IV Developments in Exchange Rate Arrangements and Markets

R. Johnston, and Mark Swinburne
Published Date:
September 1999
  • ShareShare
Show Summary Details

According to the IMF’s official classification of exchange arrangements, there has been a gradual move away from pegged regimes into more flexible arrangements. This classification scheme has, however, a number of shortcomings, and, in particular when members’ de facto exchange arrangements are taken into account, the trend to more flexible arrangements becomes much less marked. In fact, accounting for members’ de facto arrangements suggests that exchange rate targeting has remained the predominant exchange rate arrangement, and thus, the existing official classification scheme can be misleading in this regard. This section first discusses the trend in exchange rate regimes under the official classification scheme; second, it discusses the shortcomings of this scheme and proposes an alternative classification scheme; third, it reviews some of the factors underlying the evolution in exchange regimes; and finally, it reviews developments and features of foreign exchange markets and international payments instruments.

Characteristics of Members’ Exchange Rate Regimes According to the IMF Classification Scheme

The main characteristics of the official classification scheme, and the trend in arrangements classified under this scheme, are discussed in Box 1 and provided in Table 8. The percentage of members maintaining pegged arrangements declined from about 77 percent in 1975 and to 36 percent in 1997, while the percentage of countries with free floating exchange rate regimes increased from 12 percent in 1975, to 25 percent in 1997 (Figure 5). There has also been an increase in the recourse to intermediate arrangements, such as limited flexibility (mainly among industrial countries) and managed floating (mainly among developing countries).

Table 8.Exchange Rate Arrangements as of December 31, 19971
PeggedFlexibility Limited vis-à-vis a Single

Currency or Group of Currencies
More Flexible
Single currency2Currency compositesSingle


Other managed


U.S. dollarFrench francOtherSDROther

(Indian rupee)

Islamic State of5
Antigua and BarbudaBurkina FasoLatviaBotswana5Qatar6BelgiumBelarus
Argentina*CameroonBosnia and Herzegovina

(deutsche mark)
Libya4,5BurundiSaudi Arabia6DenmarkBoliviaAlbania
Bahamas, The5Central African Rep.Myanmar5Cape VerdeUnited Arab Emirates6FinlandBrazil5,8Armenia

(deutsche mark)
Djibouti*CongoBrunei Darussalam

(Singapore dollar)
DominicaCôte d’lvoireKuwaitItalyColombia11Congo, Democratic

Republic of the5
GrenadaEquatorial GuineaEstonia*

(deutsche mark)
MaltaLuxembourgCosta Rica

(Australian dollar)
SamoaPortugalCzech RepublicGambia, The
Marshall Islands12MaliSeychellesSpainDominican Rep.5Ghana
Micronesia, Federated

States of12

(South African rand)
Slovak Rep.15Ecuador13,5Guatemala

(South African rand)
VanuatuEl SalvadorGuyana
Palau12Nepal (Indian rupee)GeorgiaIndia
Panama12San Marino11

(Italian lira)
St. Kitts and NevisHonduras15,14Jamaica
St. LuciaSwaziland

(South African rand)
St. Vincent and

the Grenadines
Iran. Islamic

Rep. of5
Syrian Arab Rep.5Israel17Liberia
Kyrgyz RepublicMoldova
Lao P.D.R.Mongolia
Macedonia, former

Yugoslav Republic of
New Zealand
MalawiPapua New Guinea
NicaraguaSäo Tomé and

Pakistan5Sierra Leone
RomaniaSouth Africa
Solomon IslandsTrinidad and Tobago
Sri LankaUganda
Sudan5United Kingdom
SurinameUnited States
Tajikistan5Yemen, Republic of
Note: An asterisk, *, indicates countries with currency board arrangements.

Figure 5.Exchange Rate Arrangements of Member Countries

(In percent of total number of countries)

Sources: IMF, Annual Report on Exchange Arrangements and Exchange Restrictions, various issues.

1For continuity, the industrial and developing country classifications do not reflect the new World Economic Outlook reclassification of the three economies previously classified as developing countries. Thus, Israel, Korea, and Singapore are still included in the developing country group.

2Emerging market countries include (1) Asia Pacific: Indonesia, Korea, Malaysia, Philippines, Thailand, Singapore, and China; (2) South America: Argentina, Brazil, Chile, Colombia, Mexico, and Venezuela; (3) Eastern and Central Europe: Czech Republic, Slovak Republic, Poland, Slovenia, Hungary, and Russia; and (4) Others: India, Sri Lanka, Pakistan, Turkey, Israel and South Africa. From the group of Eastern and Central Europe, in 1979 only Slovenia is included, in 1985 Slovenia and Hungary are included, in 1991 Poland, Slovenia, Hungary, Czech Republic, and Slovak Republic are included. In 1997, all countries classified under Eastern and Central Europe are included.

In the industrial countries, over one half of the members, including countries participating in the exchange rate mechanism (ERM) of the European Monetary System (EMS), maintained pegged exchange rates until the early 1990s. The exchange rate crisis of 1992–93 within the ERM, however, brought greater flexibility to exchange rate regimes of several European countries as the ERM fluctuation bands were widened for most ERM members, and several countries suspended their formal pegs during the currency turmoil. Italy and the United Kingdom suspended their ERM membership, and Finland, Norway, and Sweden abandoned their link to the European currency unit (ECU) in Fall 1992.12

A number of countries, however, recently joined the ERM in a step toward joining the European Economic and Monetary Union (EMU). Finland and Italy (re) joined the ERM in 1996, and Greece, which maintained a de facto crawling peg mechanism under its formally announced managed floating regime, joined the ERM in 1998.

The gradual shift from fixed exchange rate systems to more flexible arrangements has been more clearly demonstrated in the developing countries. In 1975, 87 percent had some type of pegged exchange rate. By 1997, the percentage pegging their currencies had fallen to some 40 percent. A number of developing countries continued to pursue fixed rates in the context of monetary unions (e.g., the CFA franc and several Caribbean countries), and some reverted back to pegged systems after periods of applying more flexible arrangements (e.g., Angola, Argentina, Bulgaria, Guinea-Bissau, and Venezuela). Among the emerging market economies, the ratio of those members that maintained pegged exchange rates fell from around 40 percent in the late 1970s to less than 10 percent in 1997, a much smaller percentage than developing countries as a whole. Greater flexibility has in general taken the form of a managed float rather than an independent float.

In 1992, when most transition economies joined the IMF, about 41 percent adopted pegged exchange rates vis-à-vis a single currency, and an equal number adopted freely floating rates. A number of these countries were using the Russian ruble as legal tender at the time of accession to membership in the IMF. Most transition economies that joined in 1993, however, adopted more flexible regimes. By the end of 1997, 22 out of the 26 transition economies were pursuing more flexible arrangements, although more than half of this group manage their exchange rates, rather than letting them float freely. A few transition economies (e.g., Latvia and Lithuania) that started with free floating exchange rates moved to pegged regimes, and two countries, Argentina and Bulgaria, recently reverted back to a pegged exchange rate mechanism in the context of a currency board (Part III, Table A7).

Box 1.Exchange Rate Classifications

Following the collapse of the par value system and introduction of generalized floating of the major currencies in 1973, members’ obligations regarding their exchange rate policies changed significantly from those embodied under the Bretton Woods system. Under the Second Amendment of the IMF’s Articles of Agreement, members were formally given the freedom to choose their own form of exchange rate arrangements, subject only to minor limitations (namely, a peg in terms of gold and pursuit of multiple exchange rates), reflecting a recognition that the precise nature of such arrangements was of less importance than the manner in which a member conducts its policies under the exchange arrangements of its choice. Members agreed to comply with certain broad obligations, including to assure that their exchange rates and macroeconomic policies foster orderly balance of payments adjustment. The IMF would exercise firm surveillance over the exchange rate policies of members, and each member would be required to provide the IMF with the information necessary for such surveillance. Figure 5 shows the current status of members’ exchange rate arrangements.

Members are obliged to notify the IMF of the exchange arrangements of their choice, within 30 days of becoming a member and promptly thereafter of any changes in their arrangements. Based on these notifications, the IMF summarizes members’ arrangements by an exchange rate classification scheme that was first introduced in mid-1975 as part of its oversight function of the evolution of the international monetary system. This classification system grouped members’ exchange rate arrangements according to the degree of flexibility with which the arrangements have been implemented, and has been broadly unchanged for over 14 years.1

The classification scheme distinguishes between three main groups: pegged exchange rate arrangements, where the exchange rate is fixed vis-à-vis a single currency or a currency composite; limited flexibility, where the exchange rate is allowed to move within bands vis-à-vis a single currency or within a cooperative arrangement; and more flexible arrangements, in which the exchange rate is managed or allowed to float freely. The distinction between managed and independently floating arrangements aims to reflect the policy stance for full or limited market determination of the exchange rate. Under independently floating regimes, supply and demand is, in principle, in continuous equality, with intervention limited to smoothing excessive short-run fluctuations in the exchange rate without establishing a particular level for it. In countries with managed floating systems, the foreign exchange market does not necessarily clear without intervention by the central bank.

