Reforming the International Monetary and Financial System

Comments: Let Exchange Rates Float

Alexander Swoboda, and Peter Kenen
Published Date:
December 2000
  • ShareShare
Show Summary Details

When I agreed to discuss this paper, I thought I was signing on to rebut arguments for greater fixity among major exchange rates, such as target zones. But when the draft arrived, I found instead an intelligently argued paper taking eminently reasonable positions—including “doubt on the effectiveness of arrangements based on numerical targets for exchange rates, such as target zones.” I will, therefore, say no more about target zones. Instead, I would like to take some time to question the premise that greater constancy in nominal exchange rates is something desirable per se, wholly apart from any practical difficulties in achieving it. Then I will turn my attention to the authors’ major concerns and policy recommendations.

To begin with, we all know that real exchange rates should adjust for a number of classical reasons, such as the following.

  • Differential productivity growth. If, for example, trend productivity growth in Japan exceeds that in the United States, as has been true for most of the postwar period, then the yen should exhibit an upward trend vis-à-vis the dollar—as indeed it has. But, while the process of international convergence toward equal productivity levels across nations has gone a long way, it has not stopped. And that means that productivity growth should continue to be faster, on average, in the laggard countries than in the United States, which maintains a productivity lead over the other G-7 countries.1 For that reason alone, a fixed yen-dollar or even euro-dollar exchange rate should not be viewed as any sort of long-run norm. Indeed, the forward exchange rates implicit in current long-term interest rates are forecasting a much higher yen five and ten years from now.

  • Terms-of-trade shocks. Real exchange rates also must change from time to time to accommodate terms-of-trade shocks. Gyrations in the relative price of energy are only the most prominent of such shocks, not the only ones. Large relative price changes among tradable goods may be less important within the G-7 than they are in the world as a whole, but they do occur. Canada’s sensitivity to various commodity prices is perhaps the most obvious example; and the Canadians have long understood that it is a good reason to float. But the fortunes of other rich nations—such as Norway, Finland, and Australia, to name just three—are also closely tied to various commodity prices. Real exchange rate flexibility is vital to such countries.

To these two classical reasons for real exchange rate flexibility, I would like to add a third, “Keynesian” reason:

  • Asynchronous business cycles. Currently, for example, the business cycles in the United States, Europe, and Japan are dramatically out of phase, with the United States at or near a cyclical peak, Japan (we hope) at a cyclical trough, and Europe somewhere in between the two. Under such circumstances, a cyclically high dollar and a cyclically low yen serve a valuable macroeconomic purpose—they shift aggregate demand away from the United States, which has too much, and toward Europe and Japan, which have too little. Fixed exchange rates would block that vital channel of international macroeconomic adjustment.

These three elements also provide reasons why nominal exchange rates should be flexible—unless national price levels assume the entire burden of adjustment, which is neither likely nor desirable. In addition, of course, nominal exchange rates should adjust to reflect differences in inflation rates across countries. It is true that inflation is now extremely low everywhere in the G-7, but it is, nonetheless, negative in Japan, about 1 percent in the euro area, and closer to 2.5 percent in the United States.2 Such differences call for changes in nominal exchange rates.

I conclude, therefore, that it is far from obvious that movements in nominal exchange rates among the three major currencies are something to be avoided. Having said that, I do sympathize with the authors’ two main concerns: that floating exchange rates have displayed both excessive (highfrequency) volatility and major (low-frequency) misalignments

Excessive volatility is, of course, in the eye of the beholder. A true believer in the wisdom of markets will assure you that every daily exchange rate movement is justified—and it will be hard to prove him wrong. Nonetheless, my impression is the same as Cœuré and Pisani-Ferry’s. It sure looks as if much of the high-frequency volatility in foreign exchange markets is disconnected from fundamentals.

But is that a big problem? I think not. At least within the three major currencies, it is relatively easy to hedge foreign exchange risk efficiently with forward contracts, futures, and derivatives. G-7 governments surely have more pressing problems to worry about.

Nonetheless, most of the modest and sensible institutional changes that Cœuré and Pisani-Ferry propose—including greater transparency, more information on positions, better supervision and monitoring, and so on—are far more likely to do good than harm. I have, therefore, no problem supporting them.3 But I doubt that adopting these proposals would reduce short-term volatility in foreign exchange markets very much. As everyone at this conference knows, exchange rates these days are essentially asset-market (not goodsmarket) prices; so we must expect the dollar, yen, and euro to behave more like IBM stock than like IBM computers. I tend to agree with the authors that the world would be a better place if exchange rates moved more like IBM stock than like Yahoo! stock. But I do not think we know how to accomplish that objective. Nor do I think it is terribly important.

