficiency [22].

Hypothesis 2b with a negative sign is the efficiency-risk hypothesis: more efficient banks can generate higher returns, which can partially substitute for equity capital to protect banks in the event of financial distress. Hypothesis 2b with a positive sign is the franchise-value hypothesis: efficient banks will maintain a high capital ratio to protect the franchise value associated with high efficiency [10].

Bank decisions on the levels of risk and capital are interrelated, as both decisions are affected by leverage, deposit insurance, and regulation.

Hypothesis 3a. Bank capital causes a change in bank risk.

Hypothesis 3b. Bank risk causes a change in bank capital.

Negative relationship between capital and risk indicates the moral hazard behavior as low capitalization Granger causes high nonperforming loans, because managers have less capital to lose in the event of default, and they benefit from higher returns on risky investments [8]. Causation of bank risk on bank capital supports the regulatory hypothesis where regulators require banks to hold capital commensurate with their risk [28], so an increase in the risk of problem loans can force managers to replenish bank capital [9].

The baseline analysis is extended in several dimensions to study whether these hypothesized relationships are stronger for certain types of banks. “Cost skimping” is expected to occur in efficient banks [8]. Managers of efficient banks are tempted to pursue expansion through investing in risky assets or controlling cost to attain higher efficiency, thereby increasing risk. Therefore, we look at subsamples of high- and low-efficiency banks to analyze the difference in behavior of the two groups of banks.

We extend our work to assess the relationships among risk, capital, and efficiency with subsamples of different size, ownership structure, and pre- and postcrisis periods. Literature finds the impact of market structure on bank risk taking [29, 30, 31] and efficiency [4, 17, 32, 33]. Large banks span their operations in both domestic and international markets, thus having a diversified portfolio. Small banks operate in smaller geographical or regional areas, so they have limited power in the market [34]. Large banks can improve efficiency and reduce risk, as they benefit from economies of scale and portfolio diversification [15, 17, 35]. Large banks also have easier access to the capital market and thus can operate with proportionately smaller capital [19, 36, 37]. In addition, several studies have found differences in the behavior of banks with different ownership. Udell et al. [38] find that on average foreign banks perform less well than private domestic counterparts in developed countries, while reverse result is found in developing countries [39]. Finally, the 2008 global financial adversely hit the global banking sector and ASEAN banking system experiences the same effect. Banks may behave differently in response to the global crisis; therefore, subsamples of the two different periods of before and after crisis need further investigation.

3.2 The model

We model the contemporaneous relationship of capital, risk, and efficiency using panel vector autoregration (VAR), initially developed by [40] and subsequently elaborated by [41]. Panel VAR treats all variables as endogenous and can capture their dynamic interdependencies. Impulse response functions (IRFs) identify the reaction (response) of one variable to a shock (impulse/innovation) in another variable while holding all other shocks at zero [40]. This process can explain the underlying causality among endogenous variables of the model.

The system of simultaneous equations from the model is regressed using system-based GMM. Dynamic simulations are implemented involving the estimation of impulse response functions and variance decompositions [7]. There is a key identifying assumption in setting the order of variables: variables that occur earlier in the ordering affect the following variables contemporaneously, while variables that appear later in the ordering affect previous variables only with lag [40]. This sequential order is a preferred identification strategy, which is referred to as Choleski decomposition. In this study, we make an assumption that risk could be more exogenous in relation to the other variables. An economic shock, or “bad luck” [8], immediately increases nonperforming loans, which proxy for risk. Banks respond by adjusting their capital cushion. Bank capital directly affects costs by providing a source of funding other than deposits [42]. Hence, capital and cost efficiency come later in the order.

We specify the three-variable VAR model as follows:

$${\mathit{CE}}_{i,t}={\alpha }_{10}+{\beta }_{11}{\mathit{\text{RISK}}}_{i,t-1}+{\beta }_{12}{\mathit{CAP}}_{i,t-1}+{\beta }_{13}{\mathit{CE}}_{i,t-1}+f_{1i}+d_{1c,t}+{\epsilon }_{1i,t}$$E1

$${\mathit{CAP}}_{i,t}={\alpha }_{20}+{\beta }_{21}{\mathit{\text{RISK}}}_{i,t-1}+{\beta }_{22}{\mathit{CAP}}_{i,t-1}+{\beta }_{23}{\mathit{CE}}_{i,t-1}+f_{2i}+d_{2c,t}+{\epsilon }_{2i,t}$$E2

$${\mathit{\text{RISK}}}_{i,t}={\alpha }_{30}+{\beta }_{31}{\mathit{\text{RISK}}}_{i,t-1}+{\beta }_{32}{\mathit{CAP}}_{i,t-1}+{\beta }_{33}{\mathit{CE}}_{i,t-1}+f_{3i}+d_{3c,t}+{\epsilon }_{3i,t},$$E3

where CE_i,t, CAP_i,t, and RISK_i,t represent cost efficiency, capital, and risk respectively. f_i denotes fixed effects that allow for individual heterogeneity. d_c,t are time dummies.

