The target zone proposal for exchange rate management has received considerable attention from both economists and policy-makers as a means of improving macroeconomic performance. For such an approach to be adopted, it is clearly a necessary condition that it should produce a better (expected) performance than existing policy-making procedures. Such improvements may be due, however, to the better design of fiscal and monetary policies, and hence to the increased effectiveness of conventional policy instruments, and not to the target zone scheme per se. This is therefore not a sufficient condition for a target zone mechanism to be incorporated into policy, which requires that the expected results should improve on the best that could be achieved in the absence of exchange rate targets.
In Discussion Paper No. 393, Research Fellow Andrew Hughes Hallett tests for the effectiveness of the target zone proposal for the G5 countries during the 1980s under both conditions, using the Federal Reserve Board's multi-country econometric model in a game-theoretic context. He finds that the gains under the necessary condition are ten times larger than those under the sufficient condition, which suggests that the major gains in performance arise from an increased effectiveness of conventional fiscal and monetary policies, rather than from exchange rate targeting itself.
Further experimenting with alternative feedback mechanisms for the specification of exchange rate target paths indicates, however, that target zones nevertheless have certain practical advantages. Exchange rate targeting offers policy-makers a simple and fairly uncontroversial means of preventing weak or undesirably competitive policies, but the policies produced by this `safety net' mechanism are not optimal and do not substitute well for explicit cooperation. On the other hand, targeting is helpful when stability in the foreign exchange markets is an explicit objective. Hence this `safety net' property is important for establishing stability, credibility and discipline, but not particularly effective in generating strictly efficient policy choices.
Hughes Hallett also notes that the distinction between necessary and sufficient conditions accounts for a number of apparent differences in the results of previous empirical tests of the target zone proposal, and that his own numerical results are supported by those obtained from other models, time periods and specifications of policy objectives.

Target Zones and International Policy Coordination: The Contrast Between the Necessary and Sufficient Conditions for Success
A J Hughes Hallett

Discussion Paper No. 393, April 1990 (IM)