Fixed Rate Arrangements

Peg: Single Currency. The country pegs to a major currency—usually the U.S. dollar or the French franc—with infrequent adjustment of the parity.

Peg: Currency Composite. A weighted composite is formed from the currencies of major trading or financial partners. Currency weights are generally country-specific and reflect the geographical distribution of trade, services, or capital flows. They can also be standardized, such as those of the SDR and the ECU.

Flexible Rate Arrangements

Flexibility Limited vis-à-vis a Single Currency. The value of the currency is maintained within certain margins of fluctuation around the de facto peg, corresponding empirically to volatility within the regime of wider margins that preceded the Second Amendment.

Flexibility Limited: Cooperative Arrangements. This regime refers to countries in the exchange rate mechanism (ERM) of the European Monetary System (EMS) and is a conceptual cross between a peg of each EMS currency to others in the system (currently within wide margins), and a float of all EMS currencies jointly vis-avis non-EMS currencies.

More Flexible: Adjusted According to a Set of Indicators. The currency is adjusted more or less automatically in response to changes in selected quantitative indicators. A common indicator is the real effective exchange rate that reflects inflation-adjusted changes in the currency vis-à-vis major trading partners; another is a fixed, prean-nounced change. This category was merged with the More Flexible: Managed Float category below since 1997.

More Flexible: Managed Float. The central bank quotes and supports the rate but varies it frequently. Indicators for adjusting the rate are broadly judgmental, including, for example, the balance of payments position, international reserves, or parallel market developments, and adjustments may not be automatic.

More Flexible: Independent Float. Rates are market determined, with any intervention aimed at the moderating rate of change, rather than at establishing a level for the rate.

1The scheme was revised slightly after the Second Amendment of the Articles and in 1982.

Shortcomings of the Existing Classification Scheme and Proposed Modification

The existing exchange rate classification scheme has a number of shortcomings. As a result, a new exchange rate classification scheme has been developed.

Existing Classification Scheme

The existing exchange rate classification scheme has a number of well-recognized shortcomings.13 First, there is considerable ambiguity as to what the scheme is intended to measure. There is a wide spectrum of exchange rate regimes beyond the traditional “fixed versus flexible” exchange rate dichotomy used in the official classification scheme with each regime affording a varying degree of monetary policy independence, which is not always apparent from the existing exchange rate classification. This is best illustrated by the group of countries that are classified as following managed floating regimes but use a range of different approaches to limit the flexibility of the exchange rate and to assign varying roles to the exchange rate as a nominal anchor. Several countries, for example, informally peg their currencies rather than using a single currency or a currency basket (e.g., Maldives and the former Yugoslav Republic of Macedonia). In a few others (e.g., Brazil and Ukraine), the exchange rate is pegged within a horizontal band. In some others, the currency is allowed to depreciate periodically at a rate that is preannounced or adjusted according to a set of indicators, with (e.g., as in Israel and Poland) or without bands around the central parity (e.g., in Costa Rica and Nicaragua). In these groups of countries, the de facto pegged exchange rate arrangement imposes similar constraints on monetary policy flexibility as in those with formal pegged arrangements though the restrictiveness of such constraints would clearly depend on the degree of flexibility of the pegged arrangement. The remaining countries within the managed floating group influence the movements of the exchange rate through active intervention without specifying a preannounced path for it and thus without constraining their monetary policies to that end.

The difficulties with the present classification scheme are also illustrated by the group of countries that are classified as maintaining pegged exchange rates. In countries that operate currency board arrangements (CBAs) (e.g., as in Argentina, Estonia, and Djibouti), in which the currency of another country circulates as legal tender (e.g., the Australian dollar in Kiribati) or maintain a currency union under which a common currency circulates at par among the members (e.g., the CFA franc zone), there is little or no scope for discretionary monetary policies. However, in countries that peg their currencies, including those pegging to a currency composite (e.g., Bangladesh, Fiji, and Malta), or that maintain pegged systems that allow the exchange rate to fluctuate within bands (e.g., Iceland, and the Slovak Republic), there is some limited degree of monetary policy discretion, with the degree of discretion depending on the band width.

The second, and perhaps more important, shortcoming of the present classification scheme is that there are sometimes important differences between the official classification, based on members’ formally announced regimes, and the actual, de facto, exchange rate arrangements followed by members; such differences may reflect political considerations of the authorities, as well as policy dilemmas arising from trade-offs between various economic objectives. For example, some countries that are classified with pegged exchange rates have engineered frequent changes in the parity of their exchange rates (e.g., as in Bangladesh and Solomon Islands), making the arrangement less distinguishable from a more flexible regime. Other countries that are classified with managed or independently floating exchange rate arrangements have, de facto, followed fixed exchange rate arrangements (e.g., Georgia, Lebanon, and the Philippines until mid-1997). Some countries have also announced pegs to a currency basket but followed de facto pegs to a single currency (e.g., Thailand’s peg to the U.S. dollar until July 1997 and Jordan, which has formally announced its peg to the SDR, but follows a de facto peg to the U.S. dollar). This divergence between members’ formally announced and de facto regimes reduces the transparency of members’ policy actions and the role of their exchange rate regimes in the overall policy framework, thus making effective surveillance over the policies of member countries more difficult.

In the period 1994–97, 33 members’ exchange arrangements were reclassified from less flexible to more flexible arrangements, and 35 from more flexible arrangements to less flexible ones (see Part III, Tables A8 and A9 for further details on these changes). A closer look at the nature of these shifts indicates that some of the reclassifications from less flexible to more flexible involve the continued maintenance of exchange rate targeting of some form (e.g., in Venezuela and Hungary, the move from pegged exchange rates to greater flexibility took the form of a forward-looking crawling band system, thus maintaining the essence of the pegged regimes). In all, many of the exchange regime reclassifications in 1994–97 involved a move to intermediate exchange rate regimes from either pegged regimes or free floating systems, with about 50 percent of these intermediate regimes conveying the features of a pegged exchange rate system (e.g., El Salvador and the former Yugoslav Republic of Macedonia informally peg their currencies; Brazil, Croatia, Sudan, and Ukraine peg their currencies within formal or informal bands; Costa Rica maintains a crawling peg; and Honduras and the Russian Federation operate crawling band regimes). Among the 14 transition economies that manage their exchange rates, 9 do so in the context of regimes that share the features of a pegged exchange rate regime. The former Yugoslav Republic of Macedonia, for example, maintains an informal peg against the deutsche mark; Belarus (the latter part of 1997), Croatia, and Ukraine operate within formal or informal target bands; and Hungary, Poland, and Russia (between mid-1996 and the end of 1997) pursue crawling band regimes where the exchange rate is allowed to depreciate at a preannounced or forward looking rate. De facto arrangements indicate that a much larger percentage of the membership follows different forms of exchange rate targeting than is suggested by the official classification of members that peg their exchange rate. It also shows that exchange rate targeting has remained the predominant monetary arrangement followed by almost two-thirds of the membership (see Table 9, columns 3 and 4).

Table 9.Evolution of Pegged Exchange Rate Regimes in Official Versus De Facto Classification
Pegged exchange rates (official classification)81526536
Pegged exchange rates, including de facto fixed pegs and formal and informal bands1102669653
Pegged exchange rates, including de facto fixed pegs, formal and informal bands, crawling pegs, and crawling bands11187611463
Source: Information provided by the national authorities.

Expanded Information on Exchange Arrangements

In an attempt to overcome some of the shortcomings of the existing classification scheme, supplementary information on members’ exchange rate arrangements has been developed:

  • Additional information was provided on the nature of the exchange rate arrangements maintained by members that were classified as following pegged and managed floating arrangements under the present scheme, and on the choice of a nominal anchor in countries following floating exchange rates (Figure 6).

  • Supplementary information compared members’ formally announced exchange rate arrangements, as recorded in the current classification system, with their de facto arrangements, with the latter ranked to indicate the general role of the exchange rate as an anchor of monetary policy (Figure 7).

