Random forests-based early warning system for bank failures

- This paper introduces a random forests-based early warning system (EWS).
- Our approach is significantly more accurate than other conventional EWSs.
- There are 730 banks in danger with assets equivalent to about 95.3 million US dollars in total.

This paper introduces a novel random forests-based early warning system for predicting bank failures. We apply this method to the analysis of bank-level financial statements, in order to find patterns that identify banks in danger of failing. The experimental results show that our method outperforms conventional methods in terms of prediction accuracy.