Equation (1) tests the impacts of risk and capital on cost efficiency. The estimated coefficients for risk (${\beta }_{11}$) and capital (${\beta }_{12}$) constitute evidence for the bad luck hypothesis (H1a), agency cost, the shareholders-managers/shareholders-debtholders hypothesis (H2a).

Equation (2) examines the effects of risk and cost efficiency on capital. The estimated coefficients for cost efficiency (${\beta }_{23}$) and risk (${\beta }_{21}$) are used to test the efficiency-risk hypothesis/franchise value hypothesis (H2b) and the regulatory hypothesis (H3b).

Equation (3) investigates the impact of capital and cost efficiency on risk. The coefficients for cost efficiency (${\beta }_{33}$) and capital (${\beta }_{32}$) provide evidence of the bad management/cost skimping hypothesis (H1b) and the moral hazard/regulatory hypothesis (H3a, H3b).

We also introduce fixed effects, denoted fi in the model above, to allow for individual heterogeneity in variables [42]. The fixed effects are correlated with the regressors because the dependent variables are lagged, so in order to create unbiased coefficients, we need to eliminate fixed effects by using forward mean differencing, known as the Helmert procedure [43]. This process removes the forward mean and preserves the orthogonality between transformed variables and lagged regressors. So we can use the lagged regressors as instruments in our system GMM regression.

In addition, country-specific time dummies, d_c,t, are included to capture the country-specific macroeconomic variables. These dummies are eliminated by subtracting the means of each variable calculated for each country-year.

We analyze the impulse response functions (IRFs) to examine the reaction of one variable to a shock in another to infer the causality of the variables. Monte Carlo simulations method is used to derive a draw of coefficients. Then the matrix of variance decompositions is determined to explain the cumulative percentage of variation in a variable explained by the shock in another.

3.3 Data and variables

Our data comprise 1404 observations of 146 commercial banks in Thailand, Malaysia, the Philippines, Indonesia, and Vietnam for the period from 2005 to 2015 in an unbalanced panel data. Banks’ financial statements were obtained from Bankscope. To measure bank risk, we use the ratio of loan loss provisions (LLOSS) to loans to be the proxy for credit risk. This measure is commonly used to account for bank risk [6, 9, 15] as it focuses on credit risk and derives from accounting data. The ratio of equity to total assets is used as a measure of bank capital (CAP). This widely used proxy captures the bank’s financial cushion to absorb loan losses [6, 8, 9]. To measure bank efficiency, we opt to use cost efficiency (CE) determined by stochastic frontier approach, which is widely used in literature [14, 44].

4. Empirical evidence

4.1 Result of full sample

We report the parameter estimates of risk, capital, and efficiency in a system of equations that account for fixed effects and country-time effects (see Table 1). The relationship of risk and cost efficiency is not evidenced as the coefficient is not significant. The small but significant and positive coefficient of L.CAP indicates that higher capital ratio can improve cost efficiency as larger capital reduces the conflict between shareholders and debtholders, thereby lower agency cost [22]. In the five ASEAN countries, commercial banks that are better capitalized appear to be more efficient. This finding is consistent with results from the study of Tahir and Mongid [13] in ASEAN region and Prakash et al. [14] in Indian banks. Evidence of reverse causation from efficiency to capital is not supported. Lagged capital significantly decreases credit risk as proxied by loan loss; and vice versa, risk decreases capital significantly (at the 5% level). This bidirectional relationship supports moral hazard behavior also found by other researchers [1, 8, 45].

Dependent variable	Independent variable	Coefficient
CE	L.LLOSS	0.000; (0.572)
	L.CAP	0.002^***; (0.000)
	L.CE	0.417^**; (0.014)
CAP	L.LLOSS	−0.157^**; (0.037)
	L.CAP	0.626^***; (0.000)
	L.CE	−21.66; (0.458)
LLOSS	L.LLOSS	0.797^***; (0.000)
	L.CAP	−0.059^***; (0.000)
	L.CE	−4.476; (0.530)

Table 1.