  • An additional dimension was introduced to indicate members’ choices of alternative nominal anchors in conducting monetary policy including the exchange rate anchor, targeting of monetary aggregates, and direct targeting of inflation (Figures 6 and 7). This additional information indicates that as members move toward greater exchange rate flexibility, they adopt additional or alternative anchors to ensure price stability. In some cases, countries maintain monetary targets along with exchange rate anchors (e.g., Greece, Poland, Sri Lanka, and so forth). A number of other countries (e.g., Australia, Canada, New Zealand, Sweden, and the United Kingdom) have chosen a direct targeting of inflation as the anchor of monetary policy, while some countries (e.g., Colombia, Finland, Israel, and Spain) combined inflation targets with exchange rate anchors, in some cases (e.g., Colombia and Israel) to reinforce price stability. It should be acknowledged, however, that it would not be possible, for practical reasons, to infer from Figure 7 which nominal anchor plays the principal role in conducting monetary policy. In addition, a number of countries adopt IMF-supported or other monetary programs, which implicitly impose limits on their credit aggregates such as net domestic assets or domestic credit (e.g., Peru and Romania). It should be recognized, however, that such credit aggregates are not effective nominal anchors, and countries typically missed targets for both money and inflation due to unexpectedly large foreign exchange inflows; thus many countries with independently floating exchange rates do not have an effective anchor.

Figure 6.Expanded IMF Classification of Exchange Rate Arrangements as of December 31, 19971

Source: IMF, Quarterly and Semiannual Reports on Exchange Rate Arrangements, various issues.

1 Member countries of the ERM that maintain an exchange rate regime of limited flexibility within a cooperative arrangement are not reported in this figure.

2The following countries also have an IMF-supported program: Argentina, Bangladesh, Benin, Bulgaria, Burkina Faso, Cameroon, Chad, Republic of Congo, Côte d’lvoire, Djibouti, Estonia, Gabon, Guinea Bissau, Jordan, Latvia, Lithuania, Mali, Niger, Senegal, and Togo.

3This group includes Benin, Burkina Faso, Cameroon, Central African Republic, Chad, Republic of Congo, Côte d’lvoire, Equatorial Guinea, Gabon, Guinea Bissau, Mali, Niger, Senegal, and Togo.

4This group includes Kiribati, Marshall Islands, Micronesia, Republic of Palau, Panama, and San Marino.

5This group includes Antigua and Barbuda, Argentina, Bosnia and Herzegovina, Brunei Darussalem, Bulgaria, Djibouti, Dominica, Estonia, Grenada, Lithuania, St. Kitts and Nevis, St. Lucia, and St. Vincent and Grenadines.

6This group includes Angola, Bahamas, Barbados, Belize, Bhutan, Comoros, Iraq, Jordan, Lesotho, Namibia, Nepal, Oman, Syrian Arab Republic, and Swaziland.

7This group includes Cyprus (+/-2.25%), Iceland (+/-6%), Libyan Arab Jamahiriya (+/-47%), and Slovak Republic (+/7%).

8This group includes Bangladesh, Botswana, Burundi, Cape Verde, Fiji, Kuwait, Latvia, Malta, Morocco, Myanmar, Samoa, Seychelles, Tonga, and Vanuatu.

9The following countries also have an IMF-supported program: Algeria, Bolivia, Croatia, Egypt, El Salvador, Ethiopia, Georgia, Hungary, Kazakhstan, Kenya, Kyrgyz Republic, Federal Yugoslav Republic of Macedonia, Mauritania, Pakistan, Romania, Russia, Thailand, Ukraine, and Uruguay.

10This group includes Bolivia, Costa Rica, Greece, Nicaragua, Solomon Islands, Tunisia, and Turkey.

11This group includes Algeria, Belarus, Cambodia, Czech Republic, Dominican Republic, Ethiopia, Kazakhstan, Kenya, Kyrgyz Republic, Lao Peoples Democratic Republic, Malawi, Malaysia, Mauritania, Mauritius, Nigeria, Norway, Romania, Singapore, Slovenia, Suriname, Tajikistan, Thailand, and Uzbekistan.

12This group includes the following: Chile (back ward-look ing crawl); Colombia, Honduras, Israel, Poland, Sri Lanka, Venezuela (forward-looking crawl); and Ecuador, Hungary, Russia, Uruguay (preannounced rate of crawl). Sri Lanka changed classifications to reflect the current information.

13This group includes Brazil (R$0.97-R$1.06/US$l), China (+/-0.3%), Croatia (HRK3.5-3.8/deutsche mark), Sudan (+/-2%), Ukraine (HRV 1.7-1.9/US$1), and Vietnam (+/-10%).

14This group includes Egypt, El Salvador, Georgia, Islamic Republic of Iran, Federal Republic of Macedonia, Maldives, Pakistan, and Turkmenistan.

15This group includes The Gambia, Ghana, Guinea, India, Jamaica, Korea, Mongolia, Philippines, São Tomé and Príncipe, Sierra Leone, South Africa, Switzerland, and Zimbabwe. Ghana, Guyana, Philippines, São Tomé and Príncipe, and Sierra Leone also have and IMF-supported or other monetary programs.

16This group includes Lebanon.

17This group includes Armenia, Azerbaijan, Haiti, Indonesia, Madagascar, Moldova, Mozambique, Peru, Tanzania, Uganda, and Zambia, which all have IMF-supported programs. This group also includes Albania, Mexico, Rwanda, Trinidad and Tobago, and Yemen, which have other monetary programs.

18This group includes Australia, Canada, New Zealand, Sweden, and the United Kingdom.

19This group includes Democratic Republic of Congo, Eritrea, Guatemala, Japan, Papua New Guinea, Paraguay, and the United States (which do not have explicit nominal anchor) and Afghanistan, Liberia, and Somalia (for which there is no recent information).

Figure 7.Exchange Rate Arrangements: Comparison of IMF Classification, De Facto Arrangements, and Choices of Monetary Anchors, as of December 31, 19971

1The number next to each pie slice indicates the number of countries following the classification outlined in the bottom axis.

2This exchange rate arrangement indicates the actual policy followed by the member country, as opposed to the formally announced exchange rate arrangement.

3This category indicates that the country adopts more than one nominal anchor (inflation targeting, exchange rate anchor, explicit monetary targets, or monetary targets of an IMF-supported program or other monetary program) in conducting monetary policy.

4This category indicates that there is no explicitly stated nominal anchor but the country monitors a multiple set of indicators in conducting monetary policy, or that no recent, relevant information is available for the country.

This additional information serves to bring out the similarities between exchange rate arrangements and use of nominal anchors by members whose exchange rate arrangements are classified under different regimes according to the official classification system. For example, the arrangements in which the exchange rate fluctuates within bands can be found under the formal peg arrangements, as well as in limited flexibility and managed floating systems. Of the group of countries that are officially classified as maintaining more flexible arrangements, 9 have de facto fixed pegs, 6 limit the flexibility of their exchange rates within some formally announced or informal fluctuation bands, 7 allow their currencies to depreciate at a preannounced rate, and 11 do the same within bands crawling at a preannounced rate.

This additional information also indicates that within the group of exchange rate arrangements classified as pegged, there has been a marked shift away from conventional single currency or basket pegs toward more rigid forms of pegged exchange rate regimes (such as CBAs, currency unions, and regimes in which the pegged currency is the legal tender) (Figure 8). Moreover, within the more flexibly managed floating systems, there has been a marked shift toward arrangements where the authorities manage their exchange rates without establishing a preannounced path. These trends suggest that members have been moving toward more extreme regimes and may support the view that, as capital flows increase, the more extreme regimes become more viable: namely, floating regimes, where the exchange rate moves regularly in response to market forces, and truly fixed exchange rates with the strongest commitment to the exchange regime, whose sustainability depends on the implementation of a consistent set of economic policies (see Fischer, 1996, and Eichengreen and others, 1998).

Figure 8.Evolution of Exchange Rate Regimes—Expanded Classification

Source: IMF, Annual Report on Exchange Rate Arrangements and Exchange Restrictions, various issues.

Possible Revised Classification Scheme

Not characterizing members’ exchange rate arrangements accurately may cloud the policy implications derived from the research and policy papers that base their analysis on the IMF’s classification scheme. Moreover, classifying members’ exchange rate arrangements based on their de facto rather than their formally announced policies could bring greater transparency to members’ policy actions and thus contribute to improve surveillance over the exchange rate policies of members. Bringing together information on both exchange rate arrangements and nominal anchors of monetary policy can also help make potential sources of inconsistency in the monetary-exchange rate policy mix more transparent and illustrate that different forms of exchange rate regimes can be consistent with similar monetary frameworks.

Accordingly, Table 10 and Box 2 present a revised classification scheme that addresses some key shortcomings of the existing system. Members’ exchange rate regimes could be based on their de facto regimes. In addition, the new classification would present members’ exchange rate regimes against alternative nominal targeting frameworks, with the intention of using both criteria as a way of providing greater transparency in the classification scheme. The scheme would continue to rank exchange rate regimes on the basis of the degree of flexibility of the exchange rate arrangement. However, it would distinguish between the more rigid forms of pegged regimes (such as CBAs);14 other (conventional) fixed peg arrangements against a single currency or a currency basket; exchange rate bands around a fixed peg; crawling peg arrangements; and exchange rate bands around crawling pegs. The degree of flexibility of the band arrangements would depend on the width of the bands chosen. It would also introduce a new category to distinguish the exchange arrangements of those countries that have no separate legal tender. Adopting the currency of another country as legal tender or joining a monetary union in which the same legal tender is shared by the members of the union are forms of ultimate sacrifices for surrendering monetary control, where no leeway is left for national monetary authorities to conduct monetary policies. Traditionally, the classification scheme has treated such arrangements as pegged regimes. It should also be noted that some countries with pegged exchange regimes may have a bimonetary legal tender. This characteristic has been proven to enhance stability (e.g., in Argentina) by facilitating orderly redenomination of financial assets in domestic financial markets during heightened uncertainty.

