Predicting bank failures by using banking performance and condition ratios

Abstract

I analyze the relationship between bank failures that occur in the financial sector and the banking performance and condition ratios. The performance ratio that I chose to use in my analysis is the return on equity and the condition ratios are the capital adequacy ratio, the leverage ratio and the loan-deposit ratio. To achieve this goal, I use Estrella's model (2000) as my main framework and observe the banking sector over the recent credit crisis. I apply logit regression analysis to see if making predictions about future bank failures is possible with the available past data. In the regression analysis, I use both three-factor models and single factor models to assess the efficiency of more previously collected data. In order to see the motivators behind the banking firms' decisions, I also apply correlation tests between each performance and condition ratio.