Results of full sample.

Note: CE, LLOSS, CAP stand for cost efficiency, credit risk, and capital ratio respectively. L.LLOSS, L.CAP, L.CE are lagged value of the three variables. p-value reported in parentheses.

***, **, and * indicate 1, 5, and 10% significance levels respectively.

To confirm the effects of the three variables, we look at how one variable responds to the shock in another variable. The graphs of impulse-response functions and variance decomposition (Figure 1) can help explain those relationships.

Figure 1.

Impulse-response functions for risk (LLOSS), capital (CAP), and cost efficiency (CE).

Row 2, column 1 of Figure 1 displays the response of cost efficiency to a shock in capital, confirming the result of Table 1 in visual form. The positive response of cost efficiency to the impulse in capital supports the agency cost shareholders-debtholders hypothesis. The response reaches the peak in year 2 and then reverts to zero after 10 years.

Row 2, column 3 of Figure 1 shows the response of credit risk to the shock in capital. The response is negative and significant, bottoming out in year 3 and converging to equilibrium after 10 years. Large confidence interval after year 3 suggests cautious conclusion on long-term causality. The result supports moral hazard theory where capital can have impact on risk-taking. Row 3, column 2, depicts the response of capital to a shock in risk. Capital appears to decrease following an increase in credit risk. This negative influence bottoms out after 2 years and then reverts to zero. We observe a bidirectional causal relationship between capital and risk from the IRF graph, confirming evidence of moral hazard behavior.

The variance decomposition (VDC) analysis reported in Table 2 shows the percentage of variation in one variable that is explained by the shock in another variable.

Full sample	LLOSS	CAP	CE
LLOSS	90.35%	9.26%	0.39%
CAP	2.78%	93.16%	4.06%
CE	2.69%	12.25%	85.06%

Table 2.

Variance decompositions for the full sample.

Note: Each cell indicates the percentage of variation in the row variable over a 10-year period explained by a shock to the column variable.

Twelve percent of variation in cost efficiency is explained by the shock in capital and only 2.69% of cost efficiency justified by risk. The variance decomposition of cost efficiency confirms again the hypothesis of agency cost shareholders-debtholders. The explanatory power of capital on variation in risk is 9%, whereas the explanation of efficiency on risk is negligible. This result suggests that banks with low capitalization take on higher credit risk due to nonperforming loans, implying the moral hazard behavior. Credit risk and cost efficiency explain only 2.78 and 4.06% of the variation in capital. The significant influence of bank capital on credit risk and cost efficiency advocates the enhancing capital base of banks from regulatory capital requirement as enforced by international Basel standards.

4.2 Result of subsamples

In what follows, we divide our data two subsamples, one of banks with lagged cost efficiency higher than the median, and one of banks at or below median efficiency.

Table 3 presents results for subsamples of high- and low-efficiency banks. In the efficiency equation, capital causes an increase in cost efficiency, but the result is significant only for the group of low efficiency banks. The result of capital equation shows contrasting relation between efficiency and capital for the two groups. A shock in cost efficiency results in increase in capital of low-efficiency banks implying evidence of franchise value hypothesis. Low-efficiency banks tend to preserve their franchise value generated from lower returns to protect the banks from financial distress; hence, they will not assume the risk of lowering their capital base. Nevertheless, in high-efficiency banks, increase in efficiency causes a decrease in capital, supporting efficiency risk hypothesis. High efficiency banks can take on higher leverage and maintain less capital because of lower expected costs of financial distress. These findings confirm the breakpoint in the association of capital and efficiency between high-efficiency and low-efficiency banks in the study of Bagntarasian and Mamatzakis [25] in European banks. The risk equation shows a significant negative causation of capital on risk for both subsamples. The lagged cost efficiency has negative coefficient but significant only at 10% suggesting weak evidence of bad management behavior for both high and low efficiency banks.