Table 10.New IMF Exchange Rate Classification as of December 31, 19971
Monetary Policy Framework
Exchange Rate Regime (number of countries)Exchange rate anchorMonetary aggregate targetInflation targeting frameworkIMF supported or other monetary programOther
Exchange arrangements with no separate legal tender (26)

currency as

legal tender
ECCM2CFA franc zone
KiribatiAntigua and


Burkina Faso*Central African

DominicaCote d’lvoire*
PalauSt. Kitts and

Mali*Congo, Rep. of*

San MarinoSt. LuciaSenegal*
St. Vincent

and the


Burkina Faso*



Congo, Rep. of*

Côte d’lvoire*






Currency board arrangements (6)Argentina*




Other conventional fixed peg arrangements (including de facto peg arrangements under managed floating) (38)Against a single currency (22)


Bahamas, The







El Salvador*4


Iran, Islamic Rep. of4





Macedonia, FYR*4






Syrian Arab Republic

Against a composite (15)



Cape Verde















El Salvador*4




Macedonia, FYR


Pegged exchange rates within horizontal bands (26)Within a cooperative arrangement (12)












Other band arrangements (6)









Saudi Arabia

Slovak Republic



United Arab






Crawling pegs (7)4Within a cooperative arrangement (12)

Costa Rica
Other band arrangements (14)




Solomon Islands5
Exchange rates within crawling bands (11)4Colombia*






Sri Lanka*



Sri Lanka*




Managed floating with no preannounced path for exchange rate (23)Czech Rep.







Kyrgyz Rep.











Independently floating (44)Ghana








São Tomé and Principe*

Sierra Leone*

South Africa




New Zealand


United Kingdom
















São Tomé and Principe*

Sierra Leone*


Trinidad and Tobago



Afghanistan, Islamic State of6

Congo, Dem. Rep. of the5


Gambia, The6





Papua New Guinea


United States5
Sources: Information provided by country authorities.

In addition to their exchange rate regimes, members would also be classified by their choices of alternative nominal anchors in conducting monetary policy. Countries that maintain multiple anchors of monetary policy appear in more than one column in Table 10. Replacement of the national currencies of countries participating in EMU with a common currency, the euro, will raise an issue for the classification scheme.

The exchange rates of the national currencies were irrevocably locked on January 1, 1999. National currencies had continued to circulate as legal tender in their country of issue since January 1, 1999. Since that time, they have acted, however, no longer as separate currencies but as subdivisions of the euro. Euro notes and coins will be introduced on January 1, 2002. National currencies will be withdrawn and after July 1, 2002 will cease to be legal tender. The question will arise how to classify the exchange arrangements of the individual countries of EMU. It would seem reasonable to classify the EMU members jointly under the new category for the countries that have no separate legal tender with the appropriate nominal anchor (e.g. direct inflation targeting). The proposed new classification became effective on January 1, 1999.

The new classification of members’ exchange rate arrangements in Table 10 shows that contrary to the trends that emerge from the existing classification scheme, exchange rate arrangements, in which the exchange rate plays the role of the nominal anchor of monetary policy, are still the dominant type of monetary arrangements. It is also interesting to note that a significant part of the IMF membership (about 56 percent) that maintains some sort of pegged arrangements seems to continue to adopt conventional peg arrangements, with or without bands.

The adoption of the new classification system would involve a departure from the long-standing practice of classification by members’ formally announced regimes. The revised scheme requires an interpretation by IMF staff of the actual exchange rate arrangements and the monetary anchors that are pursued by members. Such an understanding of members’ exchange arrangements is already a central part of the IMF’s surveillance over the exchange rate policies of its members in the context of Article IV consultations. The adoption of a new classification scheme would make the assessment a more formal requirement and have the advantage of making more transparent the role of members’ exchange rate arrangements as part of their overall macroeconomic and monetary policies.

Box 2.New IMF Exchange Rate Classification System

The IMF’s new classification system is based on the members’ actual, de facto, regimes that may differ from their officially announced arrangements. The system ranks exchange rate regimes on the basis of the degree of flexibility of the arrangement. It distinguishes between the more rigid forms of pegged regimes (such as currency board arrangements); other conventional fixed peg regimes against a single currency or a basket of currencies; exchange rate bands around a fixed peg; crawling peg arrangements; and exchange rate bands around crawling pegs, in order to help assess the implications of the choice of exchange rate regime for the degree of independence of monetary policy. This includes a category to distinguish the exchange arrangements of those countries that have no separate legal tender. The new system presents members’ exchange rate regimes against alternative monetary policy frameworks with the intention of using both criteria (i.e., policy frameworks and current regime) as a way of providing greater transparency in the classification scheme and to illustrate that different forms of exchange rate regimes could be consistent with similar monetary frameworks. The following explains the categories.

Exchange Rate Regime

Exchange Arrangements With No Separate Legal Tender

The currency of another country circulates as the sole legal tender or the member belongs to a monetary or currency union in which the same legal tender is snared by the members of the union. Adopting such regimes is a form of ultimate sacrifice for surrendering monetary control where no scope is left for national monetary authorities to conduct independent monetary policy.

Currency Board Arrangements

A monetary regime based on an implicit legislative commitment to exchange domestic currency for a specified foreign currency at a fixed exchange rate, combined with restrictions on the issuing authority to ensure the fulfillment of its legal obligation. This implies that domestic currency be issued only against foreign exchange and that new issues are fully backed by foreign assets, eliminating traditional central bank functions such as monetary control and the lender of last resort and leaving little scope for discretionary monetary policy; some flexibility may still be afforded depending on how strict the rules of the boards are established.

Other Conventional Fixed Peg Arrangements

The country pegs its currency (formally or de facto) at a fixed rate to a major currency or a basket of currencies, where a weighted composite is formed from the currencies of major trading or financial partners and currency weights reflect the geographical distribution of trade, services, or capital flows. In a conventional fixed pegged arrangement the exchange rate fluctuates within a narrow margin of at most ±1 percent around a central rate. The currency composites can also be standardized, such as those of the SDR (special drawing right). The monetary authority stands ready to maintain the fixed parity through intervention, limiting the degree of monetary policy discretion; the degree of flexibility of monetary policy, however, is greater relative to currency board arrangements (CBAs) or currency unions, in that traditional central banking functions are, although limited, still possible, and the monetary authority can adjust the level of the exchange rate, although infrequently.

Pegged Exchange Rates Within Bands

The value of the currency is maintained within margins of fluctuation around a formal or de facto fixed peg that are wider than ±1 percent around a central rate. It also includes the arrangements of the countries in the exchange rate mechanism (ERM) of the European Monetary System (EMS) (replaced with ERM-II on January 1, 1999). There is some limited degree of monetary policy discretion, with the degree of discretion depending on the band width.

Crawling Pegs

The currency is adjusted periodically in small amounts at a fixed, preannounced rate or in response to changes in selective quantitative indicators (past inflation differentials vis-à-vis major trading partners, differentials between the target inflation and expected inflation in major trading partners, and so forth). The rate of crawl can be set to generate inflation adjusted changes in the currency’s value (“backward looking”), or at a preannounced fixed rate below the projected inflation differentials (“forward looking”). Maintaining a credible crawling peg imposes constraints on monetary policy in a similar manner as a fixed peg system.

Exchange Rates Within Crawling Bands

The currency is maintained within certain fluctuation margins around a central rate that is adjusted periodically at a fixed preannounced rate or in response to changes in selective quantitative indicators. The degree of flexibility of the exchange rate is a function of the width of the band, with bands chosen to be either symmetric around a crawling central parity or to widen gradually with an asymmetric choice of the crawl of upper and lower bands (in the latter case, there is no preannouncement of a central rate). The commitment to maintain the exchange rate within the band continues to impose constraints on monetary policy, with the degree of policy independence being a function of the band width.

Managed Floating With No Preannounced Path For the Exchange Rate

The monetary authority influences the movements of the exchange rate through active intervention in the foreign exchange market without specifying, or precommitting to, a preannounced path for the exchange rate. Indicators for managing the rate are broadly judgmental, including, for example, the balance of payments position, international reserves, parallel market developments, and the adjustments may not be automatic.

Independent Floating

The exchange rate is market determined, with any foreign exchange intervention aimed at moderating the rate of change and preventing undue fluctuations in the exchange rate, rather than at establishing a level for it. In these regimes, monetary policy is in principle independent of exchange rate policy.

