Determinants of Bank Closures: Do Levels or Changes of CAMEL Variables Matter?

Abstract

This study examines the role of levels and changes in bank balance sheet variables in explaining bank closure. Using a unique set of monthly bank-level panel data from Russia, we estimate determinants of bank license withdrawals during 2013M7-2017M7. We make two key findings. First, changes in CAMEL indicators are always significantly correlated with probability of bank closure, and the magnitude of parameter estimates decreases with the lag length. Second, while the one-month lagged levels of capital, earnings, and liquidity are significantly associated with the probability of bank closure in the subsequent month, the level of liquidity is the only significant indicator for longer lags. Our key contribution that changes in CAMEL variables matter more than levels is robust to various robustness checks.