A Discriminant Analysis of Commercial Bank Profitability

Abstract

The research objective of this study is to isolate a few "key" financial ratios significantly explaining variation in commercial bank profitability. A methodological objective is to test the operational and statistical results of regressing only variables that strongly discriminate between groups of banks exhibiting large differences in profitability.

Discriminant analysis was utilized to find a subset of explanatory variables that explain differences in bank profitability for two subsamples of banks with large differences in profitability. Multiple regression was used to test the relevance of this variable set for all sample banks. The regression results demonstrated that the variables that accounted for differences in profitability between the two groups also performed well for the entire sample. By isolating the variables that critically affect profitability, attention is focused on those variables that may be operationally introduced into profitability planning and evaluation.

The evidence points to the overwhelming importance of operating expenses and non operating items is determining bank profitability. Operating expenses such as "interest on time and savings deposits" and "all other expenses" are particularly important. The evidence also indicates that banks could improve their profitability by investing a greater portion of their assets in tax-exempt securities and by charging loan interest rates commensurate with the actual credit risks. Aside from this, changes in asset composition and asset returns offer little in the way of improving relative profitability.