There are many studies of bank performance and bank failure in the literature. Most of these studies used banking ratios as variables in their models without giving consideration to their appropriateness, nor was much consideration given to the stability of those ratios through time and across asset size. Many studies also failed to recognize that bank structure may differ by asset size. This study evaluates a large number of banking variables in order to identify stable ratios. These ratios are then used in disaggregated logistic models to predict bank failure. The study finds that the disaggregated models with stable variables were better predictors of bank failure than aggregated models used in earlier studies.

JEL Codes

L25, G21, G33


Bank, Small Bank, Failure