Dependent variable	Independent variable	High-efficiency banks		Low-efficiency banks
		Coeff	p-value	Coeff	p-value
CE	L.CAP	0.001	0.00	0.00	0.01^***
(1)	L.CE	0.64	0.00^***	0.15	0.25
CAP	L.LLOSS	−0.09	0.29	−0.07	0.37
(2)	L.CAP	0.61	0.00^***	0.53	0^***
	L.CE	−32.8	0.04^**	53.1	0.00^***
LLOSS	L.LLOSS	0.86	0.00^***	0.67	0.00^***
(3)	L.CAP	−0.03	0.04^**	−0.08	0.01^***
	L.CE	−7.47	0.08^*	−14.3	0.06^*

Table 3.

Result for subsamples of high- and low-efficiency banks.

***, **, and * indicate 1%, 5%, and 10% significance levels, respectively.

An impulse response function graph may help explain the impact of cost efficiency on bank capital (Figure 2).

Figure 2.

Impulse response functions for subsamples. Subsample of high-efficiency banks. Subsample of low-efficiency banks.

The response of capital to a shock in cost efficiency is displayed in Row 1, column 2 of Figure 2. Contrasting responses are shown in the graphs of the two subsamples. Visual evidence in high-efficiency banks confirms the efficiency risk hypothesis where the shock in efficiency causes negative response in capital and the effect reaches the trough after 3 years and reverts to zero. High-efficiency banks expect high earnings from better efficiency to substitute for equity capital in the event of financial distress [10]. For low-efficiency banks, following the shock in cost efficiency, banks respond by an increase in capital with peak reached in 2 years and subsequently converge to equilibrium after 10 years. This result provides visual evidence for franchise value hypothesis. Low-efficiency banks tend to protect their economic rent by increasing capital.

The result of variance decomposition of the two subsamples in Table 4 is consistent with findings from IRFs. In the group of high-efficiency banks, capital shock can explain 19.72% of the variation in cost efficiency. The franchise value hypothesis has strongly evidenced in VDC analysis of low-efficiency banks with 20.49% explanatory power of cost efficiency on capital. On the contrary, in high-efficiency bank subsample, evidence of efficiency risk hypothesis is weaker with only 8.27% explanation. Bad management behavior is apparently evidenced in low efficiency banks with 12.59% risk variation explained by cost efficiency as compared to only 3.61% in the group of high-efficiency banks. Capital explains 11% of forecast error variance in risk for high-efficiency banks but only 8% for low-efficiency banks. The result supports the moral hazard hypothesis.

Variable	High-efficiency banks			Low-efficiency banks
	LLOSS	CAP	CE	LLOSS	CAP	CE
CE	2.3%	19.7%	78%	2.7%	3.7%	93.6%
CAP	0.1%	91.7%	8.3%	7.2%	72.4%	20.5%
LLOSS	85.3%	11.1%	3.6%	79.1%	8.3%	12.6%

Table 4.

VDC for subsamples of high- and low-efficiency banks.

4.3 Sensitivity analysis

The empirical study is extended to how the causal relationship of capital, risk, and efficiency is sensitive to the differences in ownership type, size of banks, as well as the 2008 global crisis.

4.3.1 Results for foreign banks and domestic banks

Literature suggests that the behavior of banks can alter with different types of ownership. Therefore we divide the sample into two groups of domestic and foreign banks. Table 5 reports parameter estimates of the model for the subsamples of foreign banks and domestic banks. There exists weak evidence of bad luck in foreign banks. The results indicate positive causation running from capital to efficiency, which supports agency cost shareholders-managers hypothesis for both foreign and domestic banks. However, the reverse causation is different among the two groups. In foreign banks, higher efficiency causes an increase in capital while efficiency in domestic banks causes a decline in capital. Bank capital negatively causes risk regardless of type of ownership, which is consistent with regulatory hypothesis.

Depen-dent variable	Indep- endent Variable	Foreign banks		Domestic banks		Large banks		Small banks
		Coeff	p-value	Coeff	p-value	Coeff	p-value	Coeff	p-value
CE
	L.LLOSS	−0.00	0.09^*	0.00	0.22	0.00	0.68	0.00	0.98
	L.CAP	0.00	0.00^***	0.00	0^***	0.00	0.01^***	0.00	0^***
	L.CE	0.13	0.66	0.60	0^***	0.74	0^***	0.39	0.05^**
CAP
	L.LLOSS	0.11	0.61	−0.12	0.10	−0.24	0.02^**	−0.07	0.61
	L.CAP	0.56	0^***	0.61	0^***	0.65	0^***	0.56	0^***
	L.CE	84.5	0.07^*	−40.9	0.08^*	−41.7	0.02^**	8.88	0.85
LLOSS
	L.LLOSS	0.62	0^**	0.84	0^***	0.75	0^***	0.78	0^***
	L.CAP	−0.05	0.1^*	−0.07	0^***	−0.17	0^***	−0.05	0.01^***
	L.CE	−5.05	0.77	−0.46	0.92	−8.69	0.23	−7.23	0.42

Table 5.

