Bank Runs and Intervention: Efficient Deposit Freeze Policy

Our paper is a study of a benevolent banking authority freezing deposits in response to a bank run that has already begun. After having announced the freeze, the banking authority involves a court system and verifies depositors’ types to define which fraction of depositors need to withdraw funds immediately, similar to the events that occurred during the crisis in Argentina in 2001. To analyze the effects of such ex-post intervention policies on ex-ante incentives of depositors to participate in the run, we extend the setting of Ennis and Keister’s model (AER 2009) by introducing the distortion in the banking authority’s verification ability. More specifically, we examine how imperfect verification ability and widespread deposit freezes affect the banking system fragility. We find that the banking system is more prone to bank runs when the banking authority has weaker verification abilities. Our result is explained by the lack of commitment of the banking authority to freeze deposits ex-ante optimally, once the bank run is underway, and this type of time-inconsistency problem prevails more if the banking authority fails to identify a sufficiently large fraction of depositors. By analyzing the ex-post intervention policies in the spirit of the influential work of Diamond and Dybvig (1983), we contribute to the existing literature on bank runs caused by self-fulfilling prophecies.