Skip to main content

The International Financial Architecture

Jeffrey A. Frankel

The recent financial crises in many emerging market economies have raised anew questions about the appropriate exchange-rate regime and the use of capital controls as policy instruments. The use of both mechanisms should be tailored to each country’s unique circumstances. Fixed exchange-rate mechanisms, such as dollarization (adopting the dollar as legal tender in place of the national currency), are suited to small open economies or those desperate to import monetary stability. Larger economies, such as the European Union (EU) and the United States, should allow their currencies to float. Intermediate regimes that fall between fixed- and floating-rate regimes–such as bands, baskets, and crawls (See Figure 1 for definitions)–are still appropriate for some countries. Certain well-targeted restrictions on the composition of capital flows might be appropriate for some emerging-market countries as temporary measures when inflows are particularly high.


The G-7 and the rest of the international financial policymaking community has, over the last eight months, made a variety of reforms to try to reduce the frequency and severity of international financial crises. The reforms include steps to improve transparency, strengthen financial systems, and involve the private sector more fully in rescue packages. These measures are worthwhile, regardless of whether critics are right that they are too small to merit the label new financial architecture.

However, innovations in several areas would be so fundamental as to merit unquestionably that appellation. One is the question of liberalization of international capital flows, and how rapid it should be. Another is the question of exchange rate regimes, and how flexible they should be.

No single currency regime is right for all countries or at all times. The choice of exchange-rate arrangement should depend on the particular circumstances facing the country in question. This proposition may sound obvious. But it needs to be stated. Recent experience has many lessons to offer. There is a danger of over-generalizing, however–of applying a given lesson to all countries regardless of circumstances. Five such propositions are currently heard.

Five Common Propositions Regarding Exchange-Rate Regimes

The first proposition is that countries should generally move to increased exchange-rate flexibility. We hear this particularly from policymakers who have tried over the last two years to help fight speculative pressures against exchange-rate targets in Thailand, Korea, Indonesia, Russia, and Brazil, where the attempts ended in costly crashes. When exchange rates float, there is no target that needs defending.

A second (diametrically opposed) proposition is that all countries should move toward enhanced institutional commitment to fixing the exchange rate. After all, none of those crisis-influenced currencies had been literally or formally fixed to the dollar. Enthusiasts point to currency boards that have successfully weathered the storms in Hong Kong and Argentina. Some even go further and suggest full official dollarization. They take encouragement from the euro-eleven’s successful move to a common currency on January 1, 1999.

A third proposition is that countries in general must move increasingly in either direction on exchange-rate regimes–free-floating or firm-fixing–but that the intermediate regimes such as target zones are no longer tenable (See Figure 1 for definitions). This proposition, too, is in danger of being applied too broadly.

Figure 1:
Definitions of Nine Major Exchange Rate Regimes, Ranged Along the Continuum from the Most Flexible to the Strongest Fixed-Rate Commitment

1. Free floating–the absence of regular intervention in the foreign exchange market
2. Managed float–the absence of a specific target for the exchange rate

3. Target zone, or band–a margin of fluctuation around some central rate
4. Basket peg–fixing not to a single foreign currency but to a weighted average of other currencies
5. Crawling peg–a preannounced policy of devaluing a bit each week
6. Adjustable peg–fixing the exchange rate, but without any open-ended commitment to resist devaluation or revaluation in the presence of a large balance of payments deficit or surplus

7. Truly fixed peg–fixing, committing to buy or sell however much foreign currency is necessary at a given exchange rate, with a firm and lasting intention of maintaining the policy
8. Currency board–three defining characteristics: fixing not just by policy but by law, backing increases in the monetary base one-for-one with foreign exchange reserves, and allowing balance of payments deficits to tighten monetary policy and thereby adjust spending automatically
9. Monetary union–the adoption of a foreign currency as legal tender. This includes the special case of official dollarization

The fourth proposition is the prediction that the world is breaking up into a few big currency blocs as European countries give up their currencies for the euro and Western Hemisphere countries give up theirs for the dollar. (The perceived trends in the first two blocs have recently become more plausible in light of the success of the European Monetary Union (EMU) and talk in Latin America of dollarization. Not so the yen bloc, however.)

The final proposition is the opposite of the fourth. Some, particularly among European leaders, believe that the most important reform is to stabilize the cross-rates between the dollar, euro, and yen and let smaller countries do what they will, rather than stabilizing the exchange rates within each bloc and letting the three major currencies continue to float freely.

Some Countries Should Fix Firmly

What are the characteristics that make a country more suited for fixed rather than flexible rates? The classic list includes: small size, openness to trade, high labor mobility, availability of a fiscal mechanism to cushion downturns, and a high correlation of the local business cycle with that of the country to which a currency peg, or fix, is contemplated. These attributes are well known among economists as criteria for political units to join in an optimum currency area (OCA). Countries that have these characteristics are likely to see big benefits from exchange-rate stability, and are also less likely to need monetary independence in the first place. Easy examples are Panama (dollar) and Luxembourg (euro).

