In this study, we investigate the effect of efficiency and market competition on bank performance in the Middle East and North Africa (MENA) region. We also incorporate in our analysis the effect of the recent financial crisis. The empirical literature on bank risk and performance during the COVID-19 pandemic is growing rapidly. During the initial stage of the crisis in March 2020, there was a massive draw-down of existing credit lines, named “dash for cash” (Acharya and Steffen ; Lei et al. ). Most of the draw-down was done by large companies from large banks (Berger et al. ). In this context, Kapan and Minoiu () find that “banks with greater exposure to credit line drawdown risk reported tighter lending standards on new loans to small and large firms, and reduced the supply of large, syndicated loans and small business loans since March 2020” (p. 40). In a recent paper of Mirzae et al. (), the authors confirm that “efficiency levels are found to have positive effects on bank performance during both normal times and subsequent financial crisis times” (p. 3). Therefore, investigating efficiency incorporating the COVID-19 pandemic effects is important contribution to the existing literature. Show
The impact of efficiency on bank risk and performance has attracted the attention of many researchers. It draws even more attention after the recent COVID-19 pandemic outbreak for the following reasons. First, the recent global financial crisis offers a better quasi-natural experiment than the global financial crisis (GFC) and prior crises to study bank performance since the COVID-19 shock is plausibly exogenous to both the lending institutions and the borrowers. Thus, COVID-19 allows us to study the question of whether the same factors determined bank performance in the recent financial crisis as in the pre-crisis periods. Second, we focus our attention on bank efficiency since we expect efficiency level to significantly impact bank performance during the COVID-19 outbreak. Third, prior research has demonstrated that Islamic banks (IBs) had relatively high efficiency and profitability during periods of financial uncertainty (including COVID-19) than conventional banks (CBs). Therefore, we add to the current empirical literature by addressing the following two questions: (1) Is the bank profitability during the COVID-19 outbreak dependent on efficiency (and competition)? (ii) How different is the impact of bank efficiency (and competition) on the profitability of Islamic and conventional banks? While most of the previous studies investigate the impact of the COVID-19 pandemic on the banking sector, mainly in North America and the European Union, our research targets the MENA region's emerging economies. There are three main reasons for this. First, the MENA region is considered as “a bridge connecting Europe and Asia, and with the oil-rich countries of the Gulf Cooperation Council (GCC), it is one of the richest parts of the world in terms of resources” (Boukhatem and Ben Moussa , p. 233). Second, like any other region in the world, the COVID-19 crisis “is having a significant impact on the economies of the MENA region, leading to the fall in oil production, hospitality and tourism are worst affected and decline in GDP per capita income” (Hassan et al. , p. 53). Third, the link between Islamic finance and long-terms economic growth and stability in the MENA region became an important question for both policy makers and researchers. Another notable aspect is the importance of Islamic finance for economic development in this region, which faces a spectacular growth in recent years. Bitar et al. () claim that “in oil exporting countries, Islamic banks can benefit from oil revenues as a way to generate deposits and channel them to investments in large governmental and infrastructure projects” (p. 22). Therefore, estimating the efficiency of the Islamic banking sector is an important issue for researchers and regulators. We focus our attention on bank efficiency since we expect the level of efficiency to have a significant impact on banking system stability during the COVID-19 outbreak. Our motivation for this research emanates from the fact that existing empirical literature provides mixed results for the impact of efficiency on bank performance. Mirzae et al. () argue that “efficiency levels are found to have positive effects on bank performance during both normal times (Fiordelisi et al. ; Hughes and Mester ) and subsequent financial crisis times (Assaf et al. )” (p. 3). In a single-country study, Cristian et al. () investigate the effects of competition, efficiency, and risk on the performance of banks in Indonesia and find that the efficiency of costs and revenues has no effect on banking performance (measured by ROA and NIM), whereas profit efficiency positively affects the net interest margin (NIM). In a recent study of the stock performance of Islamic banking institutions during the COVID-19 outbreak, Mirzae et al. () found that “better stock performance of IBs during the current health crisis can be attributed to their pre-crisis levels of efficiency” (p. 29). Comparing efficiency across the developed and emerging economies, Saâdaoui and Khalfi () report that “on the one hand, the results of the experiments show that, in the emerging region, there is evidence of a strong linkage between Islamic banking efficiency and gross domestic product. On the other hand, in the developed region, the efficiency is rather based upon Sharia Supervisory Board and board committees” (p. 1). This ambiguous relationship between efficiency and bank performance remains a great puzzle. One of the main reasons is that many other factors affect the link between efficiency and bank profitability. For example, the two most researched factors are the type of banking system (conventional or Islamic) and the level of banking market competition. To fill this gap, we investigate the relationship between efficiency, market power, and bank profitability across different banking systems (Islamic and conventional) using a sample of 225 banks in 18 MENA countries. We find a strong positive association between efficiency and bank profitability as well as the level of banking market competition and performance. The outcomes of the analysis indicate that for banks operating in markets with a high and medium level of competition, striving for high efficiency will decrease banks’ profitability. This effect is not significantly different between the two banking systems but is more enhanced during the COVID-19 pandemic outbreak. For the low level of market competition, the analysis indicates that efficiency level does not influence bank performance and is insignificant for both IBs and CBs. Our results indicate that competitive pressure is not always good for the banking industry. We differentiate from previous research on emerging economies by two means. First, unlike our predecessors who examine a single country’s bank performance (Cristian et al. ; Moudud-Ul-Huq ), we investigate the impact of COVID-19 and efficiency on bank performance in the MENA region. We report that efficiency and market power (measured by the Lerner index) are strongly associated with improved bank performance across the sample. Our empirical results are relatively robust to different bank profitability and efficiency measures. Second, prior research on the link between efficiency and bank performance during the COVID-19 pandemic mainly focuses on stock return as a measure of bank profitability. For example, Mirzae et al. () investigated the stock performance of Islamic banking institutions during the COVID-19 outbreak. They found that “pre-crisis levels of efficiency can explain the higher stock returns for IBs relative to their conventional peers” (p. 1). The higher efficiency level can explain the better performance of Islamic banking institutions, but other factors like market competition and risk behavior are not explored. In contrast to their findings, our evidence show that improved efficiency does not necessarily lead to better profitability of IBs. In contrast to the prior research focusing mainly on bank stock performance during the COVID-19 outbreak (Demirgüc-Kunt et al. ; Mirzae et al. ), we investigate the effect of efficiency on bank profitability using both accounting and market measures. As such, our study contributes strongly to advancing existing empirical literature on bank efficiency and stability. The reported results indicate that efficiency has a strong and positive effect on bank performance, and this effect is not significantly different in the initial phase of the COVID-19 pandemic. Likewise, market competition (measured by the Lerner index) positively influences bank performance in MENA countries. Previous research provides limited evidence on the differential effect of efficiency and competition on IB's performance compared to their conventional peers (Srairi ; Sun et al. ). We are the first to formally test whether efficiency and market competition explain the observed differences in bank performance between the two banking systems. Our results indicate that efficiency and marker competition have a greater impact on the profitability level of CBs than IBs. However, we did not find strong evidence that this impact was augmented during the COVID-19 outbreak. These findings deepen the notion that efficient banks are more resilient during global financial crises and emphasize the importance of bank regulatory reforms that boost efficiency in standing against the negative consequences of the recent (COVID-19) crisis. The reminder of the paper is organized as follows: Sect. contains the literature review and the main hypotheses related to the concentration/efficiency-performance relationship. Section describes the data set and the empirical model. Section provides a detailed analysis of the empirical results and their interpretations. Section includes a robustness check and alternative specifications. Finally, Sect. presents our conclusions and the main direction of future research. 