Bernanke (2005) builds on this previous work and he assumes that the US current account deficit is explained by a global saving glut mainly in emerging countries. The current account surpluses of emerging countries have two origins. Firstly, in the aftermath of the 1997 financial crisis, Asian emerging countries have accumulated reserves to cover against a possible sudden flight of foreign capital. At the same time, they maintain their exchange rates undervaluated which has boosted their exports and their growth rate. Secondly, because of oil prices soaring OPEC countries also emit a surplus of savings that they want to invest abroad, since their absorption capacity is limited due to their small population. The saving surplus of the emerging countries flows to the United States, the economy that presents the best features (Bernanke, 2005). These inflows of investments lead an appreciation of the dollar, an increase in asset prices and, after 2000, lower interest rates and an increase in household wealth. The combination of these factors motivates households to reduce their savings and increase consumption.
Since Chinn & Prasad (2003)’s seminal paper, empirical literature estimates current account determinants by running panel estimation on the current account balance (as a share of GDP) on several macroeconomic variables (fiscal policy, relative prices, demographic factors, financial series…). As demonstrated by Kao (1999), if the data include stochastic trends then OLS panel estimation could suffer from the spurious regression issue. This paper aims at testing the long-run determinants of the current account in 21 OECD countries with panel cointegration techniques. Fiscal policy, the real effective exchange rates, productivity and to a lesser extend real interest rates differentials are the main determinants of current account in the long run. Mean group panel DOLS estimation provides a robust framework to calculate equilibrium current account balance for each country. We then compute a measure of imbalance by subtracting this equilibrium value of the current account to the original series. A striking feature of this new measure of imbalance is that the observed current account does not fluctuate around its equilibrium value. There is significant imbalance in the current account, which can last for more than a decade. Countries in deficit (namely the United States and South European countries) as well as countries in surplus (e.g. Germany and Japan) have imbalances of their current account in surplus relative to this standard. Therefore the rebalancing of current account imbalances among OECD countries requires economic policy changes in both deficit and surplus countries.