Sensitivity analysis for subsamples.

***, **, and * indicate 1, 5, and 10% significance levels, respectively.

The IRF graphs (Figure 3) confirm the regression results. The effect of one standard deviation shock in risk on cost efficiency is negative, bottoming out within a year and reverts to zero as shown in row 3, column 1. Row 2, column 1 indicates the shock of capital on efficiency with longer impact of more than 2 years for domestic banks as compared to 1-year influence in foreign banks. Different responses of the two groups are visualized in row 1, column of Figure 3 as foreign banks have higher capital in a year and domestic banks reduce capital in 2 years following an efficiency shock.

Figure 3.

Impulse response functions for foreign and domestic banks. Subsample of foreign banks. Subsample of domestic banks.

Variance decompositions for the two groups of foreign and domestic banks are displayed in Table 6. In total, 15 and 11% of variations in efficiency are explained by capital for foreign and domestic banks, respectively. The explanatory power of risk in efficiency variance is 7% in foreign banks but only 1% in domestic banks explains 7% of variance in efficiency for foreign banks but just 1% of efficiency variance for domestic banks. Twenty percent of capital variance can be justifìed by efficiency disturbance, whereas the explanation of risk is only 9% in foreign banks and 11% in domestic banks.

	Foreign banks			Domestic banks
	LLOSS	CAP	CE	LLOSS	CAP	CE
CE	7.3%	14.5%	78.2%	1.0%	10.6%	88.4%
CAP	2.6%	77.2%	20.2%	1.6%	78.7%	19.7%
LLOSS	87.0%	9.0%	4.0%	84.6%	11.9%	3.6%

Table 6.

VDC for subsamples of foreign banks and domestic banks.

4.3.2 Results for large banks and small banks

Bank size does influence the relationship of capital, risk, and efficiency. The sample is divided into two groups of large and small banks. Both groups display the same impact of capital on risk and efficiency as shown in Table 5. However, the behavior is different in capital equation as both risk and efficiency negatively cause changes in capital of large banks while such causation is not evidenced in small banks.

Figure 4 reports the IRFs for large and small banks. Row 2 column 1 of Figure 4 reports the capital shock on efficiency. The effect is positive for both large and small banks, but impact on large banks lasts more than 3 years as compared to 1 year in small banks. Large banks also experience a decrease in capital following the shock in efficiency, but such response is not significant in small banks. The negative and bidirectional causal relationship between risk and capital is observed in both large and small banks. Moral hazard behavior is found in ASEAN banks regardless of bank size.

Figure 4.

Impulse response functions for large and small banks. Subsample of large banks. Subsample of small banks.

VDC estimations are presented in Table 7. With regard to the variation of cost efficiency, capital of large banks can explain 14% and capital of small banks explains 10%. Risk can explain only negligible percentage of efficiency forecast error variance. In strong contrast, 43% of capital variation for large banks is explained by efficiency, but the explanatory power is only 1% for small banks. The influence of efficiency on capital is clearly observed in large banks only. That is, efficiency influences capitalization only in large banks. In large banks, capital and efficiency have strong explanatory power over the risk variance, but the response in small banks is opposite.

	Large banks			Small banks
	LLOSS	CAP	CE	LLOSS	CAP	CE
CE	3.6%	14.2%	82.2%	0.7%	10.2%	89.1%
CAP	2.1%	54.4%	43.4%	1.9%	97.6%	0.6%
LLOSS	65.9%	23.6%	10.6%	88.3%	8.7%	2.9%

Table 7.

VDC for subsamples of large banks and small banks.

4.3.3 Results for the precrisis and postcrisis periods

The 2008 global crisis has adverse impact on banking and financial system around the world, and ASEAN region is not an exception. It is important to see the difference in behavior of banks before and after the crisis. To study the impact of crisis on the interactions among capital, risk, and efficiency, subsamples of pre and postcrisis are investigated. Regression results indicate that increased capital causes improvement in efficiency and reduction in risk both before and after the crisis. The impact of cost efficiency on capital displays different causation for the two periods as shown in Table 8. Efficiency helps improve bank capital in the precrisis period. However, such causation is not seen in postcrisis period. After the crisis, banks tend to maintain their capital in order to protect themselves against negative shocks.

