This research aims to create early warning models for predicting financial distress on Indonesian commercial banks. Early warning models is made by estimating bank indicators that led bank run into financial problems which to be shut down by the government. Estimation use two methods, discriminant analysis, logistic regression. The data used to create models is bank's financial ratios in 1994-1997 gathered from Direktori Perbankan Indonesia (DPI). Out of the sample test use the data in 1998. All model were used to predict the bank's financial distress after 2000. This study found some characteristics of distressed banks that will be in the state of failure in two or three years. Finally, this research found that early warning system models is able to predict the probability of financial distress on commercial banks.
- Bank failures, Banking crisis, Discriminant analysis, Financial distress, Logistic regression