2 Theoretical background and hypotheses2.1 The impact of efficiency on bank performanceAccording to the existing theoretical literature, two main concepts may explain the link between banking market structure and performance. First, competition in the banking industry is frequently mentioned as the main factor in the performance of the banking industry. According to the Structure-Conduct-Performance (SCP) theory, market power or market concentration is the main driving force for the higher performance of a bank (Goldberg and Rai ). Thus, more concentrated banks tend to compete for less to obtain higher profits. The existing empirical research also provides evidence that competition in the banking industry can explain bank performance (García-Herrero et al. ; Chortareas et al. ; Tan ; Fang et al. ). In contrast with the traditional SCP hypothesis, the efficient structure (ES) hypothesis argues that it is superior efficiency rather than collusive behavior, which improves bank profitability (Hannan ). Moreover, the ES hypothesis predicts that in concentrated market structures, the more dominant banks achieve higher profitability with increased efficiency. The literature identifies many elements that are related to bank efficiency, or the production of outputs, using a different set of inputs. Factors that are shown to influence bank efficiency include the size of banks (Camanho and Dyson ), the ownership structure (Altunbas et al. ; Berger et al. ), the age of banks (Fries and Taci ), and deregulation and liberalization (Chen et al. ). Kozak () investigates “the impact of shock increase in the value of non-performing loans on the equity level and profitability of 141 banks in 18 countries of Central Eastern South Europe” (p. 1). The study reports that “better cost efficiency does not appear to be sufficient to improve banks' profitability. In most banks, the net interest margin (NIM), ROE, and ROA ratios are lower than their sample mean values” (p. 8). López-Andión et al. () investigate the issue of “whether the economic crisis that so severely affected Portugal had a significant impact on the effects of securitization on solvency” (p.15) and report a positive relationship of bank efficiency and profitability with the risk-adjusted ROA. In addition, bank efficiency measured by the cost-to-income ratio (CIR) enhances capital requirements and solvency. Saâdaoui and Khalfi () find evidence that “on the one hand, in the emerging regions, there is a strong linkage between Islamic banking efficiency and gross domestic product; on the other hand, in the developed regions with Islamic banking presence, the efficiency is rather based upon Sharia Supervisory Board and board committees” (p. 1). Previous research reports that the bank stability of Islamic institutions is higher than conventional banks in the MENA region (Albaity et al. ). The efficiency of IBs also brings a lot of attention in recent years, but the results are inconclusive. For example, Aghimien et al. () found no difference in efficiency level between IBs and CBs, Kamarudin et al. () reported that IBs are less efficient than CBs, supporting the findings of an earlier study by Isnurhadi et al. (). Next, Saeed and Izzeldin () show “that the relationship between profit efficiency and default risk of banks across the sample, for CBs and for the GCC is such that a decrease in default risk is associated with lower efficiency levels” (p. 1). Moreover, Bitar et al. () report that “Islamic banks underperform their conventional counterparts in more democratic political systems but outperform them in hybrid and Sharia’a-based legal systems” (p. 3). The empirical evidence about the efficiency of IBs during the global financial crises also remains mixed and inconclusive. For example, Hasan and Dridi () demonstrate that Islamic banks had relatively higher efficiency and profitability than their conventional peers during financial crises. Moreover, Belanès et al. () report a negative impact of the subprime crisis on the efficiency of Islamic banks in the GCC region just like their conventional peers worldwide. However, Islamic banking institutions remained almost efficient and witnessed a slight decrease in their efficiency level during the 2007–2008 crisis. In a recent study of 426 banks from 48 countries, Mirzae et al. () documented “the dominance of IBs over CBs in terms of the higher levels of pre-crisis efficiency” (p. 28). To shed some light on this puzzling issue, we further investigate the efficiency of different banking systems in the MENA region by testing a few hypotheses presented in Sect. . 2.2 The impact of the COVID-19 pandemic on banking sector performanceThe COVID-19 pandemic “features the most unanticipated large and widespread exogenous economic shock of all time—it was even more global than the Global Financial Crisis (GFC). To the extent that such shocks are plausibly exogenous, that is, they affect the real economy or financial system, but are not themselves caused by economic or financial forces” (Berger and Demirgüç-Kunt , p. 2), it will be difficult to estimate the overall effect of COVID-19 on the financial system stability. Carletti et al. () claims that the pandemic crisis “hits the real economy directly, combining huge supply and demand shocks. The former is due to disruptions in the global supply chain and the latter—to demand declines because of lockdowns” (p. 1). Furthermore, “as the sources of cash started to deplete, borrowers faced dire situations with decreased ability to make loan repayments. Consequently, credit losses in the form of non-performing loans are expected to increase, and the exposure to credit risk to amplify” (p. 4). On the customers’ side, “the uncertainty about the scale of the pandemic development contributed to a decline in demand for financing investment and current capital, as well as consumer goods and services” (Kozak, , p. 1). According to Mirzae et al. (), “lower consumer spending in retail, and fewer assets under management, are likely to lower fee income for banks. These adverse effects of the crisis on the banks’ balance sheets are further exacerbated by significant increases in operating cost” (p. 4). Likewise, the COVID-19 pandemic had a negative impact on banks’ lending activities. Kozak () witnesses that “the reduction in the number of potential borrowers and the opportunity for granting new loans significantly reduced banks’ interest income, as well as other fees and commissions related to granting loans. As a result, banks weakened their ability to raise equity capital through retaining the net profit and to pay dividends to shareholders, which made it harder for them to attract new capital from the market” (p. 4). Current research finds that the decrease in lending activities observed during the COVID-19 pandemic is contingent on both specific characteristics of the banks (e.g., size) and the external environment’s conditions. For example, based on an analysis of banks in 125 countries during the COVID-19 pandemic, Colak and Oztekin () report that “bank lending is particularly weak among smaller banks and those with lower returns on assets as these banks likely have more funding constraints and heightened credit risk” (p. 4). The role of banks’ size on systematic risk prompted by the COVID-19 crisis was reconfirmed by Borri and di Gorgio (), who reported that “larger banks and banks with a business model more exposed to trading and financial market volatility contribute more” (p. 1). The banks’ financial stability issue during the COVID-19 pandemic