Currency bands entail an exchange rate float within a defined band with official intervention taking place only at the edges. In Discussion Paper No. 1031, Research Fellow Marcus Miller and Lei Zhang provide a theoretical rationale for such an ERM, showing that precommitment to narrower bands would yield a welfare gain. Their model has the central bank as an optimizing agent, seeking to stabilize the rate conditional on private sector expectations, and facing costs that are strictly proportional to the size of an intervention. The private sector determines &nbspthe behaviour of the exchange rate as a function of fundamentals and the intervention trigger.

A target zone model provides the framework to investigate the optimal and credible central bank policy. The bank can achieve some exchange rate stabilization simply by announcing its intention to support a currency band. However, since the exchange rate is taken to be a forward-looking asset price the behaviour of which is influenced not only by random walk velocity shocks but also by the expectation of future intervention, time inconsistency arises because the bank may be tempted to announce a narrower band than it will actually be willing to defend. This problem is solved by optimum control techniques which yield a credible band. At the same time, other non-credible solutions are observed which, if they could be rendered credible, would result in cost savings. Here, an ERM is a way of committing the bank to intervene at pre-specified bands. Numerical simulations suggest quite a significant effect: the ERM permits the optimal band to be reduced by half, so that bands of +/-4.5% chosen without precommitment fall to +/-2.25% with an ERM.

Optimal Target Zones: How an Exchange Rate Mechanism Can Improve Upon Discretion
Marcus H Miller and Lei Zhang

Discussion Paper No. 1031, June 1994 (IM)