The paper examines over seventy episodes of current account adjustment in industrial and major emerging market economies. It argues that these episodes were characterised by strongly divergent economic developments. To reduce this divergence, the paper classifies episodes with similar characteristics in three groups, using cluster analysis. A majority of cases was characterised by internal adjustment through a slowdown of domestic demand and did not involve significant exchange rate movements. In some cases, the adjustment was mainly external, facilitated by a relatively modest exchange rate depreciation and without economic slowdown. Finally, some cases involved a crisis-like combination of a severe slowdown and a significant currency depreciation. Using a multinomial logit, we find that this classification of episodes helps improve the predictability of current account adjustment.
cluster analysis current account adjustment external imbalances multinomial logit