The classic OCA criteria list needs to be updated, particularly if we are talking about prerequisites for the most rigid institutional arrangements–a currency board, full dollarization or a monetary union. Argentina, for example, is not an especially small open economy. But it has had a currency board since 1991 that has been largely successful in the face of severe challenges. (The government announced in January that it was considering abandoning the peso altogether in favor of official adoption of the U.S. dollar as legal tender.) First and foremost to be added to the list of criteria for a rigid peg is a strong need to import monetary stability due to either a history of hyperinflation, an absence of credible public institutions, or unusually large exposure to nervous investors. The willingness of Argentina to give up monetary independence derives from its past history of hyperinflation and a domestic political consensus that the experience must not be repeated.

It is also useful for a fixed-exchange-rate candidate to have extensive integration with one particular large trading partner or currency area, or a craving for future integration of this sort. The appropriateness of currency boards in Estonia, Lithuania, and Bulgaria, for example, derives from their desire for integration with the European Union.

The next requirements are access to an adequate level of reserves, and a strong, well-supervised and regulated financial system. Otherwise, the country might simply convert currency-crisis vulnerability into banking-crisis vulnerability. Finally, the existence of the rule of law is a necessary condition for a currency board, though not necessarily for dollarization. Otherwise, putting the currency peg in the law accomplishes little. Proclaiming a currency board does not, as is sometimes asserted, automatically guarantee the credibility of a fixed-rate peg (see Figure 1); a currency board is not credibility in a bottle. It is unlikely to be successful unless accompanied by solid fundamentals.

The Dollar and Euro Should Float Freely

Fixing is not the right course for all countries. To begin with the opposite extreme, the United States meets the criterion for an independent, free-floating currency. We have a large economy. The states of the union are more highly integrated with each other than they are with the rest of the world. There is more movement of trade, labor, and fiscal transfers–and a higher correlation of the business cycle–within our borders than across our borders. Fluctuations in the exchange rate are simply not as important to us as they are to most countries. We do not want to subordinate our monetary policy to conditions abroad. Thus, the advantages of floating overwhelm the advantages of fixing.

This does not mean that the U.S. authorities should never intervene in the foreign-exchange market at all. An occasional purchase or sale of foreign exchange is appropriate, if necessary to maintain in the marketplace a sense of two-way risk (rates can go up as easily as down), or to nudge the dollar exchange rate on those few occasions when it is far out of line.

Purely parenthetically, the question has arisen of what U.S. exchange rate policy might be like under the newly nominated Secretary of the Treasury. One cannot resist putting forward the alliterative theory of the dollar (Figure 2). Over the last twenty-five years, the secretaries whose names begin with the letter “B” have been perceived as bashing the dollar, as it depreciated during their terms, whereas the dollar was revived under secretaries whose names begin with “R.” In the mid-1970s, it remained steady under the last secretary whose name began with “S.” This is probably the appropriate characterization for the current outlook under Lawrence Summers, as well.

Figure 2:
Trend of Dollar Exchange Rates by Term of Treasury Secretary, 1974-1998
Percent Per Annum

Treasury Secretary Term G10 Effective
Exchange Rate
William E. Simon May 8, 1974 – January 20, 1977 2.48
W. Michael Blumenthal January 23, 1977 – August 4, 1979 -7.04
G. William Miller August 7, 1979 – January 20, 1981 3.28
Donald T. Regan January 22, 1981- February 1, 1985 15.28
James A. Baker, III February 4, 1985 – August 17, 1988 -17.10
Nicholas F. Brady September 15, 1988 – January 17, 1993 -1.26
Lloyd M. Bentsen January 20, 1993 – December 22, 1994 -2.18
Robert E. Rubin January 11, 1995 – December 31, 1998 1.58
Source: Federal Reserve Board (1999) and U.S. Department of Treasury (1999).

What of the new EMU? The eleven members do not satisfy the classic criteria for an OCA as well as the fifty American states do. And still less do the other four members of the European Union who have not yet decided whether to join. Labor mobility, income correlations, and federal fiscal cushions are all lower than within the United States. Nevertheless, trade integration and labor mobility within Europe are increasing over time, and as they do, the correlation of national business cycles can be expected to increase as well. Thus, European countries will, with the passage of time, come increasingly to satisfy the criteria for a common currency. But the theory of the optimum currency area suggests that the euro, dollar and yen should continue in the future to float freely against each other.

Are Intermediate Regimes No Longer Feasible?

Most countries are somewhere in between the big United States and little Luxembourg. Until recently, many experts believed that countries intermediate with respect to size, openness, and the other optimum currency area criteria were probably suited to intermediate exchange-rate regimes. However, suddenly the view has become common that such regimes are not sustainable in a world of large-scale financial flows and, therefore, countries are being pushed to the corners of either firm-fixing or free-floating.

What is the logic behind the proposition that countries must choose between firm-fixing and free-floating? At first glance, it appears to be a corollary to the principle of the impossible trinity. That principle says that a country must give up one of three goals: exchange-rate stability, monetary independence, or financial-market integration. It cannot have all three simultaneously (Figure 3). One can attain any pair of attributes–the first two at the apex marked capital controls, the second two at the point marked monetary union, or the other two at the point marked pure float. But one cannot be on all three sides simultaneously.