Monetary Policy Framework

Members’ exchange rate regimes are presented against alternative monetary policy frameworks to present the role of the exchange rate in broad economic policy and help identify potential sources of inconsistency in the monetary and exchange rate policy mix.

Exchange Rate Anchor

The monetary authority stands ready to buy and sell foreign exchange at given quoted rates to maintain the exchange rate at its preannounced level or range (the exchange rate serves as the nominal anchor or intermediate target of monetary policy). These regimes cover exchange rate regimes with no separate legal tender, CBAs, fixed pegs with and without bands, and crawling pegs with and without bands, where the rate of crawl is forward looking.

Monetary Aggregate Anchor

The monetary authority uses its instruments to achieve a target growth rate for a monetary aggregate (reserve money, Ml, M2, and so forth) and the targeted aggregate becomes the nominal anchor or intermediate target of monetary policy.

Inflation-Targeting Framework

A framework that targets inflation involves the public announcement of medium-term numerical targets for inflation with an institutional commitment by the monetary authority to achieve these targets. Additional key features include increased communication with the public and the markets about the plans and objectives of monetary policymakers and increased accountability of the central bank for obtaining its inflation objectives. Monetary policy decisions are guided by the deviation of forecasts of future inflation from the announced inflation target, with the inflation forecast acting (implicitly or explicitly) as the intermediate target of monetary policy.

IMF-Supported or Other Monetary Program

An IMF-supported or other monetary program involves implementation of monetary and exchange rate policy within the confines of a framework that establishes floors for international reserves and ceilings for net domestic assets of the central bank. As the ceiling on net domestic assets limits increases in reserve money through central bank operations, indicative targets for reserve money may be appended to this system.


The country has no explicitly stated nominal anchor but rather monitors various indicators in conducting monetary policy, or there is no relevant information available for the country.

As a practical matter, it could be envisaged that the Article IV consultations would be a first step in discussing with the authorities, in cases where there are significant divergences between de facto and formally announced regimes and the merits of reporting more realistic descriptions of exchange rate regimes. This should also help minimize the potential risks of misinterpretation by IMF staff of the true nature of the members’ arrangements, as well as any potential disagreements with the authorities regarding the description of the nature of their arrangements. Exchange regimes for a particular currency can vary between Article IV consultations; thus, keeping the classification scheme current would require more frequent monitoring of monetary frameworks and exchange of information with the authorities. Both schemes would be maintained during a transition period over which the de facto and formally announced regimes are expected to conform with each other.

Factors Underlying the Evolution of Exchange Rate Regimes

Although less pronounced than suggested by the IMF classification scheme, there has been, in recent years, a shift away from pegged regimes to more flexible ones. This reflected a variety of factors including the changing economic conditions and policy objectives of countries over time, the liberalization and globalization of financial markets, and the accompanying increase in capital mobility.

Tensions Between Economic Objectives

The emergence of tensions between the objectives of lower inflation and external competitiveness over short time frames has been one significant factor in countries’ adopting more flexible exchange rate arrangements. During the initial phase of stabilization, many countries adopted pegged exchange rate regimes to help stabilize inflationary expectations and increase economic policy credibility. The advantages of fixed exchange rates have been particularly relevant for countries that suffered from hyperinflation and those in the process of transition to a market-based economy. The disadvantage is that fixed exchange rates have also been associated with sizable appreciations in real exchange rates reflecting the period needed to reduce inflation rates to the levels of major trading partners. In some cases, real exchange rate appreciations reflected the choice of an exchange rate peg that did not reflect the pattern of international trade, such as in some Asian economies. Real appreciations have in turn been associated with a deterioration of competitiveness and external imbalances, particularly when exchange rate policy was not adequately supported by consistent structural policies to improve competitiveness, or where the real exchange rate was not significantly undervalued initially. In such circumstances, a number of countries have abandoned pegged exchange rate arrangements. In some, the pegged exchange rates were abandoned as a result of speculative attacks. Factors that contributed to such crises included contagion effects, inconsistencies of fiscal policies with the pegged exchange rate, as well as severe problems in the financial system, high public or private indebtedness, and economic slowdown, which made it costly for the authorities to maintain and defend fixed parities (e.g., Finland, Italy, the United Kingdom, and Sweden in 1992; Mexico in 1994; and the Czech Republic, Indonesia, Korea, the Philippines, and Thailand in 1997). In several other countries, the move to greater exchange rate flexibility has been the result of a deliberate policy choice. Some countries adopted a progressive, step-wise approach toward more flexible exchange rates, and others a pragmatic approach.

A number of members have sought to address the tensions between inflation and external objectives by adopting crawling band arrangements. Chile, Israel, and Poland, for example, started their stabilization efforts with fixed exchange rates but moved to crawling pegs or horizontal bands, and eventually to crawling bands in an attempt to allow for greater exchange rate flexibility that, nevertheless, continued to maintain the anchor role of the exchange rate. Such arrangements retain an anchor role for the exchange rate by committing to a preannounced schedule of mini-devaluations, while, at the same time, avoiding serious real exchange rate misalignment that would hurt competitiveness.15 The choice of the rate of crawl of the band in general reflected differing policy priorities assigned to price stability and competitiveness objectives: in most cases, the rate of crawl was set in a “forward-looking” manner where the authorities assigned a greater weight to disinflation; in a few cases, the rate of crawl was determined in a “backward-looking” manner where competitiveness was given a higher priority (see Part III, Table A10). As indicated in Figure 8, the number of members adopting crawling band arrangements increased significantly in recent years (from about 10 percent of the managed floating arrangements in 1991 to about 22 percent in 1997), reflecting the advantages of such arrangements to minimize the trade-off between inflation and competitiveness objectives, as well as to provide greater monetary policy autonomy in the face of increasing capital mobility (see below).

Moves Toward Currency Convertibility

Moves toward greater currency convertibility have been associated with the adoption of more flexible, market-based exchange rates. First, as countries eliminated exchange restrictions on payments and transfers for current international transactions and liberalized capital movements, conditions were created for the development of domestic foreign exchange markets where exchange rates could be determined in a more flexible manner in response to supply and demand conditions. There is a positive correlation (0.87) between the share of members with Article VIII status and with flexible exchange rate regimes (managed or free floating regimes). The percentage of countries with Article VIII status increased from about 35 percent in 1978 to 78 percent in 1997, while the share of members with flexible exchange rates more than doubled within the same period.

Second, the elimination of exchange restrictions and in particular multiple exchange rates in itself often involved the adoption of market-determined exchange rates. Recent trends in multiple exchange rate practices are discussed in more detail below. Of a total of 39 countries that moved to greater exchange rate flexibility during 1991–94, 14 adopted floating rates upon unification of exchange rates and 9 introduced interbank foreign exchange markets or auctions and shifted transactions to the free market.16 In 1994–97, the elimination of multiple exchange rates was also associated with the move to floating exchange rates in the case of Angola, Azerbaijan, Lao People’s Democratic Republic, Venezuela, Yemen, and Zimbabwe, and with the introduction of interbank foreign exchange markets (e.g., Guinea, Madagascar, Mauritius, and Papua New Guinea). Of the 43 countries that maintained multiple exchange rate systems at the end of 1997, about 56 percent pursued less flexible exchange rate regimes (Table 11).17 The exchange rate classification scheme classifies members’ arrangements based on the dominant foreign exchange market, and thus, multiple exchange rate systems are also found in some countries with floating exchange rate systems.

Table 11.Exchange Rate Regime and Various Aspects of Exchange Systems, End of 1997(Number of countries)
Exchange Rate



Multiple Exchange Rate SystemsMemo Item:

Total Members
With MCPsWithout MCPsTotal
Pegged exchange rate4047551066
Limited flexibility1100016
Managed float (with preannounced exchange rate path)1618861429
Managed float (with no preannounced exchange rate path)1418911021
Independent float152881949
(In percent of total members within each exchange arrangement)
Pegged exchange rate60.671.27.67.715.2100.0
Limited flexibility6.
Managed float (with preannounced exchange rate path)
Managed float (with no preannounced exchange rate path)66.785.742.95.347.6100.0
Independent float30.657.116.31.918.4100.0
Source: IMF, Annual Report on Exchange Arrangements and Exchange Restrictions (1997).

Third, many countries support their pegged exchange rates through administered constraints on the sale of export proceeds and have abandoned pegged exchange rates as they eliminate such constraints. Some two-thirds of countries that still require export receipts to be repatriated and surrendered to the central bank or the banking system maintain, formally or informally, pegged exchange rate regimes.

