Recent years have seen renewed interest in the regulation of interbank markets. A review of the literature in this area identifies two gaps: first, the literature has tended to make ad hoc assumptions about the interbank contract space, which makes it difficult to generate convincing policy prescriptions; second, the literature has tended to focus on ex-post interventions that kick in only after an interbank disruption has come underway (e.g., open-market operations, lender-of-last-resort interventions, bail-outs), rather than ex-ante prudential policies. In this paper, I take steps toward addressing both these gaps, namely by building a simple model for the interbank market in which banks optimally choose the form of their interbank contracts. I show that the model delivers episodes that qualitatively resemble the interbank disruptions witnessed during the financial crisis. Some important implications for policy then emerge. In particular, I show that optimal policy requires careful coordination between ex-post and ex-ante interventions, with the ex-ante component surprisingly doing most of the heavy lifting. This suggests that previous literature has underemphasized the role that ex-ante interventions have to play in optimal interbank regulation.