May 13, 1998
|Contact:||John Williamson||(202) 458-9008|
Washington, DC—The dollar is overvalued by at least 30 percent against the Japanese yen, in terms of sustainable medium-term currency relationships, according to a new study of G-7 exchange rates released today by the Institute for International Economics. This large misalignment helps to explain why the US trade and current account deficits are soaring to record levels, on the order of $250 billion in 1998, and why Japan's global trade and current account surpluses have already reached record levels in real terms.
The new study concludes that the dollar is also overvalued against the German mark, and the future euro, but by a considerably smaller amount on the order of 5-15 percent. This will permit the euro, if it actually starts up at such a rate at the beginning of 1999, to experience the modest appreciation that would naturally accompany its objective of becoming a "strong currency" without undue adverse effects on the European economy. All these calculations are derived by Rebecca Driver and Simon Wren-Lewis in Real Exchange Rates for the Year 2000, which updates earlier estimates of fundamental equilibrium exchange rates (FEERs) developed by John Williamson in earlier Institute publications.
The new study also concludes that current exchange rates among the French franc, German mark and Italian lira are consistent with their medium-term trends. Hence the initial conversion rates for Economic and Monetary Union (EMU) in Europe appear to be sustainable. By contrast, the pound sterling is overvalued against the DM by 15-25 percent so the United Kingdom was correct not to enter EMU at this time.
The concept of FEERs is based on estimates of sustainable current account positions, presented for all G-7 countries in an appendix to the new study by John Williamson and Molly Mahar, and calculations of the level of real (i.e., inflation-adjusted) exchange rates that will produce those balances. It seeks to avoid spurious precision by estimating fairly wide ranges for each FEER, consistent with the concept of target zones advocated by the authors. Driver and Wren-Lewis note that FEERs can also be viewed as forecasting exchange rates over the medium term, and that FEERs performed better than the main alternative concept (purchasing power parity) in predicting the exit of sterling and the lira from the European Monetary System in 1992.
These results do not in themselves suggest by how much current market exchange rates are "wrong." Economies where demand is weak, and where monetary policy is (and is expected to be) easy, are likely to have exchange rates in the short run that are undervalued compared to medium term trends, and vice versa. This goes some way to explain why the yen is currently undervalued and why the dollar and sterling are overvalued.
The main contribution of this study is to indicate how far the dollar, yen and sterling have moved away from sustainable levels. The analysis suggests that exchange rates were much closer to trend levels in 1995 than they are today. Hence the G-7, IMF and others should consider reform of the exchange-rate system as they continue their current reassessment of the "international financial architecture."
The study acknowledges that FEER calculations involve a large degree of uncertainty. They depend on the stability of historical trends in export and import behavior and on assumptions about trends in output, fiscal policy and private sector savings behavior. The study therefore includes a sensitivity analysis that indicates how FEERs would change under different assumptions about sustainable current accounts or trend output.
The attached table summarizes the authors' calculations of FEERs for the G-7 countries and compares them with private market rates. The main conclusions are that:
Fundamental Equilibrium Exchange Rates in 2000
|Sustainable current account
(as percentage of GDP)
(against the dollar)
|Present market rate
(against the dollar)