The relationship between the choice of the exchange rate system and the degree of restrictiveness of the exchange regime, for a sample of 41 countries for which the degree of restrictiveness of the current account, capital account, and the overall exchange regime has been calculated in terms of an index, is examined in Part III, Table A11. The countries identified as having more restricted exchange regimes seem to pursue more flexible exchange rate systems according to the official exchange rate classification. However, taking into consideration the members’ de facto policies shows that more than 75 percent of the countries with more restricted exchange regimes have some kind of a pegged exchange rate arrangement. Among those countries with less-than-average restrictiveness, no clear pattern is identified. This is consistent with the fact that the introduction of convertibility did not involve floating exchange rates in a number of cases where monetary policy was clearly subordinated to maintaining the exchange rate objective.

Responses to Increased Capital Flows

The increase in capital flows places a premium on countries following consistent monetary and exchange rate policies. Thus, countries attempting to maintain a fixed or a crawling peg have a limited capacity to set their domestic interest rate independent of the foreign interest rate without risking significant capital flows.

In general, the first policy reaction to strong capital inflows with a pegged exchange rate has been to conduct sterilized intervention to limit the impact of such flows on monetary aggregates. However, such intervention involves quasi-fiscal costs and is generally of limited effectiveness since it serves to keep interest rates high, attracting further capital inflows. A variety of other instruments and policies have, therefore, been brought into play in responding to capital inflows. These included a tightening of fiscal policy, imposition of controls on capital inflows, relaxation of controls on outflows, and a tightening of financial and prudential regulations, or a combination of a number of different measures.18

As an alternative response to increased capital inflows, and in many cases as a supplementary measure, a number of countries have moved toward more flexible exchange rate arrangements, inter alia, to insulate the domestic money base from the expansionary effects of capital inflows and to increase perceived exchange risk that may help deter speculative and potentially destabilizing short-term inflows. Several countries allowed their exchange rate to adjust through an appreciation of the exchange rate in a floating system; a revaluation of the pegged exchange rate; and an introduction of greater flexibility within their pegged or managed exchange rate systems (see Table 12). The latter has taken many different forms, including the introduction of fluctuation bands around a central parity, a move from horizontal band systems to crawling bands, successive adjustments in the band width, or the abandonment of the pegged systems altogether and moving to managed or floating exchange rates.

Table 12.Exchange Rate Regime Response to Capital Inflows
CountryExchange Regime at Start of Inflows (De Facto Regime)Exchange Regime Response
BrazilIndependent floatAllowed the exchange rate to appreciate within a flexible regime.

Switched to managed float in 1994 and widened the de facto band in 1995-97.
ChileManaged float (crawling band)Widened the band width five times in 1988–97 from ±3 percent to ±12.5 percent in 1997.

Revalued the central rate several times in 1992–96.
ColombiaManaged float (crawling band)Allowed the exchange rate to move within the de facto ±7 percent crawling band.

Revalued the band twice in January and December 1994.
Czech Rep.Basket pegWidened the trading band from ±0.5 percent to ±7.5 percent in February 1996.

Abandoned the horizontal band in May 1997 and moved to managed float with no preannounced path.
HungaryBasket pegSwitched to managed float in 1995.

Reduced the rate of crawl twice in 1997.
IndonesiaManaged float (crawling peg)Introduced a crawling band in 1994.

Widened the band width seven times in 1994–97.

Switched to free float in August 1997 under market pressure.
IsraelManaged float (horizontal band)Switched to crawling band in January 1992.

Widened the band three times in 1993–97 from ±5 percent to ±7 percent then to ±15 percent, finally widened only the upper band keeping the lower band unchanged.
KoreaManaged floatIntroduced a horizontal band, widened the band width in steps in 1992–95.
MalaysiaBasket peg with bandMoved to managed float with no preannounced path for the exchange rate and allowed for greater degree of flexibility.
MexicoManaged float (crawling peg)Moved to a gradually widening crawling band in November 1991, but in practice heavy intramarginal intervention was used to keep the rate within a tight inner band.
PeruIndependent floatExchange rate was allowed to adjust within a flexible regime.
PhilippinesIndependent floatExchange rate was initially allowed to adjust within a flexible regime.

Starting from 1995 a de facto peg was adopted.
PolandManaged float (crawling peg)Switched to crawling band in 1995, but in practice kept a narrower inner band.

Revalued in 1996 and reduced the crawl rate in five steps in 1991–96 to curb appreciation.
RussiaManaged float (de facto horizontal band)Switched to crawling band in 1996.

Adjusted the hand width on a number of occasions.
South AfricaIndependent floatExchange rate was allowed to adjust within a flexible regime.
Slovak Rep.Basket pegWidened the trading band in two steps from ±1, 5 percent to ±7 percent in 1996–97.
TurkeyManaged float (crawling peg)Devalued the lira in 1994 and allowed for more flexibility within a flexible regime.
VenezuelaIndependent floatExchange rate was allowed to adjust within a flexible regime in response to market forces in early 1990s.

Subsequently moved to managed float, and then to single currency peg, and back to managed float with a crawling band in 1996.

In some countries, the prolonged maintenance of pegged exchange rates may have been seen as implicit exchange rate guarantees and encouraged external borrowing. This, in turn, led to excessive exposure to foreign exchange risk in the financial and corporate sectors and increased vulnerability to a sudden change in market sentiment. In several Southeast Asian countries (Indonesia, Korea, the Philippines, and Thailand), as well as Mexico in 1994, floating the exchange rate was the main response when countries faced a subsequent significant outflow of capital and a weak international reserve position, and when increases in domestic interest rates proved insufficient to encourage a reversal of the capital outflows. Among other factors that contributed to the crises in these countries are a buildup of overheating pressures; lack of enforcement of prudential rules and inadequate supervision of financial systems, together with direct-lending practices by the government; problems in disclosure and transparency of data and information and in governance; and movements in international competitiveness due to wide fluctuations in exchange rates.

Developments in Foreign Exchange Markets

At a global level, exchange rate arrangements have generally become more market-based, with greater reliance on interbank exchange markets as the core of wider currency markets in developing and emerging market economies. At the same time, in the most developed markets, traditional interbank foreign exchange trading appears to be declining somewhat in relative importance.

Foreign Exchange Markets in Industrial Countries

In the major industrial countries, the importance of traditional interbank operations in spot and forward foreign exchange markets appears to have been declining in importance relative to other forms of foreign exchange trading, especially trading in derivatives. As discussed in Part II, Section IX, Bank for International Settlements (BIS) data on total foreign exchange trading for 1995 show continued growth of around 30 percent between the triennial surveys, with total surveyed trading averaging $1.26 trillion daily. Traditional spot, forward, and swap transactions still account for the majority of this by far, but trading in currency futures and options is growing rapidly, and accounted for around 6 percent of daily turnover in 1995. A similar pattern is reflected in bank income trends: foreign exchange dealing as a profit center for major international banks is reported to be declining, with greater emphasis being placed on client business, especially in relatively complex derivative products. Other factors contributing to the trend include the growing power of nonbank financial institutions and other end users that, inter alia, has made it somewhat more difficult for traditional dealers to read short-term market developments (their traditional speciality); intense competition from electronic deal-matching systems; and to some extent also the convergence toward EMU. More generally, greater exchange rate confidence and lower volatility reduce the demand for speculative position taking and, to that extent, the return to market making. With reduced volatility of the exchange rate, the number of institutions actively making markets often tends to decline.

At the same time, a trend toward growing concentration in the major foreign exchange markets continues. In addition to factors such as those mentioned above that are more specific to foreign exchange markets, this trend is in part symptomatic of broader developments in financial markets. It reflects the same sorts of intense competitive pressures in the financial sector that are encouraging rationalizations and mergers and acquisitions, as previously distinct (if not protected) market segments are breaking down internationally, and as the differences between commercial banks and other financial institutions also continue to erode significantly. While regulatory changes act as a spur to some of these trends, the regulatory changes are themselves in part a response to underlying market trends—including those associated with globalization, innovation, and rapid technological progress—that would simply make their presence felt in other ways in the absence of regulatory changes. A case in point is the trend toward rationalization in the United States, as previously separate regional banking markets integrate further, and as the barriers between investment and commercial banking erode, notwithstanding the failure to date of attempts to repeal the Glass-Steagall Act itself. Similar market pressures have their reflection at the supervisory level, where some countries have moved to more closely integrate domestic supervision and regulation of a range of different financial market sectors, and where coordination has been growing between international supervisory bodies for banking, insurance, and securities markets. At an organizational level, there is some trend also toward closer global integration of major financial institutions’ activities internationally, including in some cases a global centralization of institutional risk management functions.19

The review of exchange market behavior in Part II, Section IX concludes that the key issue for public policy is to improve the environment in which markets work, rather than to intervene directly in market behavior. In particular, while there are aspects of shorterterm dealer behavior that are not immediately driven by an analysis of economic “fundamentals,” the review notes that this does not mean that such behavior is unconstrained by those fundamentals. Indeed, one would expect the payoff to good economic analysis to be largest precisely when the fundamentals are changing significantly. In general, then, extrapolating from short-term dealer behavior may give a misleading impression of (the lack of) “rationality” or “efficiency” in the foreign exchange market. At the level of the dealers specifically, direct interventions in dealer activity that restrict their ability to earn dealing profits will likely have the negative consequences of either reducing the amount of market-making activity (a form of public good) below what is appropriate given the general market environment or result in a relocation of dealing activity to a different jurisdiction to avoid the effect of the regulatory intervention. More often that not, it is not particularly difficult technically to trade a currency in offshore markets, even if indirectly through various sorts of synthetic instruments.20

Foreign Exchange Markets in Developing Countries

The market structures for foreign exchange trading in developing countries have been organized as both interbank markets (as observed in developed and more advanced developing countries) and auction-type arrangements. The choice between market structures in general depends on a range of factors, including the initial institutional and regulatory circumstances.

