In a recent IMF Staff position note  Olivier Blanchard and Gian Maria Milesi-Ferretti provide a useful classification of current account imbalances. They argue that deficits and surpluses on current accounts are "good" if they reflect optimal allocation of capital across time and space. That is the case, for example, when savings ratios differ across countries because of different ageing profiles or when investment ratios differ because of different productivity trends.
However, imbalances are “bad” if they reflect distortions that cause suboptimal saving or investment behavior. These distortions may range from lack of social insurance (creating too much household savings) or poor firm governance (creating unwarranted corporate savings) to excessive public borrowing or excessive build up of foreign exchange reserves. A widespread distortion is that borrowers commonly underestimate the volatility of capital flows and the related risks and consequently over-borrow.
So far, so good. But, as B&M acknowledge, it is far from easy to determine the character of actual imbalances. Were global imbalances  over the last decade good or bad? B&M provide an interesting assessment of the past and (predicted) future imbalances. And it shows indeed that such an assessment is not straightforward, as also their judgment is both defendable and debatable. But let’s not go into that now.
Let me state a more obvious point: You don’t have to be a fan of spaghetti westerns (but who isn’t?) to realize that something is missing in the analysis. If there is a good one, and a bad one, then there must be an ugly one, too. Which dimension is missing in the paper of B&M?
In my opinion the missing dimension is the imbalance between global demand and supply of goods and services. If a country runs a current account deficit (that means that, for either good or bad reasons, spending exceeds income) then that imbalance can easily get ugly if global spending already exceeds global production capacity. Conversely, in a situation of insufficient global effective demand current account surpluses are likely the ugly ones. More specifically, over the last decade the U.S. current account deficits (underpinned by ample creation of liquidity) would be ugly if there was already more than enough effective demand in the world. The surpluses in some European countries, in oil exporting countries, and more recently in China would be ugly if there was a chronic lack of effective demand.
To determine which imbalance was ugly we take the World Bank’s measure of global capacity utilization and add that to Figure 1 in B&M’s paper, which contains current account imbalances over time. (see chart below)
The increase in U.S. deficits  since 2003 coincided with a tightening global real economy, as reflected in our measure of global capacity utilization, which could also be illustrated with low unemployment numbers and which ultimately showed in sharp increases in commodity prices. In that period the deficits looked a bit uglier than the surpluses. That changed obviously in 2008 with the global crisis. Surpluses in oil-exporting countries and in China came down sharply, but they became uglier too, because concerns dramatically shifted to lack of global effective demand.
This third dimension not only provides a more complete description of the character of imbalances, it is also a crucial element of the policy debate. U.S monetary or fiscal policy should tighten if one worries about the U.S. current account deficit, but also if one worries about too much global demand. China should stimulate domestic demand if one thinks that China’s current account surplus is unwarranted, but also if one worries about insufficient global demand.
In that regard the discussion about exchange rate policies is an intriguing one. It is often presented as a way to reduce current account imbalances. The story goes as follows. If a country stimulates domestic demand, which would decrease its current account surplus, then the price of domestic goods must rise relative to the price of foreign goods. To prevent domestic inflation, a smooth way of achieving that relative price change is through appreciation of the nominal exchange rate. (By the way, as they compare two possible future scenarios, B&M argue that if in China increased domestic demand is not accompanied by appreciation of the currency, then the reduction in the current account surplus will not occur. That is unlikely. More likely is that it would lead to domestic inflation.)
In that sense the exchange rate is related to adjustments in current account, albeit in an indirect way.
However, the situation is very different if we see exchange rates as an independent policy instrument, independent of more fundamental changes in domestic demand. Then changes in nominal exchange rates have a lot more to do with stimulating or slowing down the economy. For example, an appreciation of their currency  does not necessarily decrease a country’s current account surplus. An appreciated currency makes a country less competitive and production will slow. As a result, the investment ratio will fall and a current account surplus has a tendency to increase. Conversely, countries can stimulate their economy through depreciation, which could create a boom, higher investment rates, and a decrease in their current account surplus, or an increase in their deficit.
From a global perspective, merely changes in exchange rates do little to stimulate or slow down the economy. The advantage of one country is the disadvantage of another country. That is why a focus on changes in domestic demand is more important than changes in nominal exchange rates, not only to correct good or bad current account imbalances, but also to restore equilibrium in the global markets for goods and services, and thus make the world economy less ugly.
Source: World Bank and IMF