Global Imbalances - Two Tales from the Recent Past for the G20

The FT reported yesterday that "Deep divisions over the sources of global economic fragility intensified on Thursday before the Group of 20 meeting of finance ministers and central bank governors, undermining talk of greater international co-operation. The US strengthened its rhetoric on inflexible exchange rates as the main causes of economic imbalances, while China’s president insisted that the most pressing concern was inadequate development of emerging economies." It may help if participants were reminded of some basic facts (just the facts, Ma'am!).
Global imbalances arguably started in the mid 1990s when the United States started to increase its deficit on the current account, a deficit that was covered by the sum of corresponding surpluses by 100+ countries around the world. My colleague Chris Garroway has kindly put up the video chart on http://vimeo.com/22433167. Watch it (clickstart on the bottom left), and watch it again: you will be able to see egg and hens, or hen and eggs.


Global reserves and imbalances in the current account, 1990-2008 from Chris Garroway on Vimeo.

Despite this evidence, it is increasingly clear also to the Chinese authorities that the renminbi has to appreciate further to dampen socially costly inflation and to help shrink the country's enormous surplus that has developed over the last decade. A new OECD Development Centre paper* shows that current account imbalances (savings-investment imbalances, that is) are not all structural: exchange rate changes can help bring them down. Using data for 128 countries between 1960 and 2008, we have found 25 episodes of large sustained exchange rate revaluations, which we define as both nominal and real effective exchange rate appreciations of 10% (and more) within a two year window (or less). We argue that these cases represent instances of exogenous appreciation shocks that we can use to estimate the macroeconomic impact of large appreciations and assess the robustness of estimates based on a wider definition of appreciation and revaluation events. Using a dummy-augmented autoregressive panel model we could indeed show that such large appreciations episodes have strong macroeconomic effects. Most importantly, we established four key stylised facts that can prove useful in the ongoing debate about the role of exchange rate adjustment for global rebalancing. The graph summarises, for the group of developing countries, the effects of (exogenous) appreciations on output and current accounts (and its components)over the subsequent ten years: 

• First, the current account balance typically falls strongly in response to large exchange rate revaluations. Three years after the revaluation, the current account balance deteriorates by about 3 pp. relative to GDP. This is due to a reduction in aggregate savings without a concomitant fall in investment. The effect on the current account balance is statistically significant and robust to variation in the country sample and the definition of appreciation events.
• Second, the effects on output seem limited. Our point estimates suggest a negative effect of output growth, albeit of relatively small magnitude: on average, the aggregate level effect on output amounts to about 1% after six years. The confidence intervals are also considerably wider than for the current account. The output effects are statistically not significant.
• Third, while aggregate output is not strongly affected, export growth falls significantly after appreciation shocks. Import growth remains by and large unchanged resulting in the observed deterioration in external balances. As aggregate economic growth is much less affected, our results point to a positive domestic demand response following appreciation episodes.
• Fourth, these effects seem to be more pronounced in developing countries. The sensitivity of the current account balance to revaluation shocks is higher. The effect reaches almost 4 percentage points of GDP after three years and is statistically significant. But also the potentially negative effects on output are larger. Our point estimates indicate a loss in output of 2% over ten years. But confidence intervals remain wide, so that these results miss statistically significant levels. Why these effects are stronger in developing countries will be an important question that we aim to address in future research.

In sum, the historical record of large exchange rate revaluations that we have studied in this paper lends some support to the idea that large exchange rate appreciations and revaluations have an impact on the current account as they lead to marked changes of savings and investment within countries. Appreciation shocks impact external balances, but this effect potentially comes at the cost of a reduction of dynamism in exports. While the domestic economy seems to pick up some of the external slack, leaving overall growth relatively unaffected, the prospect of sharp decelerations in export growth will remain a concern for policy-makers and bears watching especially in the context of developing countries.

*Kappler, Reisen, Schularick and Turkisch (2011), "The Macroeconomic Effects of Large Exchange rate Appreciations", OECD Development Centre Working Paper No. 296.