Auction-type arrangements still exist in a significant number of IMF member countries (Angola, Azerbaijan, Belarus, Bolivia, Congo, Ethiopia, Georgia, Honduras, Kazakhstan, the Kyrgyz Republic, Moldova, Papua New Guinea, Russia, the Slovak Republic, Tajikistan, Turkmenistan, Ukraine, Uzbekistan, Vietnam, and Zambia). Auctions continue to be the dominant market structure in several of these countries while in others the role of the auction market (mostly in the form of fixing sessions) has been significantly reduced concurrent with the development of interbank markets (Part III, Table A12). In the early stages of market development, when the volume of foreign exchange trading is limited, voluntary auctiontype arrangements, such as an interbank auction or fixing arrangements, have offered a useful transitional arrangement toward an interbank exchange market.

Several countries introduced interbank foreign exchange markets only recently (e.g., Azerbaijan, Guinea, Iceland, Madagascar, Mauritius, Morocco, Mozambique, Nigeria, Tunisia, and Vietnam), and in several countries, interbank markets have developed significantly in depth and sophistication with transactions taking place through electronic and computer systems (e.g., Armenia, Bulgaria, China, Chile, Colombia, Costa Rica, Egypt, Estonia, Mexico, Moldova, the Philippines, Poland, Romania, Russia, and the Slovak Republic). While there are significant benefits from continuous interbank market structures because of the relative liquidity and efficiency of such arrangements, the development of such markets can be impeded by various factors. These include the instability and low volume of foreign exchange flows; insolvency problems with commercial banks; lack of trust between market dealers; inefficiencies in banking, payments, and communication systems; and the presence of official regulations and practices and exchange controls. As Part II, Section IX notes, broader international trends have combined in recent years to heighten the interest of international banks in the foreign exchange markets of at least some developing and transitional economies. This interest can provide a significant fillip for the further development of those markets but is in turn predicated on some minimum requirements in terms of size, structure, and efficiency.

Active foreign exchange bureau markets where cash and retail transactions take place continue to exist in many developing countries. In some countries, the importance of exchange bureau markets reflects continued restrictions on interbank market trading and limitations on access to foreign exchange in the official markets. In addition, parallel markets for foreign exchange continue to exist in a number of countries where access to foreign exchange has been restricted. The existence of officially sanctioned parallel markets (e.g., the curb market in Pakistan) can also provide a significant degree of capital account convertibility.

As part of its technical assistance work on exchange systems, the IMF has assisted member countries in developing an efficient interbank foreign exchange market. In creating such a market, the staff emphasized, inter alia, the need to eliminate barriers arising from official regulations and practices, and exchange controls, that might hinder foreign exchange dealings; the establishment of codes of conduct for dealings among market participants and the imposition of information technology to facilitate interbank dealings; the strengthening of payments and clearing arrangements; and the introduction of prudential guidelines and reporting requirements for authorized dealers. Some of the aspects involved in the development of interbank markets are discussed in Part II, Section VIII.

Forward Exchange Markets

As regards forward exchange markets, most industrial countries have eliminated restrictions on access to forward markets and limits on forward exchange transactions. Except for Greece, where the central bank provides forward cover to credit institutions, all industrial countries now have market-determined forward markets. Significant progress has also been made in a number of developing countries in terms of forward market development, with many countries allowing the banking system to provide forward cover for foreign exchange risk (see Part III, Table A12). About half of the developing countries allow forward cover in financial transactions, although official approval for the provision of cover or for conducting transactions in the forward market is still necessary in some cases.

About 70 IMF developing member countries have no formalized markets for forward transactions, and in the remaining developing countries the forward markets are either not very active (e.g., Algeria, Belarus, Jamaica, Lesotho, Madagascar, and Slovenia) or are subject to certain restrictions and limitations, particularly on the type of transactions for which forward operations and cover are permitted. Table 13 presents summary information on the features of the forward markets as well as their relationship with the exchange rate regimes in these countries. As the table indicates, there appears to be a close relationship between the lack of development of forward markets and maintenance of pegged exchange rates: almost all the countries that require prior approval for forward foreign exchange transactions or allow forward cover only for trade transactions and prohibit those for financial transactions have pegged exchange rates; and more than 11 of the 19 countries in which forward cover is provided by the monetary authority maintain formal or informal pegs. This tendency reflects both the fact that the spot markets are not sufficiently developed to support an efficient and smooth functioning forward market and restrictions on forward markets, which are more often associated with pegged exchange arrangements.

Table 13.Summary Features of Forward Exchange Markets in Selected Developing Countries, End of 1997
CountryExchange Rate

Forward Cover/TransactionsOfficial Forward

Prior Approval for

Forward Transactions
Bahamas, TheSCPYesYesNoYes
Burkina FasoSCPNoYesNoYes
ColombiaMF (CB)YesYesNo
Côte d’lvoireSCPNoYesNoYes
CroatiaMF (HB)Yes
CyprusBP (HB)Yes
EcuadorMF (CB)YesYesNo
EgyptMF (DP)YesYesNo
HungaryMF (CB)YesYesNo
LebanonFF (DP)NoYesNo
Macedonia, FYRMF (DP)NoYesYes
Marshall IslandsSCPYesYesNo
PakistanMF (DP)Yes
Papua New GuineaFFNoYesNo
RussiaMF (CB)YesYesYes
Saudi ArabiaLFYesYesNo
Solomon IslandsBPYesYesNo
South AfricaFFYesYesYes
Sri LankaMFNoYesNo
Trinidad and TobagoFFYesYesNo
TunisiaMF (CP)NoYesYes
TurkeyMF (CP)YesYesNo
United Arab EmiratesLFNoYesYes
Notes: BP = basket peg, CB = crawling band; CP = crawling peg; DP = de facto peg; FF = free float; HB = horizontal band; LF = limited flexibility; MF = managed float; SCP = single currency peg.

As discussed more fully in Part II, Section IX, central bank activity in non-spot markets, including through currency futures or derivatives markets, as well as forward contracts, per se, can sometimes be quite problematic. This discussion does not relate to the use of such transactions for foreign reserves management, as distinct from exchange rate intervention. Nor does it relate to forward foreign exchange transactions where these are part of a foreign currency swap (where the two sides of the swap transaction largely eliminate the direct exchange rate effect, but affect domestic bank liquidity). Foreign currency swaps have been used as an important monetary management instrument in a number of countries, Switzerland being perhaps the best known case. The problems with some forms of forward or derivative foreign exchange intervention are that there can be a strong temptation for overuse, as the non-spot interventions economize on the immediate use of official foreign reserves and are frequently less transparent than spot operations. They can disguise significant foreign exchange losses, as has been amply demonstrated by the experiences of Thailand and Korea during the buildup to their currency crises in 1997. In addition, the Brazilian authorities used currency intervention in futures markets as part of their defense of the real during 1997–98.

Multiple Exchange Rate Systems

Some 43 IMF members maintained multiple exchange rate systems at the end of 1997. In 30, the multiple exchange rates gave rise to a multiple currency practice subject to the IMF’s jurisdiction. (See Part III, Table A13 for a more detailed description of these regimes and Box 3 for a discussion of such practices and the IMF’s jurisdiction). Most multiple currency practices arose from the practice of applying different exchange rates to different transactions, from the existence of dual markets with spreads between official and market rates in excess of 2 percent, and from the imposition of various taxes or subsidies or provision of official exchange rate guarantees (see Table 14). In the remaining 13 countries, the multiple exchange rate systems did not give rise to multiple currency practices subject to the IMF’s jurisdiction, as the spread between different effective rates remained within 2 percent or because the different exchange rates were applied to capital transactions (e.g., the Bahamas, Brazil, and Chile).

Table 14.Summary of the Nature of Members’ Multiple Exchange Rate Regimes, End of 1997
Nature of the Multiple Exchange Rate (MER) RegimeNumber of Countries Maintaining the System
Countries with MER that have jurisdictional implications (multiple currency practices)30
Dual markets or different rates for different transactions17
Exchange tax or subsidy4
Exchange guarantee4
At least two of the above5
Countries with MER that do not have jurisdictional implications13
Dual markets or different rates for different transactions5
Exchange tax or subsidy4
Exchange guarantee0
At least two of the above4
Source: Table A13.

