C O N T E N T S 9Non-technical summary
AbstractWe investigate the role of settlement in a dynamic model of a payment system where the ability of participants to perform certain welfare-improving transactions is subject to random and unobservable shocks. In the absence of settlement, the full information first-best allocation cannot be supported due to incentive constraints. In contrast, this allocation is supportable if settlement is introduced. This, however, requires that settlement takes place with a sufficiently high frequency.
Non-Technical SummarySettlement has three defining properties. It is not a welfare-improving activity by itself. Rather, settlement involves a mere transfer of an asset between participants in order to fulfill the obligations created by previous transactions. Second, it takes place periodically. Finally, settlement gives the opportunity to all participants in the system to start afresh since, after settling their obligations, they are no longer liable to the system. In this paper we employ the dynamic model of a payment system developed in Koeppl, Monnet, and Temzelides (KMT, 2005) in order to study the role of settlement. Our main finding is that settlement is an essential part of an optimal payment system as it enables agents to engage in beneficial transactions that would otherwise not be realized.Our model departs from monetary economics and emphasizes the role of private information in a way related to the dynamic contracting literature. Our model displays the following properties. First, the introduction of periodic settlement rounds does not increase welfare by itself. The increase in welfare is accomplished indirectly through the interplay between settlement and intertemporal incentives. Second, the first-best is supportable only if settlement takes place with a sufficiently high frequency. Finally, payment system participants must exit the settlement stage with identical balances. In this sense, agents start afresh as their history does not affect their future transactions.