On the duration of large current account imbalances
The sustainability of "global imbalances" or large current account deficits are analyzed using duration or survival analysis models to determine the effect of key macroeconomic variables on the duration of large current account imbalances. For the emerging market economies most often associated with current account reversals and financial crises, real exchange overvaluation and high debt levels increase the probability a deficit spell will end with a swing into surplus. Contagion increases the likelihood of a current account reversal, where contagion is measured by the average current account balance for countries in the same geographic region and at roughly the same income level, Developing Asia for example. For advanced OECD economies the length of deficit spells are affected by domestic credit expansion, inflation and the interaction of reserves with the real exchange rate: that is a combination of a strong RER and high reserves reduces/increases the likelihood a deficit spell will end. We also identify a group of chronic deficit countries that on net accumulate foreign obligations over our entire study period 1970-2007. For this group, deterioration in the net barter terms of trade also hastens the end of current account spells. A key contribution of duration analysis to the study of current account deficit sustainability is to explicitly measure time dependence by using the Cox regression coefficients to compute the evolution of the hazard function over time for each sample of countries. For emerging market and chronic deficit countries the hazard rate is lower for short spells, but rises rapidly after about 20 years for developing countries, 22 years for chronic deficit, countries. High-income countries have a higher hazard rate of for short spells, but the hazard rate then falls over time. This suggests chronic current account deficits are risky for low- and middle-income economics, but that high-income countries can run current account deficits for long periods, or at least longer periods that the 38-year span covered by this study.
Kintanar, Kristine Angela Uy, "On the duration of large current account imbalances" (2009). ETD Collection for Fordham University. AAI3384640.