The macroeconomic determinants of systemic banking crises. Panel data logit regression approach.

Abstract

This research aims to examine the macroeconomic causes of systemic banking crises across countries and to advance the existing prediction models for banking crises. This paper uses panel data logit regression with a sample consists of 47 developing countries, 32 developed countries, 28 least developed countries and 11 transitioning countries. The analysis uses annual data for banking crisis observations between 1970 and 2017 from Laeven and Valencia (2020), along with annual data for gross domestic product, domestic credit to private sectors by banks, foreign direct investment inflows and outflows and inflation, which have been collected from different data sources such as the World Bank Data, International Monetary Fund (IMF) and International Financial Statistics (IFS), depending upon the availability of the data. Our findings suggest that the growth of economies, represented by gross domestic product value, is related to the existence of banking crises across countries. There is no anticipation of the flows of money in the occurrence of banking crises across countries. Additionally, this study confirms that there is only a 0.1 percent probability of a banking crisis happening as a result of credit expansion. Furthermore, the significance of inflation in anticipating systemic banking crises in this study supports the work of other studies that link inflation with the existence of banking crises. The findings enable regulators to anticipate crises and mitigate their adverse implications. Although this study successfully demonstrates the significance of the selected variables in the occurrence of banking crises, the generalisability of the research findings is subject to certain limits. For example, the lack of a unified dataset required the collection of data from different datasets that used different methods. This may affect the outcomes of the study regarding the absence of a unified methodology for the data collection among datasets.