also attracted growing interest among the academic community. Extant research “supports the notion that during the financial crisis, capital responses to the risk-taking behavior of banks are not similar to the normal economic conditions” (Mateev et al. , , , p. 2). Moudud-Ul-Huq () finds “that higher capital promotes banks’ financial stability by lessening the risk, and higher risk impedes capital growth. He concludes that during a crisis (COVID-19), bank requires more capital to absorb credit risk, and COVID-19 also hits hard to banks' survivability” (pp. 30–31). In the same context, Li et al. () findings suggest that worse-capitalized banks could not accommodate increased liquidity demands from the crisis. Furthermore, Acharya et al. () show that the COVID-19-related decline in bank stock prices was particularly strong for worse-capitalized banks. Finally, Danisman () demonstrates that “equity markets of countries with stricter regulatory requirements on capital and liquidity tend to be more resilient to COVID-19” (p. 1). Indeed, in the context of increased risk caused by the pandemic, Aldasoro et al. () report that banks during the COVID-19 outbreak banks “underperformed significantly relative to other sectors. Still, markets showed some differentiation by bank nationality, and credit default swap (CDS) spreads rose the most for those banks that had entered the crisis with the highest level of credit risk” (p. 1). A recent study by Li et al. () investigated “the effect of the COVID-19 pandemic on the relationship between the use of non-interest income and bank profit and risk and reported, “that noninterest revenue sources are positively related to performance but inversely related to risk” (p. 1). The study recommends banks “follow a portfolio diversification approach during the pandemic that can be beneficial for banks expanding beyond the traditional lending sources of income” (Mateev et al. , p. 4). However, on the negative side, when banks move to non-interest income sources, they face higher credit risk (NPL) and more income instability (Ammar and Boughrara ). We add value to this strand of the empirical literature by testing the impact of efficiency on bank credit and insolvency risks in the MENA countries. 2.3 Hypotheses developmentCompetition in the banking industry as well as operating efficiency, are frequently mentioned as the main determining factors of bank performance. However, the existing studies do not precisely indicate how the structure of the banking market and/or the efficiency level impact the performance of banking institutions. For example, Cristian et al. () investigate the effects of competition, efficiency, and risk on bank's performance in Indonesia, and find that “competition of non-interest income market influence negatively on bank performance. Cost efficiency and revenue efficiency do not affect bank performance. Profit efficiency positively affects net interest margin, but not return on assets” (p. 1). We also expect efficiency and market competition to influence bank's performance in the MENA region strongly. Using more precise measures of efficiency, we test the hypothesis that the efficiency (and competition) effect is positive and significantly different between IBs and CBs. Hypothesis 1The level of efficiency and market competition is positively related to bank performance. This effect should be more pronounced for Islamic banks. According to Mirzae et al. (), “efficient banks have the ability to raise cheap capital, improve profitability, and therefore, create value for bank shareholders in the form of higher stock prices” (p. 13). Moreover, this is “especially important at times of economic disorder, like that brought by the coronavirus outbreak, when bankers and investors face heightened uncertainty and increased levels of risk” (p. 23). The more efficient banks seem to be more resistant to external shocks such as the COVID-19 pandemic. Therefore, we may expect the positive effect of efficiency on bank performance to be more enhanced during the COVID-19 outbreak. To examine this notion, we test the following hypothesis. Hypothesis 2The positive relation between bank efficiency and profitability is more pronounced during the COVID-19 outbreak. Rashid and Jabeen () examine banking sector performance in Pakistan and “using CAMELS denoting capital adequacy, asset quality, management, earnings, liquidity, and sensitivity to risk, has constructed the financial performance index (FPI)” (p. 19). The study finds that Islamic banks’ performance remains good compared to conventional banks because IBs increased their deposits as a main source of earnings, improved management efficiency, and introduced new products from 2013 onward. Furthermore, the efficiency effect on bank profitability could be different between CBs and IBs due to the different business models and risk management practices applied by Islamic and conventional banking institutions. For example, Hasan and Dridi () demonstrate that Islamic banks had relatively higher efficiency and profitability than their conventional peers during financial crises. However, the effect of efficiency on IBs performance during the COVID-19 outbreak remains unknown. To shed some light on this issue, we test the following hypothesis. Hypothesis 3The efficiency effect is expected to be different between Islamic and conventional banks. Pappas et al. () report that “although most studies tend to look at efficiency and stability in isolation, there are a few that examine the link between these concepts, e.g., Koutsomanoli-Filippaki and Mamatzakis () and Koetter and Porath () for applications using conventional banks, and Saeed and Izzeldin () for a comparative study between Islamic and conventional banks” (p. 2). A more recent paper by Saeed and Izzeldin () is the first study that “consider the efficiency-default risk paradigm in a comparative setup which includes IBs. The analysis shows that the relationship between profit efficiency and default risk of banks across the sample, for CBs and the GCC, is such that a decrease in default risk is associated with lower efficiency levels” (p. 1). Therefore, we also examine the efficiency-default risk paradigm by testing the following hypothesis: Hypothesis 4There is a positive association between efficiency, financial stability, and bank performance. 3 Data and methodology3.1 Data and sampleThis study builds on Mateev and Nasr () dataset and includes bank-level data for the period of 2006–2020. We use unbalanced panel data for 225 banks in the MENA region obtained from Bureau van Dijk’s Bankscope database and the annual reports of the sample banks. Following Mateev et al. (), we use “other secondary data sources such as the World Bank's World Development Indicators (WDI), International Financial Statistics, and annual reports of the central banks to collect macroeconomic data” (p. 6). The analysed period is extended to 15 years and represents the years for which financial and accounting data are available for each bank in the MENA countries. Banks with missing information are excluded from the list. As a result, 225 banks from 18 MENA countries are included in our final sample. The number of banks in our data set varies between 4 and 25 per country. Summary statistics for banks per country and by type (that is, Islamic or conventional) are provided in Table . The total number of bank-years of observations is 3375. Out of 225 banks, 162 institutions (or 72.0%) are conventional, and the rest are Islamic banks. In line with previous research, the bank-level explanatory variables “are winsorized at the 1 and 99% levels to mitigate the effect of outliers” (Bitar et al. , p. 5). Table 1 Composition of banks by country Full size table 3.2 Empirical specificationTo assess whether efficiency and market competition have a significant impact on bank performance and where this effect is more pronounced during the COVID-19 outbreak, we estimated the following model: $$Prof_{it} = \vartheta_{0} + \vartheta_{1} Efficiency_{it} + \vartheta_{2} Competition_{it} + \vartheta_{3} D_{t} + \vartheta X_{it - 1} + \varepsilon_{it}$$ (1) $$Prof_{it} = \vartheta_{0} + \vartheta_{1} Efficiency_{it} \times Islamic_{t} + \vartheta_{2} Effiiency_{it} \times {COVID{{-}}19}_{t} + \vartheta_{3} Competition_{it } + \vartheta_{4} D_{t} + \varphi X_{it - 1} + \varepsilon_{it}$$ (2) The dependent variable, Profit is a measure of financial performance (profitability) of bank i in year t; Efficit variable represents DEA efficiency scores, and Competit is a measure of banking market competition (the Lerner index); Xit is a vector of bank-specific and country-level variables that are known to explain banks’ performance; ϑ1 to ϑ4 are the regression coefficients, and εit is the error term in the model. Following