Analysis of systemic risk in the payments system

This paper investigates systemic risk in multilateral netting payments systems. A four-period model is constructed to investigate the effects of random liquidity shocks. There are three different types of agents in this model: banks, the payments system operator, and the central bank. Banks pay one another via the payments system. The payments system operator sets the rules for participation. These include total asset requirements, collateral requirements and net debit caps. The central bank serves as a source of liquidity during a financial crisis. The model is constructed along the lines of Diamond and Dybvig (1983). In period 0, banks optimize their holdings of noninterest-earning central bank reserves to meet their payment obligations and any additional liquidity obligation. Their alternative is to invest in a nonliquid asset that earns a rate of return R. In period 1, a number of banks are unable to make their payments. The number of banks defaulting is random and realized after banks decide their optimal reserve holdings. In period 2, the remaining banks must cover the net payments of defaulting banks minus the defaulting banks' collateral holdings. Each remaining bank has three options to meet its liquidity event: deliver reserves it holds at the time of the shock, borrow funds in the interbank market from banks in net credit positions, or default since it cannot meet its additional obligations. In period 3, final wealth of each due to bank is calculated. All banks want to maximize final wealth in period 3. ; The model provides the following results. The model calculates the threshold point where the payments system collapses. An interbank funds market increases the efficiency of the payments system. Implementation of policy options such as total asset requirements, collateral requirements, and net debit caps decrease systemic risk. The central bank's role as provider of liquidity to the financial system is investigated in the context of the model.