Dependent variable	Independent variable	Precrisis		Postcrisis
		Coeff	p-value	Coeff	p-value
CE
	L.LLOSS	−0.0001	0.865	−0.0008	0.373
	L.CAP	0.0013	0***	0.0017	0***
	L.CE	0.2346	0.259	0.4986	0.001***
CAP
	L.LLOSS	0.2105	0.292	−0.0681	0.543
	L.CAP	0.3766	0.002***	0.6221	0***
	L.CE	178.1687	0.002***	−24.8413	0.277
LLOSS
	L.LLOSS	0.7075	0***	0.7890	0***
	L.CAP	−0.0433	0.06*	−0.0510	0.003***
	L.CE	−9.0253	0.468	−1.3368	0.81

Table 8.

Sensitivity analysis for subsamples of precrisis and postcrisis.

Figure 5 displays the IRF graph. Before the crisis, the effect of capital shock on efficiency gradually declined, while after the crisis, the response rose over 2 years and subsequently decreased thereafter. The reverse causation, from efficiency to capital, differs between the two periods with an increase before the crisis but decrease after the crisis.

Figure 5.

Impulse response functions for pre- and postcrisis periods. Precrisis. Postcrisis.

Table 9 reports the variance decomposition. It reveals that 29% of efficiency variance is explained by capital before crisis, but the percentage drops to 12% after the crisis. Before the crisis, efficiency disturbances explain close to 34% of the forecast error variance of capital over 10 years. The shocks in capital and efficiency explain 18 and 13% of the variation in risk in precrisis period. But after the crisis, the explanatory power of capital and efficiency is rather small. The VDC results imply a strong response of the three factors to exogenous shock in precrisis rather than post crisis period.

	Precrisis			Postcrisis
	LLOSS	CAP	CE	LLOSS	CAP	CE
CE	3.9%	29.0%	67.1%	2.9%	11.6%	85.5%
CAP	1.1%	65.1%	33.7%	0.6%	91.3%	8.1%
LLOSS	68.5%	18.5%	13.0%	94.9%	4.9%	0.2%

Table 9.

VDC for subsamples of precrisis and post crisis.

5. Conclusion and policy implications

This chapter empirically investigates the intertemporal and causal interdependencies among bank capital, risk, and efficiency in the five emerging countries in ASEAN region, a growing dynamic part of the global banking system. We summarize our findings as follows. Firstly, our study finds evidence supporting the bidirectional causality between capital and risk in ASEAN commercial banks. The results also confirm that banks with better capitalization are more efficient. Secondly, we observe different behavior between high-efficiency banks and low-efficiency banks. Following the shock of increasing efficiency, the high-efficiency banks tend to maintain low capital indicating evidence of efficiency risk hypothesis, whereas low-efficiency banks increase their capital ratio to protect franchise value. Lastly, sensitivity analysis of causations among the three factors of capital, risk, and efficiency reveals that stronger capitalization helps improve efficiency regardless of ownership, bank size, and pre or postcrisis period.

The study confirms prior research suggesting that capital, risk, and efficiency of ASEAN commercial banks are causally intertwinned. The analysis substantiates the positive impact of capitalization on bank efficiency. We provide new evidence on different behavior among high-efficiency and low-efficiency banks in the ASEAN region. We further contribute to literature on the influence of ownership, size, and period over the trade-off among capital, risk, and efficiency.

The results from this study provide relevant implications for bank managers and regulators. The result of negative causal relationship between capital and risk and positive causality running from capital to efficiency suggests the importance of bank capital in limiting risk-taking and improving bank performance. Thereby, imposing stronger capital requirements under Basel framework by regulators in the ASEAN region can help achieving both lower risk and higher efficiency. As globalization takes place rapidly in the region, a push to the adoption of the international standards on capital requirement from Basel Accord can help banks in ASEAN region to increase capitalization, resulting in improvement in performance and achieving greater competitiveness in the global market.

A limitation of our study is that the results are based mainly on accounting measure of risk, capital, and efficiency. The measure of risk focuses on credit risk. Other types of risks including market risk, operational risk, and liquidity risk are not captured in the model. Future studies may adopt a more comprehensive measure of risk that incorporates different types of banking risks to determine how the nexus changes in response to different risk factors.