Mateev et al. () approach, we introduce “a vector of dummy variables (Dt) that includes an ISLAMIC dummy variable that equals 1 if a bank is an Islamic banking institution and 0 otherwise, and a CRISIS time dummy (COVID-19) that takes the value of 1 for the year 2020, and 0 otherwise.We also control for the country fixed effects in each model” (p. 11). Next, we examine the effect of efficiency on bank performance when interacting with the crisis dummy variable. Accordingly, Eq. () incorporates an interaction term of the efficiency measure with the COVID-19 dummy together with all the explanatory and control variables of Eq. (). We also add an interaction term of the efficiency measure with the ISLAMIC dummy to investigate if the positive impact of efficiency is more pronounced in the sample of IBs. Likewise, we test the differential impact of market competition on IBs. Using the t-statistics of ϑ1 and ϑ2 coefficients in Eq. (), we can test the significance of efficiency and market competition effects on bank profitability. According to Hypothesis 1, both coefficients should be positive and significant. Next, we want to check if the efficiency effect on bank performance is more pronounced during the COVID-19 outbreak, so we add an interaction term of the efficiency measure with the COVID-19 dummy variable. According to Hypothesis 2, this coefficient should be positive and significant. On a similar note, and in line with Hypothesis 3, there should be a significant differential effect on IBs, so we expect the ϑ1 coefficient in Eq. () to be positive. Different alternative methods could be used when dealing with panel data modelling. Sun et al. ) report that previous studies “applied less powerful models, which could be shown to have serious statistical or econometric limitations, e.g., Fixed Effect Model (FEM) Ordinary Least Square (OLS). The problem with this and other similar approaches (e.g., Pooled OLS) is that the parameter estimations are measured with errors” (p. 199). Following Mateev et al. (), we use “fixed effect/random effect specifications and perform a Hausman test where the null hypothesis is that the preferred model is random effects vs. the alternative fixed effects” (p. 11). Additionally, we use the robust standard error clustered at the bank level. For robustness purposes, we run Eqs. () and () with “two-step system GMM estimator” (Generalized Method of Moments or GMM estimator). This method helps to “overcome the endogeneity problem that there is no clear demarcation between what independent and dependent variables are” (Sun et al. , p. 198). The estimated values of the correlation matrix coefficients (not presented here to conserve space) for explanatory and control variables are low, indicating the lack of significant collinearity between efficiency, market competition indicators, and other explanatory (and control) variables. 3.3 Variable selection3.3.1 Dependent variableFollowing Mateev and Bachvarov (), we use accounting ratios and, more specifically, “earnings to total assets (EARTA) and earnings to gross loans (EARGL) to measure the level of bank profitability” (p. 10). As alternative measures, we employ two market-based indicators of bank performance: Tobin’s Q, measured as the natural logarithm of the book value of assets less the book value of equity plus the market value of equity divided by book value of assets, and market to book value of equity (Abdallah and Ismail ). Following Chortareas et al. (), we also use the net interest margin ratio (NIM) as an alternative profitability measure. Appendix displays all the dependent and independent variables used in this study. 3.3.2 Independent and control variablesGuided by Louati et al. () and Mateev et al. (), we use the Lerner index to measure market competition. According to Hamza and Kachtouli (), the Lerner index “is a direct measure of competition through the distance between the price and the marginal cost, therefore capturing the extent to which a firm can raise its prices beyond its marginal cost” (p. 37). Moreover, “the value of the Lerner index varies between 0 and 1, such as a high Lerner value (close to 1) indicates a monopoly situation. However, when the Lerner index tends towards 0, the competition level is said to be very high. A Lerner index < 0 implies a price below the marginal cost that could occur due, for example, to a nonoptimal banking practice” (Louati et al. , p. 196). We follow Mateev et al. () approach to estimate the Lerner index over the period of 2006–2020. In the empirical literature, two methods are usually employed to calculate efficiency: Data Envelopment Analysis (DEA) and Stochastic Frontier Analysis (SFA). For the estimation of bank efficiency in this paper, we use the DEA approach, which was first proposed by Charnes et al. (). Banker et al. () extended the original method to include variable return-to-scale (VRS). According to Haque and Brown (), “this method solves linear programming problems to construct production possibilities of a decision-making unit (DMU) such as a bank, based on the envelopment of multiple input and output variables” (p. 277). Furthermore, the study indicates that previous research (e.g., Barth et al. ) “observed that, unlike the parametric technique (such as stochastic frontier analysis), a nonparametric DEA-based efficiency measure is not dependent on the misspecification of the functional form. Banker and Natarajan () also find that “DEA-based estimation is superior to parametric techniques in estimating the efficiency of a decision-making unit (DMU)” (p. 277). Guided by previous research (Safiullah and Shamsuddin ; Mirzae et al. ), we consider “three outputs: gross customer loans; other earning assets, including investment securities, loans and advances to banks, and other investment; and non-interest income, including net fees and commissions, net gains/losses on trading and derivatives, and other operating income. Non-interest income can be viewed as a proxy for off-balance-sheet activities, which is an important component of banking business” (Mirzae et al. , p. 23). The input variables used in this study are similar to those of many other studies (for more details, see Appendix ). Correspondingly, in Appendix , we present the summary statistics of DEA’s input and output variables of each group of banks (CBs and IBs). We follow previous research (see, e.g., Moudud-Ul-Huq et al. ; Mateev and Bachvarov ; Mirzae et al. ) and use several bank-level and country-specific characteristics that may impact bank profitability and risk. These include “bank net fees and commissions, deposits, loans to assets ratio, capital adequacy ratio, bank size, liquid assets, and equity to total assets ratio, employed in the regression analysis as control variables” (Mirzae et al. , p. 16). We assume that “bank fee income to total income ratio is a proxy for management quality that captures the effect of bank income diversification” (Mirzae et al. , p. 17). We also introduce equity to total assets ratio “as well-capitalized banks face a lower risk of bankruptcy and have a greater capacity to withstand financial shocks (p. 16). Next, we use the “bank liquid assets to total assets ratio as an indicator that an increase in the share of liquid assets lowers the amount of risky assets and strengthens bank resilience to external shocks” (p. 17). Finally, the capital adequacy ratio (CAR) measures a bank's financial strength (Bitar et al. ; Ghanem ). In line with previous studies, we use the “non-performing loans to gross loans (NPL/GL) ratio as a measure of credit risk, and distance-to-default (Z-Score) as a proxy for the bank’s financial stability” (Mateev et al. , p. 11). As mentioned before, we include several country-specific characteristics as control variables. First, we include an index of the institutional quality of a host country (institution), which is the mean of the six variables (voice accountability, political stability, government effectiveness, regulatory quality, the rule of law, and control of corruption) for each country in the sample (Kaufmann et al. ). The overall meaning is that “a higher value of the index indicates a better institutional environment in the sample country” (Mateev et al. , p. 13). Second, following Mateev et al. (), we use “GDP growth rate, inflation, and the growth in domestic credit to the private sector as a percentage of GDP to control for macroeconomic differences across the countries in our sample” (p. 15). 