Islamic banks and their conventional peers: an empirical investigation on 115 banks from MENA countries

Purpose Exploiting a sample of 80 conventional banks (CBs) and 35 Islamic banks (IBs), this study aims to distinguish the IBs’ performance from their conventional peers in 7 Middel East and North Africa (MENA) economies over the period 2005–2014 covering the 2008 GFC. Design/methodology/approach To avoid misleading results, this research used panel-corrected data from outliers effects by quantile method. Then, following the use of the two-sided Student’s t -test and the discriminant function analysis (DFA), we adopt nonlinear panel models (Random Logit and Pooled Probit) to further distinguish between banks. Then, we focus on the stability side through dynamic Generalized method of moment (GMM) linear models and interaction variables to capture the 2008 global financial crisis (GFC) impact on IB performance. Findings Univariate tests show that IBs are, on average, less profitable, more liquid and capitalized, less stable, have higher credit risk and are more solvent than CBs. In addition, the difference between the two types of banks was significant pre- and post-GFC; IBs are more profitable pre-GFC and more solvent post-GFC. In accordance with the univariate t -test results, the nonlinear pooled probit model (random logit) confirms that banks, which have more liquidity, are better capitalized, more solvent and less stable (less stable) are more likely to be IBs. From the DFA, stability was the first financial ratio important to discriminate between the two types of banks. In line with the DFA results, from the dynamic models, once the interaction variables are integrated, the GMM estimation result suggests that stronger macroeconomic stability and higher profitability, capital adequacy ratio (CAP) and liquidity are linked to increased IBs stability in the 7 MENA economies post-2008 GFC. Originality/value The present study contributes to the ongoing debate by conducting a formal empirical analysis, taking account of a range of considerations (outliers correction, interaction variables and 2008 GFC impacts) that to the best of our knowledge have not been considered by prior studies for the MENA zone.