The Early Prediction of Bank Defaults by Central Banks

Abstract

This article elucidates the approach used by central banks to monitor bank stability at an early stage. Existing default prediction models use CAMELS to estimate bank stability. This research, using Russian regional data, suggests that the internal bank data available to central banks are crucial for default prediction and are complementary to CAMELS. The combined use of central bank payment system data and Basel norms improves the quality of default prediction, benefiting from the absence of information asymmetry. The empirical results support the need to use three categories of indicators to predict banks’ financial stability: CAMELS, the indicators available to central banks, and the indicators of the external economic environment. Central banks can use internal payment system data to analyse banks’ financial conditions in real time. The use of external indicators is especially significant in Russia, given the wide disparities in the economic development across Russia’s regions.

JEL: H12