4 Empirical results and discussions4.1 Descriptive statistics and univariate analysisWe empirically investigate whether IBs and CBs performed differently during the observation period (2006–2020) and if the efficiency level can explain that difference. For this purpose, we provide descriptive statistics for bank performance and efficiency and compare the mean values of different variables used in this study across the two banking systems (see Table ). Table 2 Descriptive statistics of banks Full size table In Table we show the means, minimum and maximum values, and standard deviation for each variable by bank type, as well as the t-test results for mean differences. When the efficiency scores obtained from the DEA analysis are compared between Islamic and conventional banks in our sample, we do not find significant evidence that IBs have higher efficiency levels than CBs. Moreover, we observe a higher value of interest expense-to-gross income ratio for CBs than IBs (1.63 vs. 1.18); however, the mean difference is not significant. We measure bank profitability using EARTA and EARGL, and both show the dominant performance of IBs over their conventional peers (see Fig. ). When measured by Net Interest Margin (NIM), the analysis for bank profitability indicates that IBs performed much worse than CBs; however, the mean difference is statistically insignificant. These results align with Sun et al. () and their findings that IBs are not significantly different in behavior or dynamics from CBs and that CBs appear to be the benchmark for IBs. Fig. 1  Bank profitability development (CBs. vs. IBs) Full size image Table allows us to draw some interesting results when comparing the bank-level characteristics of the two groups of banks (CBs and IBs). For example, we observe that the capital adequacy ratio’s (CAR) mean value for the two groups of banks is between 19 and 28%, which is well above the minimum capitalization required by the Basel agreements. However, IBs seem to be better capitalized than CBs (28.5% vs. 19.3%). The data indicated that IBs have a lower deposit-to-assets ratio but outperform CBs in terms of loans to total assets (50.4% vs. 48.5%). Moreover, CBs have a higher amount of liquidity assets, while the level of equity capital is much higher in the IB group. The fee-to-total income ratio (FEE/INC) shows better management quality in the group of CBs (1.64 vs. 0.68). We also present some estimates for the Lerner index and Boone indicators for the sample countries included in the analysis. In line with Mateev et al. (, , ), we do find “a significant difference in the level of market power as measured by the Lerner index and Boone indicator between the two samples. The overall Herfindhal-Hirschman index (HHI) is 0.20, which is considered “moderately concentrated” for all countries in the sample, with the mean difference strongly significant at the 1% level of significance” (p. 13). Our results in Table support previous research findings on banks in the MENA region (Haque, ; Mateev and Bachvarov ; Mateev et al. , , ). Table confirms the Mateev et al. () finding; “when we compare the insolvency risk (measured by distance to default, or Z-score) between the two groups of banks, we find that CBs are much less risky than IBs (2.70 vs. 2.13). Moreover, we observe that IBs have higher credit risk measured by non-performing loans to gross loans (NPL/GL) ratio” (p. 10). Finally, in support of Mateev et al. (, , ), we find that “the percentage of ownership concentration is larger for CBs than IBs (49.2% vs. 44.1%)”. Overall, the high percentage of concentration in the ownership structure of both IBs and CBs aligns with the findings of other similar studies for the MENA region (Farazi et al. ; Haque and Brown ). We also confirm the finding that the countries in the MENA region are “characterized by weak institutional environment” (Mateev et al. ). 4.2 The effect of efficiency and market power on bank performanceWe estimate a panel regression with the model specification (see Eq. ) to test Hypothesis 1. Our dependence variable is bank profitability measured by EARTA and EARGL. In line with the efficient structure hypothesis, we expect efficiency level to significantly influence banks' performance in the MENA region. On a similar note, there should be a positive association between banking competition (e.g., market power) and bank profitability. We also hypothesize that there is a strong differential impact of efficiency and competition on IBs profitability. Therefore, we run the regressions first for the entire sample and then separately for each group of banks (CBs and IBs). We report the outputs of the analysis in Table . We use the Lerner index in each regression as a measure of market power. The efficiency scores are calculated using the Data Envelopment Analysis (DEA) technique proposed by Charnes et al. (). The outcomes of the analysis are discussed below. Table 3 Panel regressions of bank profitability (All banks, 2006–2020) Full size table First, we consider the results without estimating the differential impact of efficiency on IBs (see Models 1 and 5). In line with the efficient structure hypothesis in which efficiency and profitability are positively associated (Hannan ), we find that efficiency positively influences bank profitability; however, this effect is only marginally statistically significant. The relationship between competition in the banking industry (measured by the Lerner index) and bank profitability is also positive. This supports our first hypothesis (H1) that efficiency and competition are associated with enhanced bank performance. Since the Lerner index is inversely related to market competitiveness, our results indicate that less banking market competition will be associated with higher profitability. In other words, banks with higher market power will raise their interest rates level to enhance their profits; however, in turn, high rates of interest can lead to moral hazard problems by increasing banks’ non-performing loan ratio (that is, their credit risk). We further investigate this issue by incorporating an interaction effect of efficiency measure and market power indicator. The results from an unreported test indicate that the positive influence of efficiency diminishes with the increased market power of banks. This result aligns with Cristian et al. (), who investigate the effects of competition, efficiency, and risk on bank performance in Indonesia and found that competition in non-interest income market negatively impacts bank performance. Second, we test the validity of our hypothesis that the efficiency effect on bank performance is more pronounced for IBs than CBs (H1). We do not find strong evidence in support of this hypothesis. The estimated coefficients of the interaction term between DEA efficiency scores and the Islamic bank dummy are consistently negative and statistically significant across different models. The negative sign of the interaction term indicates that the (positive) effect of efficiency on bank profitability is more pronounced for CBs. This result aligns with our univariate analysis (see Table ), which shows a slightly higher efficiency level in the group of CBs. The regression models using the EARGL variable as an alternative proxy of bank profitability show a positive and significant coefficient for the Islamic bank dummy. Therefore, we may conclude that the better performance of IBs appears to be linked more to their specific business model (based on Shariah principles) rather than the level of efficiency, which aligns with Hasan and Dridi (). Furthermore, we find that banks with higher market power are more profitable, and this effect is less pronounced for Islamic institutions (the estimated coefficient of the interaction term with the IB dummy variable is negative). These findings do not support our hypothesis that the effect of efficiency (and market power) is expected to be stronger for IBs. This can be explained by the fact that Islamic jurisprudence forces Islamic banks to operate according to a profit and loss (risk) sharing principle (Abedifar et al. ). We further test this hypothesis by running our models separately for the samples of IBs and CBs (see Sect. ). Our main results continue to hold after controlling for a number of common bank-level determinants of profitability. We find that deposits, loan-to-assets ratio, liquid assets, and default risk are strongly associated with bank profitability (see Table ). The estimated coefficients of of variables carry signs and magnitude as anticipated in the empirical literature. For instance, the coefficients on bank stability (Z-Score) and bank liquidity ratio (LA/TA) is generally positive and largely significant. The effects of bank equity position (EQ/TA) and capital adequacy ratio (CAR) are mixed and statistically insignificant in all models. Further, the size of banking institutions is expected to positively impact bank performance, especially during the COVID-19 pandemic. For example, the strengthening impact of bank’s size on the increase in credit risk costs caused by the COVID-19 pandemic is confirmed by Borri