Since late 1994, 18 members eliminated multiple currency practices or unified their exchange rate systems or both (Azerbaijan, the Czech Republic, Ethiopia, Georgia, Lao People’s Democratic Republic, Lesotho, Malta, Mauritania, Mauritius, Moldova, Namibia, Nicaragua, Poland, Slovak Republic, South Africa, Tajikistan, Ukraine, and Yemen). Some countries (e.g., Venezuela) unified their exchange rates after temporarily introducing multiple exchange rates or engaging in multiple currency practices during the same period. As already noted, a number of the unifications were associated with a move to floating exchange rate regimes (e.g., Azerbaijan, Lao People’s Democratic Republic, Mauritania, Venezuela, and the Republic of Yemen).

Box 3.Multiple Exchange Rate Systems

According to the IMF’s policy on multiple currency practices (MCPs), “action by a member or its fiscal agencies that of itself gives rise to a spread of more than 2 percent between buying and selling rates for spot exchange transactions between the member’s currency and any other member’s currency would be considered a multiple currency practice” subject to IMF jurisdiction under Article VIII. The approval policies have been flexible and responsive to each country’s circumstances. In approving MCPs, the IMF has taken into consideration whether they are intended to be temporary and introduced or maintained for balance of payments reasons, and whether they discriminate between IMF members. The approval is also contingent on the existence of a clear plan designed to bring about unification over a specific and appropriately brief period of time. The IMF’s policy toward MCPs and country experiences with multiple exchange rates were last reviewed by the Executive Board in April 1984 and February 1985.

Multiple currency practices can take a number of forms:

  • 1. Different rates for different transactions. A typical multiple currency practice may take the form of the setting by the central bank of different exchange rates applying to different categories of transactions (e.g., official, commercial, tourist transactions, and so forth), which result in spreads of more than 2 percent between these rates.

  • 2. Dual or multiple exchange markets. A multiple currency practice would arise if the authorities were to establish separate exchange markets and the coexistence of these markets results in spreads of more than 2 percent between the rates in the different markets. A multiple currency practice would also arise if, having established an official market, the authorities were also to allow the existence of a free market without establishing any mechanism to ensure that spreads of more than 2 percent would not arise between the rates in these markets.

  • 3. Exchange taxes or subsidies. A tax or subsidy payable on exchange transactions or the purchase or sale of a foreign currency may be part of the effective exchange rate on a foreign exchange transaction and thus give rise to a multiple currency practice.

  • 4. Exchange guarantees. Exchange guarantees provided by a member or its fiscal agencies to cover exchange risks are viewed as a provision of subsidized exchange rate by the member. Under many of these systems, compensation for exchange losses is not a part of the nominal exchange rate applied to the purchase of foreign exchange by the beneficiary of the guarantee, but it is a part of the effective exchange rate because the benefit of the coverage flows directly from exchange losses and thus from the exchange transaction in which these losses are realized. A system for covering exchange risk managed by the official authority would not be a multiple currency practice if it is self financed (i.e., the premium paid by the beneficiaries are sufficient to cover the exchange risk).

  • 5. Broken cross rates. Broken cross rates arise from an action by a member or its fiscal agencies that results in midpoint spot exchange rates of other members’ currencies against its own currency in a relationship that differs by more than 1 percent from the midpoint spot exchange rates for these currencies in their principal markets.

Multiple exchange rate regimes have in general been used (1) for balance of payments purposes (e.g., to prevent large exchange rate depreciations from affecting the domestic price of essential commodities, or to prevent sudden one-way pressures on the capital account from affecting trade relations and international reserves); (2) to raise tax revenue through the exchange system; (3) to promote or discourage certain types of transactions or sectors through what effectively amounts to a set of effective taxes and subsidies; and (4) as a temporary measure before liberalizing transactions.

Experience with multiple exchange rate systems has shown that such regimes have more drawbacks than advantages, namely that they can distort economic incentives and impose costs on the economy by misallocating resources for production and consumption. Moreover, the maintenance of such systems generally requires a complex and costly system of controls administered by using public sector resources, and when administrative and institutional systems are weak and scarce resources are employed toward sustaining such systems, foreign exchange pressures may reemerge, manifesting themselves either directly through reserve losses or indirectly through a growing informal sector.

Since the 1950s, there has been a trend away from multiple exchange rates, although progress has not been continuous. Significant progress was made by members in simplifying their exchange systems in the late 1950s and early 1960s, but thereafter progress was mixed, with an increased use of multiple exchange rates in the late 1960s and early 1970s, and again in the early 1980s in response to widespread balance of payments difficulties. The number of countries operating some form of multiple exchange rate regimes somewhat declined from 46 (30 percent of membership) in 1986 to 43 members in 1997 (about 24 percent).

Despite the general tendency toward unification of exchange systems, new multiple currency practices also emerged in a number of countries since 1994. Some of these countries include Cambodia, the Dominican Republic, Honduras, Mongolia, Nigeria, São Tomé and Príncipe, Sudan, Suriname, Uzbekistan, and Zimbabwe (introduced in early 1994 but abolished in July 1997). In some of these cases, multiple currency practices emerged as a result of a significant deviation between the exchange rates prevailing in different exchange markets, which were in turn the result of shortages of foreign exchange, downward pressure on domestic currencies for balance of payments reasons, or absence of a firm commitment to unified exchange rates. In Nigeria, where the official exchange rate is pegged, a dual exchange market was introduced to liberalize the foreign exchange market and enhance the role of market forces as part of an overall strategy for economic liberalization.

Payment Systems Issues

As foreign exchange markets have evolved, international payments have been characterized by a rapid growth in the volume of international transactions, as well as the application of new technologies. Electronically based technologies have become prevalent, reducing transactions costs and payment lags and improving the security and reliability of individual transactions. The number and volume of international payment transactions have increased pressures to upgrade payment systems and harmonize procedures for specific categories of payment transactions.

Central banks, and other supervisory agencies, have been focusing on means to improve risk management in international payment and settlement systems since the early 1980s.21 There have been numerous studies of the risks associated with cross-border transactions, particularly by the Bank for International Settlements and its Committee on Payment and Settlement Systems (CPSS). In 1994, recognizing the importance of foreign exchange settlement risk, the CPSS formed a Steering Group on Settlement Risk in Foreign Exchange Transactions. The group suggested a strategy for controlling foreign exchange settlement risks that relies on collective as well as individual action by private banks. The collective initiatives taken by private banks (particularly the so called “Group of 20 banks”) are seeing the development of new settlement facilities known as continuous linked settlement (CLS). This involves the creation of a limited-purpose bank, the CLS bank, to offer users multicurrency accounts. Settlement risk will be eliminated by the CLS bank settling the different legs of foreign exchange transactions simultaneously. Two existing foreign exchange clearing houses operating multilateral netting services, the Exchange Clearing House (ECHO) and Multinet International Bank, merged with CLS Services, and the new entity will thus provide both continuous linked settlement services and netting services. The central banks of the Group of Ten countries have been actively cooperating with the private sector to facilitate these developments and are closely monitoring the impacts on risks in foreign exchange markets and whether there is need for any further action to mitigate these.

Under the existing classification scheme, members of the ERM have been classified as maintaining regimes of limited flexibility within cooperative arrangements since 1979, although the ERM has the features of a formal pegged exchange rate system that involves pegging the member currencies vis-à-vis the ECU within a given fluctuation margin. Widening of the ERM bands from ±2.25 percent to ±15 percent in August 1994 did not affect the classification of the member countries’ arrangements, although it implied a de facto move toward greater flexibility.

The last major review of the classification scheme was undertaken in 1982. The main changes involved introducing new categories for the nonpegged regimes to distinguish between managed floats and independent floats. The variability of exchange rates visa-vis a number of putative pegs was also reviewed, to identify “backward looking de facto pegs.” This methodology was subsequently applied on a few occasions to help identify the de facto nature of regimes.

Countries may still exercise some degree of flexibility under such arrange men ts, however, depending on how strict the rules of the boards are established (see Baliño, and others, 1997).

If the IMF classification system were to be used, which would include the countries with informal pegs among the flexible exchange rate group, one could reach the conclusion that countries with dual exchange rate systems tend to pursue more flexible exchange rates.

For further discussion on issues related to those mentioned in the last two paragraphs, see IMF (1997b).

A recent hedge fund study reached similar conclusions about the need for caution in considering direct intervention in the market activities of institutional investors and collective investment schemes. See Eichengreen, Masson, and others (1998).

A number of the IMF’s International Capital Markets reports (especially the September 1996 report) have discussed such issues.

    Other Resources Citing This Publication