The general trend of financial integration has pushed most countries toward the lower part of the graph. If one is at the bottom leg of the triangle, the choice is narrowed down to a simple decision regarding the degree of exchange-rate flexibility. But even under perfect capital mobility, there is nothing to prevent the country from choosing an intermediate solution between floating and monetary union.

Whence, then, the hypothesis of the vanishing intermediate regime? Recent history makes it understandable that some would flee the soft middle ground of the intermediate regimes and seek the bedrock of the corners. Most of the intermediate regimes have been tried, and have failed, often spectacularly so. Contrary to claims that currencies in Mexico, Thailand, Indonesia, Korea, Russia, or Brazil were formally pegged to the dollar when their recent crises hit, they actually were following varieties of bands, baskets, and crawling pegs (See Figure 1 for definitions). Perhaps when international investors are lacking in confidence and risk-tolerance–the conditions that have characterized emerging markets from 1997 to 1999–governments can reclaim confidence only by proclaiming policies that are so simple and so transparent that investors can verify instantly that the government is, in fact, following them. Market participants can verify the announcement of a simple dollar peg simply by checking if today’s exchange rate differs from yesterday’s.

Alternatively, it may be that the search for a single regime that will eliminate currency speculation as an issue is doomed to fail (without capital flow restrictions). The rejection of the middle ground is then explained simply as a rejection of where most countries have been, with no reasonable expectation that the sanctuaries of monetary union or free-floating will, in fact, be any better. Therefore, a blanket recommendation to avoid the middle regimes in favor of firm-fixing or free-floating would not be appropriate.

When Should Capital Flows Be Restricted?

As noted, a country that has fully opened its financial markets cannot have both a fixed exchange rate and an independent monetary policy. If the exchange rate is fixed, then such a country must accept the interest rate that it is given by the world financial market. But a country with restrictions on the cross-border movement of capital can set both the exchange rate and the interest rate, without fearing a tidal wave of capital flows. One cannot help noticing that China, with its still heavily controlled capital account, has been able to withstand the recent turmoil in emerging markets to a better degree than Hong Kong, even though the latter has far sounder macroeconomic and structural policies.

A variety of measures to slow international capital flows are possible. The efficacy of restrictions on inflows is likely to be greater than controls on outflows, in part because it is easier to scare capital off than to keep it in against its will. Moreover, restrictions may have a useful role to play as a temporary measure when a country faces a large surge of inflows. After several years of controls, policymakers may have a better idea whether their country is the next tiger, justifying the inflows, or merely the subject of a speculative bubble.

Statistical evidence suggests that the composition of inflow is a significant leading indicator of the probability of currency crashes occurring. The higher the reliance on foreign direct investment, the lower the probability of crisis. The higher the reliance on foreign-currency borrowing that is short term and/or handled through banks, the higher the probability of crisis. This conclusion lends support to proposals for restrictions that would seek to change the composition of capital inflows, rather than changing the total magnitude. Penalties on short-term inflows might shift the composition toward longer maturities. Restrictions in Chile and Colombia appear to have succeeded in changing the composition of the capital inflow in this way.

One promising possibility is the idea of placing some penalty on short-term bank borrowing from abroad in foreign currency, perhaps in the form of requiring a non-interest-earning deposit with the central bank. The application of higher reserve requirements than are placed on other forms of bank borrowing need not be inconsistent with the pursuit of properly timed liberalization of the overall capital account. Rather, it would fall well within the kind of enhanced prudential banking regulation that the United States and the International Monetary Fund have for some time urged on developing countries, and for which the need has become increasingly obvious in the wake of the Asian crisis.


The overarching lesson regarding restrictions on capital flows is analogous to the lesson regarding exchange-rate regimes: there is no single answer that suits all countries. Circumstances are critical. As countries approach the level of economic development attained by the industrialized countries, they should evolve domestic financial systems that are correspondingly well-developed, and should achieve full integration into the world financial system. In the meantime, certain well-targeted inflow restrictions might be appropriate for some countries, particularly at certain stages of the boom-bust cycle.

In the case of exchange-rate regimes, other country attributes are relevant. Arrangements to fix exchange rates make the most sense for small open economies, or those that have a desperate need to import monetary and financial stability. Larger countries, like the United States, the European Union, and Japan, should float their currencies. Some countries that are intermediate in size and openness should continue to adopt intermediate regimes. While the dollar and the euro will be the two important international currencies in the coming decade, forecasts that most countries in the Western Hemisphere may or should give up their currencies in favor of the dollar, or those in the Eastern Hemisphere in favor of the euro or some other currency, are greatly exaggerated. Admittedly, however, emerging-market countries face problems that industrialized countries do not. Heightened financial integration means that finicky global investors must be satisfied, and, in some cases, this increasingly requires the kind of transparency or verifiability that is afforded only by the two corner solutions of firm-fixing or free-floating.

Get daily updates from Brookings