A Genetic Approach to Finding Optimum Currency Areas

Atish R. Ghosh
International Monetary Fund

Holger C. Wolf
New York University
HWolf@imf.org

Abstract

The Maastricht treaty, monetary disintegration in Eastern Europe, and the problems besetting the CFA zone have rekindled interest in identifying optimal currency areas (OCA). Much of the literature to date examines whether a given candidate area satisfies the OCA criteria. In this paper, we adopt an alternative approach, determining the optimal decomposition of an area into a given number of monetary zones. The approach permits the calculation of the marginal cost of reducing the number of monies in a given area by one, and thus, in combination with an estimate of the benefits from reducing the number of monies, the optimal number of monies in a given area. From a computational perspective, however, this approach is extremely demanding. Accordingly, a genetic algorithm is developed and applied to identify optimal country groupings and the associated welfare costs of forming the monetary unions. For all country groups examined - the states of the United States and Canada, Europe, the G7, the CFA zone, and the world - we find that the estimated economic cost of a single currency exceeds plausible estimates of the benefits.