Essays on bank runs, contagion and systemic risk

The first essay investigates the role of herd behavior of depositors as a source of bank runs. I show that eliminating bank runs completely has some costs. Furthermore, a deposit contract that allows for runs can achieve higher levels of depositor welfare than a contract that completely eliminates them. Since early liquidation of bank's assets is costly, a central bank that acts as a lender of last resort alleviates some of the costs associated with bank runs. Yet it cannot prevent runs on healthy banks in the absence of perfect information about the bank's asset quality. In those cases, a deposit contract, even with liquidity support from the central bank, cannot achieve the first-best efficient outcome. Any effort to give market discipline a stronger role in achieving financial stability should be accompanied by transparency and disclosure of information.; The second essay investigates the factors that determine the severity of bank runs and points out possible policies that might dampen them. We show that the more information economic agents can expect to have about an ongoing financial crisis, the more willing they are to restrain themselves in withdrawing their funds from banks once a crisis actually occurs. We demonstrate that deposit insurance, even of a limited type, can also help to diminish the severity of bank runs. We also find that the presence of insiders is welfare increasing in the sense that when such insiders exist, subjects tend to withdraw their money.; The third essay investigates the interaction of banks in choosing the correlation in their investment portfolios and the implications of these choices on systemic risk. When bank loan returns have a systematic factor, the failure of one bank conveys adverse information about this systematic factor and increases the cost of borrowing for the surviving banks. Given their limited liability, banks herd ex-ante and undertake correlated investments to increase the likelihood of joint survival. If the depositors of a failed bank can migrate to the surviving banks, then herding incentives are mitigated and this gives rise to a pro-cyclical pattern in the correlation of bank loan returns.