Title: Measuring Islamic banking efficiency using data envelopment and regression analysis

Authors: Foued Saâdaoui; Monjia Khalfi; Rim Ben Elouefi

Addresses: Institut des Hautes Etudes Commerciales (IHEC), University of Sousse, Sahloul 3, B.P. 40, Sousse 4054, Tunisia ' Center of Human and Social Sciences, Spanish National Research Council, C/Albasanz, 26-28, Madrid 28037, Spain ' ESPRIT School of Engineering, El Ghazala Technological Pole, Tunis 2083, Tunisia

Abstract: Banking efficiency is a key indicator for managers and decision makers to ensure economic stability and development. This article proposes a data-mining framework aimed at predicting Islamic banking efficiency (IBE) from a balanced sample of banks operating in developed and emerging economies before and during the subprime mortgage crisis. The variable assessing efficiency in this study is mainly measured using data envelopment analysis (DEA), while regression is performed to estimate the importance of the determinants of each type of efficiency. Experiments show that Islamic banks in developed countries are more efficient than their counterparts in emerging countries in both crisis and non-crisis periods. As Islamic banks operate under decreasing returns to scale, it is also proven that their main source of efficiency is efficiency of scale. On the other hand, for banks operating within increasing returns to scale, the source of efficiency is rather pure efficiency. The results therefore show that the most efficient banks in both regions are the best capitalised and the largest in terms of size. Accordingly, capitalisation and size can be considered as the main determinants of Islamic banking efficiency. In both regions, these findings may have several implications for risk management and corporate social responsibility.

Keywords: data envelopment analysis; DEA; regression analysis; Islamic banks efficiency; subprime crisis.

DOI: 10.1504/IJMDM.2024.138320

International Journal of Management and Decision Making, 2024 Vol.23 No.3, pp.311 - 336

Received: 23 Sep 2022
Accepted: 17 Nov 2022
Published online: 01 May 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article