Major currency misalignments are a more serious matter for policymakers, since they can lead to serious misallocations of resources—and they cannot be hedged very well. The same preliminary question arises again here, however. Do such misalignments actually occur? In particular, are there really cases when the authorities know better than the markets where exchange rates “should be”? True believers will again say no. The most extreme misalignment we have seen came in 1984 and early 1985, when the dollar soared to absurd heights. Yet many supporters of Reaganomics at the time claimed that the world was merely witnessing the rational market reaction to “morning in America.” In this debate, I once again ally myself with Cœuré and Pisani-Ferry. Major misalignments do occur, it seems to me, even if we cannot define them any better than the U.S. Supreme Court could define pornography—you know it when you see it.

But what can or should governments do to ameliorate such problems? The authors suggest that better international policy coordination would help minimize misalignments. That is fine, and cross-national coordination is probably a good thing for the world. But I would enter two major caveats.

First, no matter how much we internationalists praise cooperation and goodwill, national economic policies will sometimes be “inconsistent” in a world of sovereign states. Two prominent fiscal examples were Reaganomics in the early 1980s, which sent the dollar soaring, and German reunification after 1990, which nearly destroyed the ERM. When major fiscal tremors like that occur, shouldn’t exchange rates absorb part of the shock?

Second, there are times when national policies should be “inconsistent” in this sense. Reaganomics may have lacked any persuasive economic rationale, but there were good reasons for German reunification to increase the fiscal deficit in Germany relative to those in France, Britain, and Italy. Other examples occur whenever business cycles are out of phase. As I noted earlier, it is macroeconomically appropriate for the dollar to rise relative to the yen at present.

These two caveats severely limit what international policy coordination can be expected to accomplish. Still, I am not religiously devoted to pure floating. Dirty or managed floats with occasional interventions (especially to fight major misalignments) are fine with me. And I think such interventions sometimes—certainly not always—work. They will not work, of course, if central banks intervene with puny amounts of money against market forces that are strongly arrayed on the other side. But when market sentiment is amorphous, and especially when markets take the view that a currency is strongly over- or undervalued, forceful central bank intervention has some hope of pushing markets in the desired direction. The interventions that boosted the dollar in the spring of 1995 are a prominent recent example.

I want to close by mentioning what the authors call the “public good” aspect of G-3 currency rates—the fact that third countries, many of them poor, have a stake in the dollar-yen or dollar-euro exchange rate. Cœuré and Pisani-Ferry use the fact that movements in, say, the dollar-yen exchange rate can hurt poor countries to argue for greater exchange rate stability within the G-3.

I draw a different conclusion—emerging market countries should not peg their currencies to the dollar and rich countries (presumably via the IMF) should discourage them from doing so. An emerging market country that pegs its currency to the dollar is engaging in a risky financial practice that can, for example, make its exports uncompetitive on world markets if the dollar rises sharply. That, of course, is precisely what happened in Southeast Asia in the years leading up to the 1997 crises.

Pegged rates also tend to produce what I call “the fixed exchange rate bubble,” which works in the following manner: thinking that a dollar is just another name for, say, 25 baht or 2,500 rupiah, banks and businesses in emerging markets borrow dollars at low interest rates in international markets and then lend (or invest) in local currency at far higher interest rates. This so-called (and badly misnamed) “carry trade” produces a nice profit as long as the peg holds, but produces disaster when the exchange rate crumbles. During the 1990s, we saw the bursting of the fixed exchange rate bubble devastate one country after another—in Mexico, in Southeast Asia, in Russia, and so on. There is a simple solution—let the exchange rates float.4


Baumol (1986) first called attention to the convergence process among major countries. There is by now an extensive literature. For more recent data, see Maddison (1998).

These are (approximate) consumer price index inflation rates. Wholesale inflation rates are lower.

I do take issue with one change—the idea that the IMF should compute and publicly announce “equilibrium” exchange rates. This is liable to cause big problems for countries that fix their exchange rates.

For more on this subject, including arguments for borrowing in home currency rather than in dollars, see Blinder (1999).

    Other Resources Citing This Publication