In a unilateral exchange rate target zone such as those in Finland, Norway and Sweden, the central bank defends the target zone without the cooperation of other central banks, selling reserves when its currency is weak and buying them when it is strong. At some point target zones are usually realigned by a change in the currency's central parity or in the width of the band, or completely abandoned for another regime. Some argue, however, that unilateral target zones can be defended indefinitely, regardless of the level of reserves and without other countries' cooperation or capital controls, by focusing monetary policy exclusively on the target zone. In principle, domestic credit can be reduced to halt capital outflow, and domestic interest rates can be raised high enough to create a differential relative to the rest of the world sufficient to compensate for any devaluation risk.

In Discussion Paper No. 645, Bernard Dumas and Research Fellow Lars Svensson argue that constraints such as the political infeasibility of excessive interest rates or budget deficits that require finance via steady domestic credit expansion may preclude the exclusive focus of monetary policy on the target zone. There may also exist a floor for the money supply below which the assets of the central bank cannot be reduced. In such cases, the means to prevent reserve loss are limited and a target zone's `survival time' may depend on the current level of reserves. There will be critically low and critically high levels of central bank reserves at which the target zone is either abandoned or realigned.

Dumas and Svensson argue that the width of the target zone has only a negligible effect on the expected survival time, since this only matters after the exchange rate drifts to the weak edge of the target zone, which will be a small proportion of total unless reserves are very small. Rather, the predominant influences are reserve levels and the degree of real and monetary divergence between the country and the rest of the world. The survival time, they claim, is roughly proportional to the excess of reserves over the critical level and inversely proportional to the degree of real and monetary divergence. For plausible values of the model's parameters, they find that a 10% (annual) rate of divergence, together with a 100% buffer stock of reserves yields an expected survival time of 10-20 years. Since 10% per year is a very high level of divergence, a unilateral target zone seems to be a robust construct.


How Long Do Unilateral Target Zones Last?
Bernard Dumas and Lars E O Svensson

Discussion Paper No. 645, May 1992 (IM)