and di Gorgio (), conducted on a sample of listed European banks. Our evidence does not support these previous findings. Models that run on EARTA as a profitability measure indicate that the ‘institution’ variable is significant and negative; that is, banks located in countries with better institutional environment experience lower profitability. Macroeconomic conditions do have a strong impact on the profitability level of banks in the MENA region. Specifically, we find that banks in the MENA countries with a high GDP and low inflation experience lower profitability. This finding is commensurate with Mirzae et al. (), who report similar results for the pre-crisis (COVID-19) period. Previous research finds a significant impact of COVID-19 on bank stock performance during the COVID-19 outbreak (Demirgüc-Kunt et al. ; Hassan et al. ). A very limited number of papers, however, have a specific focus on the relationship between efficiency and bank performance in the MENA region before and during the pandemic. For example, Mirzae et al. () argue that “the observed higher stock returns for IBs during the initial phase of the COVID-19 crisis can be explained by their pre-crisis levels of efficiency” (p. 15). We further investigate this issue by including a COVID-19 dummy variable in our baseline model (see Eq. ) to account for a possible crisis effect on bank's profitability. The negative sign of the estimated coefficient indicates that banks in the MENA countries experienced low profitability during the COVID-19 outbreak. To test our second hypothesis that the positive association between efficiency and bank performance is more pronounced during the COVID-19 crisis, we employ an interaction term between the efficiency measure and the COVID-19 dummy variable (see Models 4 and 8). We do not find evidence of a significant relationship between efficiency and bank profitability during the COVID-19 outbreak. The reason could be that we used only one year of data (2020) for the COVID-19 pandemic, and its effect is yet to unfold. To shed some light on this, we follow Mirzae et al. () approach and test the relationship between banks’ stock performance during the initial stage of the COVID-19 crisis (from January 1st, 2020, to March 31st, 2020) and the pre-crisis (2019) efficiency level. The result does not indicate a significant efficiency effect on bank stock performance during this crisis period. Therefore, we have to reject our second hypothesis (H2). 4.3 The differential effect of efficiency on IBs performanceIn this section, we investigate the issue of whether there is a significant differential impact of efficiency (and market competition) on IBs performance and if this effect is more pronounced during the COVID-19 outbreak. A limited number of studies estimate the IB's efficiency as compared with CBs in emerging economies (Mirzae et al. ; Saâdaoui and Khalfi ). However, the empirical evidence remains mixed and inconclusive. To shed some light on this puzzling issue, we split the sample into two sub-samples and ran our analysis separately for each group of banks (CBs and IBs). The analysis outputs are presented in Tables and , respectively. We find a significant difference in the efficiency effect between the two banking systems, which supports our third hypothesis. Specifically, the efficiency measure (DEA scores) is strongly significant and positive only in the sample of CBs, whereas this effect is irrelevant for IBs. Therefore, improved efficiency does not lead to better profitability for Islamic banking institutions. This can be explained with the fact that Islamic finance is based on the rules and principles prescribed by Shariah. Therefore, IBs profitability appears to be linked more to their specific business model rather than the efficiency level. The results in Table for EARTA indicate that market power (as measured by the Lerner index) strongly impacts CBs profitability. When the regression analysis employs the alternative measure of bank profitability (EARGL), the results in Table suggest that IBs with strong market power enjoy higher profitability. Table 4 Panel regressions of bank profitability (CBs, 2006–2020) Full size table Table 5 Panel regressions of bank profitability (IBs, 2006–2020) Full size table Next, we estimate the efficiency effect during the COVID-19 outbreak. The estimated coefficient of the interaction term between the efficiency measure (DEA scores) and the COVID-19 dummy variable is insignificant in both samples. Therefore, we are unable to provide evidence that the effect of efficiency quality is more pronounced during the COVID-19 crisis. This can be explained by the fact that our sample includes data only for the initial stage of the COVID-19 pandemic (one year), and the expected effect on bank profitability is yet to reveal. A similar (insignificant) crisis effect is found for market competition; banks with higher market power are more profitable independent of their business model. Isnurhadi et al. () find that capital adequacy and efficiency strongly influence IB's risk-taking behavior. Specifically, they report that “efficient banks with higher capital levels tend to allocate fewer funds to monitor loans which result in higher levels of credit risk, while inefficient banks with lower capital levels tend to increase the levels of credit risk to maximize their revenues” (p. 842). In the same line of thought, we test the hypothesis that efficient banks with higher capital levels will be more profitable (see Models 4 and 8). The results show that efficient banks with higher capital levels experience lower profitability. This effect is significant only for IBs. A recent study by Mirzae et al. () reports that bank size is inversely related to bank stock performance during the crisis. To test this hypothesis, we interact the size variable with the COVID-19 dummy indicator; the estimated coefficient of the interaction term appears insignificant. Regrading the policy implications, this study suggests some guidelines for decision-makers and regulators responsible for the banking sector’s financial stability in the MENA region. Specifically, our findings related to the IB's efficiency provide some tips for governments and policymakers on how to shape their policy initiatives during the pandemic. We recommend that Islamic banks can be entrusted with a more decisive and important role in post-COVID-19 revival. Since improved efficiency does not explicitly mean better profitability for IBs, better governance-related initiatives based on Shariah principles will appear to be the factors enabling Islamic banks to withstand the most severe effects of global financial crises so far. 4.4 The interaction effect of efficiency and risk on bank profitabilityTo test the hypothesis that the effect of efficiency on bank profitability may depend on the level of bank risk, we introduce an interaction term between the DEA efficiency scores and different measures of bank risk. Specifically, we use the non-performing loans to gross loans (NPL/GL) ratio as a measure of credit risk and distance-to-default (Z-Score) as a proxy for the bank’s insolvency risk. The outcomes of the regression analysis are presented in Table . Table 6 Panel regressions of bank profitability (All Banks, 2006–2020) Full size table In Model 2, we employ an interaction term with Z-Score, and in Model 4—with use the NPL/GL measure. As expected, bank profitability decreases with a higher level of insolvency risk. The interaction term is statistically significant and negative; we may conclude that efficiency's positive effect on bank profitability diminishes with the increase of insolvency risk (financial stability). In other words, efficient banks will have a higher level of performance if the banks manage to keep a low level of default risk. This finding corresponds to our prediction in Hypothesis 4. Therefore, managers need to consider the interplay between efficiency and bank risk when deciding on the level of efficiency. Likewise, we interact the credit risk measure with efficiency and find that credit risk level does not moderate the efficiency impact on bank profitability (see Models 4 and 8). The COVID-19 dummy variable shows a significant yet marginal effect; the negative sign of the estimated coefficient speaks for the detrimental impact of the COVID-19 pandemic on bank performance. The regressions on the EARGL measure of bank profitability indicate positive and statistically significant effects on efficiency scores. However, this effect is not moderated by the level of credit risk or a bank’s financial stability. Finally, our results continue to hold after controlling for a number of a common bank- and country-level determinants of bank profitability. Empirical literature investigates the relationship between