Bailouts, monitoring, and penalties: an integrated framework of government policies to manage the too-big-to-fail problem

This paper discusses optimal government bailout policy where the costs of systemic failures and moral hazard problems are considered. We find that a three-tiered bailout policy that includes an <i>ex post</i> monitoring and bailout scheme for financial institutions with large systemic impacts ('too big to fail') is optimal. The optimal policy also requires a randomized bailout for medium-impact institutions ('Constructive Ambiguity'), and no bailout for institutions that have only minimal systemic consequences ('too small to save'). However, in a volatile, innovative market environment where individual institutions may know more than the government regulator, monitoring error could contribute to risk taking, leaving the government regulator to always play a 'catch-up' role in revising policy. Moreover, the optimal bailout policy may not be time-consistent: institutions not deemed 'too big to fail' may still have an incentive to take excessive risks and expect to be bailed out in case of insolvency, primarily due to the short-term orientation of the government. Finally, because an institution's systemic cost affects the probability of a bailout, we show that the boundary of an institution may be extended by the government subsidy.