The role of information externalities in bank runs

In this dissertation, I provide an explanation of bank runs. Bank runs are depositors' response to unfavorable information about their banks. Since there is a moral hazard problem between bank managers and depositors, a bank run can be a mechanism for monitoring bank management through liquidating banks which are likely to make inefficient investments.;However, monitoring banks by runs may be inefficient. The first come, first served rule in the deposit contract creates a negative payoff externality among depositors, and gives depositors excessive incentive to withdraw. Even if we assume that depositors always choose the Pareto dominant equilibrium, bank runs may be triggered by very noisy early information even when depositors know that more precise information will be available in the near future. Deposit insurance which reduces the payoff externality can induce depositors to use their information more efficiently, and make runs effective mechanisms for monitoring banks.;Chapter 1 of the dissertation studies the determination of the deposit contract which maximizes depositor welfare. I find that in the optimal deposit contract, there is a negative payoff externality among depositors. Compared to the case where there is no payoff externality, bank runs occur more often and depositor welfare is lower. A co-insurance system which covers less than one hundred percent of deposits may improve the efficiency of bank runs.;The main issue studied in Chapter 2 is why bank runs are contagious. The failure of one bank provides depositors at other banks with noisy information about the prospects of the banking industry, and therefore the chance that an individual bank will fail. When there are both informed and less informed depositors at each bank, the less informed depositors have a strong incentive to respond to failures of other banks. The withdrawals made by these less informed depositors trigger inefficient bank runs. By offering deposit accounts with different insurance coverage, the insurer can induce the informed depositors to monitor banks, while preventing the uninformed depositors from making inefficient withdrawals. The feasibility of using mutual insurance among banks to reduce the possibility of inefficient runs is also discussed.