efficiency and bank risk behavior. Previous research (Mirzae et al. ) shows that “managing a high level of efficiency could be to the detriment of shareholder value if it comes at the expense of excessive risk-taking” (p. 23). Recently, Isnurhadi et al. () examined 129 Islamic banks in 29 countries and found a positive association between efficiency and bank stability (measured by Z-Score) but a negative with credit risk (measured by loan loss provision to total liabilities). However, the study uses a cost-to-income ratio, an imprecise measure of (in)efficiency. Therefore, it is important to consider the relationship between efficiency and bank risk level in order to define whether bank risk-taking behavior depends on the efficiency management policy. In line with Hypothesis 4, we expect a positive association between efficiency and banks’ financial stability to existing. The regression outcome of the analysis is presented in Table . Table 7 Panel regressions of bank risk (All banks, 2006–2020) Full size table We find evidence for a significant positive effect of efficiency on bank default risk measured by the Z-score. Since “higher values of our measure of insolvency risk imply a lower probability of default” (Mateev et al. , p.11), the positive coefficient indicates that managing a higher efficiency level leads to improved financial stability. This result supports our last hypothesis (H4). However, banking competition seems to have no role in explaining the bank risk level. We also find that bank risk level is more enhanced for CBs; the lack of a strong differential effect on IBs risk behavior is confirmed by the fact that the interaction term of efficiency with the IBs dummy variable is insignificant. Next, we analyze the effect of efficiency (and market competition) on bank credit risk. Our results indicate efficient banks are prone to keep higher levels of non-performing loans. Furthermore, market competition (measured by the Lerner index) is also significantly and positively related to bank credit risk. Moreover, the positive sign of the interaction term of market power with the IB dummy variable tells us that IBs with higher marker power keep a higher level of non-performing loans. This result confirms Mateev et al. () finding that the level of banking competition is less effective in shaping the risk-taking behavior of CBs than IBs. The interaction effect of bank capital and efficiency shows that efficiency encourages banks to reduce credit risk when bank capital is relatively high (see Model 8). Finally, we find that the shock caused by the COVID-19 pandemic was mostly manifested in the form of the increased credit risk of banks in the MENA region. 4.5 The efficiency effect at different levels of banking competitionWhether the efficiency effect on bank performance may depend on the level of banking competition is still unexplored. Following Mateev et al. (), we split the sample into three sub-samples depending on the market competitiveness level (high, medium, and low) and run separate regressions for each. The outcomes of the analysis are presented in Table . Table 8 Panel regressions of bank profitability (different levels of market competitiveness) Full size table The results reported in Models 1 and 2 show that efficiency plays a significant role in explaining profitability level for banks operating in markets with a high and medium level of competition. However, this effect is not significantly different between the two banking systems (Islamic and conventional). For the low level of market competitiveness (see Model 3), the analysis indicates that the efficiency effect is positive yet insignificant in both samples. In the next three models, we present our results for the regression models that employ the earnings to gross loan (EAR/GL) ratio. These are quite similar to the results using EAR/TA measure, except for markets with a medium competition level where the efficiency effect is more pronounced for CBs group. This result indicates that, in less competitive markets, which are representative of emerging economies in the MENA region, CBs may enhance their profitability by managing high efficiency. This finding aligns with Meslier et al. (), who report that conventional banking institutions “fight” the competitive pressure of Islamic banks by setting higher deposit rates when their market power is lower. The behavior of CBs, in this case, can endanger their financial stability. One possible solution is to adopt appropriate strategies for efficiency management. Our results confirm the Isnurhadi et al. () findings that “competitive pressure is not always good for the banking industry. Increased competition can lead to high risk-taking due to reduced market power and the value of bank franchises” (p. 842). Moreover, the rise in the banking market concentration will most likely affect the efficiency of banks’ operations and their ability to finance the economy (Kozak and Wierzbowska ). This finding is expected to have strong implications for the regulatory authority and decision-makers in the MENA region to set up policies that promote bank efficiency in addition to establishing several regulations related to capital and bank supervision. This aligns with Mateev et al. () findings that “decision-makers should carefully tailor banking reform initiatives related to private monitoring and banking supervision depending on the type of the banking system” (p. 39). Moreover, the need for banks to improve efficiency will encourage them to act prudently and increase their level of stability during times of financial crises like the COVID-19 pandemic. 5 Robustness checks and alternative specificationsWe perform several robustness tests. First, in addition to the fixed and random effect models reported in Table , the analysis employs identical specifications using the Generalized Method of Moments (GMM) estimator. The difference and system GMM estimators developed by Arellano and Bond (), Arellano and Bover (), and Blundell and Bond () were designed for situations with “small T, large N” panels such as ours. They dealt well with independent variables that were not strictly exogenous. They correlated with past and current realizations of the error, with fixed effects, heteroscedasticity, and autocorrelation within individuals (Roodman ). For difference GMM, all regressors are usually transformed by differencing (also referred to as Arellano–Bond estimation). System GMM is an extension of difference GMM (also referred to as the Arellano–Bover/Blundell–Bond estimator), which augments the Arellano–Bond estimator by building a system of two equations in levels and first differences, making an additional assumption that the first differences of instrument variables were uncorrelated with the fixed effects (Otero et al. ). System GMM was developed to tackle the weak instrument problem and allowed for the introduction of more instruments and the improvement of the models’ efficiency. Hence, we use the two-step system GMM estimators with Windmeijer () corrected standard errors. The results for GMM tests of efficiency and competition effects are reported in Table . In support of our previously reported results, we observe a positive influence of efficiency (and market competition) on bank performance; these effects are more pronounced in the sample of CBs. Table 9 Panel regressions (GMM estimator) of bank profitability (All Banks, 2006–2020) Full size table Second, we check the robustness of our results using alternative measures of bank profitability. For example, instead of accounting measures, we employ two market-based indicators—Tobin Q as a proxy for firm value and market-to-book value of equity (Abdallah and Ismail ). The results indicate a positive impact of efficiency (and market competition) on bank profitability. Again, we do not find evidence that the efficiency effect is more pronounced during the COVID-19 outbreak. Next, we report the results for three different efficiency scores (CRS, VRS-OUT, and SCALE) using variations in the DEA input and output models (see Appendix ). Based on the results in Table , we can confirm the findings from our previous analysis, with signs and magnitude of the regression coefficients comparable to those reported in Table . Specifically, the estimated efficiency (and competition) coefficients are significant and positive when EARTA is used as a profitability proxy. Regressions run on the EARGL variable indicate that the efficiency effect is more pronounced for IBs. At the same time, the level of market competitiveness strongly moderated the efficiency effect in the sample of CBs. Finally, we follow Mateev et al. () and decompose the Lerner index in three dummy variables reflecting three levels of competitiveness (high, low, and medium). In an unreported test, we run our baseline model (see Eq. ) using an interaction term between DEA efficiency measures and the three dummies. The results indicate that the level of market competition strongly moderates the association between efficiency and bank profitability if banks operate in highly competitive industry sectors. Third, in addition to the most widely used measures of bank risk—the default risk (Z-Score) and the credit risk (NPL), for our robustness purposes, we employ two alternative bank risk proxies—portfolio risk (σROA) and operation risk (σNIM), respectively. Following Danisman and Demirel (), we used the standard deviation of net interest margin (NIM) as a proxy of operational risk, and the portfolio risk is our second proxy (computed by ROA divided by the standard deviation of ROA). The results (available on request) are similar to those reported in Table . Specifically, the efficiency score coefficient is significantly positive with σ(ROA) while insignificant with σ(NIM). 