How Does Corruption Undermine Banking Stability? A Threshold Nonlinear Framework

Posted: 6 May 2019

See all articles by Mohamed Sami Ben Ali

Mohamed Sami Ben Ali

Qatar University - Department of Economics

Fredj Fhima

IHEC - University of Sousse, Tunisia

Ridha Nouira

University of Monastir

Date Written: April 6, 2019

Abstract

This study assess the effect of corruption on the occurrence of banking crisis for a sample of 38 countries over the period 2000 – 2017. We consider both the direct and the indirect channels through which corruption might affect the occurrence of banking crisis. We also check using a threshold regression approach for the existence of a corruption threshold driving the existence of a regime switching in our sample countries for both high-income and low-income countries. Estimation outcomes show robust support to suggest that overall, corruption increase the probability of banking crisis. The indirect effect estimation suggest that corruption negatively affects the banks’ lending channel through excessive risk rather than their profitability channel. The panel threshold analysis provides evidence of a nonlinear corruption-banking stability relationship with the existence of two corruption-banking stability regimes. The study also provides evidence that corruption matter more for low-income than for high-income countries for their banking system stability.

Keywords: Corruption, Banking Crisis, Panel Threshold Effects

JEL Classification: D73, C54, G28

Suggested Citation

Ben Ali, Mohamed Sami and Fhima, Fredj and Nouira, Ridha, How Does Corruption Undermine Banking Stability? A Threshold Nonlinear Framework (April 6, 2019). Available at SSRN: https://ssrn.com/abstract=3367336

Mohamed Sami Ben Ali (Contact Author)

Qatar University - Department of Economics ( email )

2713 Doha
Qatar

Fredj Fhima

IHEC - University of Sousse, Tunisia

Route Hzamia Sahloul 3 - BP n° 40 - 4054 Sousse
Sahloul 3, 4054
Tunisia

Ridha Nouira

University of Monastir ( email )

Mounastir, 4100
Tunisia

Do you have negative results from your research you’d like to share?

Paper statistics

Abstract Views
338
PlumX Metrics