6 ConclusionsThis paper investigates the effect of efficiency and market competition on bank performance in the MENA region. While the previous researchers mainly focus on a single country (see, e.g., Cristian et al. ) or Islamic banking efficiency (see, e.g., Saâdaoui and Khalfi ), we are the first to analyze the impact of bank efficiency and market competition on banking sectors performance in the MENA region incorporating the effect of COVID-19 outbreak. The ongoing pandemic crisis (COVID-19) offers a better quasi-natural experiment than the global financial crisis (GFC) and prior crises to study bank performance in turbulent times. We take this opportunity to investigate the effect of banks’ efficiency (and competition) on their performance and stability during the pandemic outbreak. We document that this effect is significantly different between Islamic and conventional banking systems in the MENA region due to the fact that IBs experienced higher efficiency/profitability levels than CBs (Saeed and Izzeldin ; Batir et al. ; Saâdaoui and Khalfi ). Thus, we test the hypothesis that IBs are more profitable and stable in crises such as the COVID-19 pandemic due to their better efficiency than CBs. Our results remain robust when controlling for different bank-level and country-specific characteristics and allow us to conclude that improved efficiency does not necessarily lead to better profitability of Islamic banks. We build our analysis on the previous research findings, which report a positive association between efficiency and bank profitability. We argue that the observed positive effect can be explained within the framework of the efficient structure hypothesis (Hannan ). Likewise, market competition (measured by the Lerner index) in the banking industry is positively and significantly related to bank profitability. Since the Lerner index is inversely related to the level of market competitiveness (Mateev et al. ), our results indicate that a lower level of banking market competition is associated with increased profitability. In other words, “banks with higher market power will raise their interest rates level; in turn, high rates of interest can lead to moral hazard problems by increasing non-performing loan ratio” (Mateev et al. , , , p. 14). In addition, we address whether stronger market competition impacts the positive association between efficiency and bank profitability. Our results indicate that the efficiency effect diminishes with the increased market power of banks. This aligns with Cristian et al. () finding that profit efficiency and market competition (measured by the Boone indicator) are associated with improved bank performance. We next interact the efficiency variable with the Islamic bank dummy to estimate its differential effect on IBs profitability. The interaction coefficients appear to be consistently negative and significant across different models. The negative sign indicates that the (positive) effect of increased efficiency is actually more pronounced in the sample of CBs than IBs. This finding informs us that the better performance of IBs appears to be linked more to their specific business model (based on Shari’ah principles) rather than the efficiency level, which aligns with Hasan and Dridi (). Likewise, the results indicate that IBs with high market power become more profitable compared to CB when the level of competition in the banking market increases. Overal, our findings do not necessarily support the notion that efficient IBs are more profitable. This can be explained by the fact that Islamic jurisprudence forces Islamic banks to operate according to a profit and loss (risk) sharing principle (Abedifar et al. ). Our research justifies the concept that efficiency and banking market competition are strongly linked with banks' risk behavior independent of their business model. Our results indicate that managing a higher efficiency level leads to improved financial stability while increasing the bank's appetite for taking higher (credit) risk. This result complements Isnurhadi et al. () findings that “efficiency positively affects bank stability (measured by Z-Score) and negatively affects credit risk (measured by loan loss provision to total liabilities)” (p. 1). Finally, we contribute to the empirical literature on efficiency by offering new evidence to the puzzling issue of efficiency and bank performance in the face of enhanced banking competition. Specifically, we investigate the efficiency effect at different levels of market competition (high, medium, and low). The outcomes of the regressions indicate that for banks operating in markets with a high and medium level of competition, striving for high efficiency decreases banks’ profitability. This effect is similar across the two types of banks and is more pronounced during the COVID-19 outbreak. For the low level of market competition, the analysis indicates the efficiency level does not impact bank performance, and this effect is insignificant for both IBs and CBs. Regarding policy implications, we provide some guidance to regulators, policymakers, and bank managers. First, the perception of the efficiency effect is important as it allows regulators and policymakers to convey policies that mark the entire banking system in the MENA countries. For example, the entry of foreign banks and consolidation of smaller banks into larger ones, as well as the tightening of banking regulations, are some of the initiatives of policymakers in these countries that may help to improve banks’ efficiency. Since the role of the Islamic banking system in recovery from the COVID-19 pandemic has not been examined and analyzed so far, our findings related to their efficiency will provide some tips for the governments and policymakers in the MENA region on how to shape their policy initiatives during the pandemic. Further, we suggest that IBs can be entrusted with a more decisive and important role in post-COVID-19 revival. At the same time, the need for banks to improve efficiency will encourage them to act prudently and increase their level of stability during times of financial crises like the COVID-19 pandemic. Therefore, when making decisions, bank managers must reach an appropriate trade-off between the advantages and disadvantages of managing high-efficiency level and consider the moderating effect of banking competition. References
Xie H, Chang H-L, Hafeez M, Saliba C (2022) COVID-19 post-implications for sustainable banking sector performance: evidence from emerging Asian economies. Econ Res - Ekonomska Istraživanja 35(1):4801–4816 What are the banking priorities for 2023?The PRA's priorities include financial resilience, operational risk and resilience, risk management and governance, climate-related financial risks, diversity, equity and inclusion (DEI), regulatory reporting and data quality. What are the concerns for banks in 2023?The March 2023 banking crisis highlighted the vulnerability of the banking sector to a sudden rise in interest rates. Specifically, banks' ability to limit the pass-through of rate-hiking cycles into deposit rates allows them to benefit from higher rates, but only gradually. What are the two failed banks in 2023?Signature Bank Signature failed just two days after Silicon Valley Bank went down. It, too, had suffered a run on deposits. The failure was announced shortly before Asian markets opened on a Monday morning , as panic was spreading in the wake of SVB's collapse. What are emerging risks for banks?Cyber threats continue. Banks continue to leverage new technology to further digitalization efforts, offering innovative products and services to meet customer demands. Increasing digitalization efforts can also heighten risk of fraud and error, including fraud targeting peer-to-peer and other faster payment platforms. |
