CHAPTER 4 Setting Par Values

International Monetary Fund
Published Date:
February 1996
  • ShareShare
Show Summary Details
Margaret G. de Vries

As the par value system has evolved, many questions have confronted the Fund. Since par values are central to the Fund Agreement, one set of questions has concerned their establishment. How quickly should initial par values be agreed? What should be the attitude of the Fund toward delays by members in proposing parities? How much pressure should the Fund put on countries to set parities? What exceptions might be made to the requirement of the Articles that an initial par value must be set before a country may draw on the Fund’s resources? What considerations ought to govern appraisal of a proposed parity? Once initial par values are no longer used for any transactions, when should a new effective par value be instituted? Can a country be regarded as having a fully convertible currency—that is, subject to the requirements of Article VIII—if it does not have a par value?

Still other questions have arisen in connection with the adjustment of par values and exchange rates. By what criteria should the need for a change in an exchange rate be determined? Inasmuch as the Articles provide that devaluations shall be proposed only by members, what informal role might the Fund play in suggesting to members the advisability of changes in exchange rates? What constitutes fundamental disequilibrium and what competitive exchange depreciation, the two concepts of the Articles relating to changes in exchange rates?

The ways in which the Fund has attempted to find answers to these questions make up, in part, the policy of the Fund on exchange rates. A few of the issues involved have been covered by decisions of general applicability taken by the Executive Board. But for most of these issues, no broadly stated policy lines have been formulated. Rather, as is customary in most fields, policy is mainly an accretion of the attitudes and actions taken in particular situations over the years.

This chapter and the next three provide some generalizations of the Fund’s policies on par values and exchange rates. These generalizations are based on an examination of the decisions taken by the Board and the views expressed by several Directors in a variety of individual country situations; in a number of instances the reasoning of the staff, reflected in the papers prepared for the discussions in the Board, has also been included. The present chapter covers those policies which have applied to the setting of par values. It describes considerations which have guided the Fund in timing the introduction of par values, the criteria used for approving the parities proposed, and the views of the Fund as to whether a par value must first be agreed if a member is to draw on the Fund’s resources.

Several countries, after they had established initial par values, took measures which meant that these par values were no longer applied to any transactions. The Fund, therefore, has also had to consider the extent to which it should urge countries to institute effective par values, and whether countries without effective par values may assume the convertibility status of Article VIII. Moreover, once par values have been established, certain requirements follow for cross rates with third currencies, for the margins within which exchange rates may move around parities, and for rates in forward exchange markets. These matters are also discussed in this chapter. Policies relating to the adjustment of par values and of exchange rates are deferred until Chapter 5.

Tracing the growth of the Fund’s exchange rate policies over the years reveals the flexibility and adaptability with which the Fund has administered the par value system. Yet adaptability and flexibility have not meant the absence of basic policies. Some general guidelines can be discerned, demonstrating that the policies have not been merely a series of ad hoc reactions to particular situations. Nonetheless, since generally formulated policies cannot always be applied to the particulars of individual situations, some anomalies and inconsistencies in the application of policy have inevitably occurred.


Immediate parities in early years

In the first years the Board was anxious that members should establish par values as soon as possible. This was manifested when the initial par values for original members were agreed at the end of 1946. Although both the Board and member countries recognized the many uncertainties then attending exchange rate determination, initial par values were quickly set for most original members.

Weighty arguments could have been marshaled against setting initial par values in 1946. Many members had been devastated by the war and had to undergo extensive reconstruction. These countries, and others, were in the grip of serious inflation. Shortages of goods and services were widespread and international trade was badly distorted.

However, the Board wished to go ahead. In the first place, Article XX, Section 4, provided that initial par values for members having at least 65 per cent of the total of quotas must be established before exchange transactions could begin, and that (except for countries that had been occupied by the enemy) a par value must be agreed with the Fund before a member could use the Fund’s resources. Moreover, the Board recognized the desire of the drafters of the Articles to put the par value regime into operation at an early stage. It had been foreseen at Bretton Woods that the Fund would have to begin in a period of disorder, but the conferees were convinced that a start toward an improved world exchange rate structure could be made.

On September 12, 1946, therefore, the Fund called upon its 39 members to communicate par values, and on December 18, 1946 the initial par values of 32 countries were announced, viz., Belgium, Bolivia, Canada, Chile, Colombia, Costa Rica, Cuba, Czechoslovakia, Denmark, Ecuador, Egypt, El Salvador, Ethiopia, France, Guatemala, Honduras, Iceland, India, Iran, Iraq, Luxembourg, Mexico, the Netherlands, Nicaragua, Norway, Panama, Paraguay, Peru, the Philippines, the Union of South Africa, the United Kingdom, and the United States.

Initial par values were also agreed with Belgium for the Belgian Congo; with France for Algeria, Cameroun, French Antilles, French Equatorial Africa, French Guiana, French possessions in India, French possessions in Oceania, French Somaliland, French West Africa, Madagascar and dependencies, Morocco, New Caledonia, New Hebrides, Reunion, St. Pierre and Miquelon, Togoland, and Tunisia; with the Netherlands for Curacao and Surinam; and with the United Kingdom for Bahamas, Barbados, Bermuda, British Guiana, British Honduras, Burma, Ceylon, Cyprus, Falkland Islands, Fiji, Gambia, Gibraltar, Gold Coast, Hong Kong, Jamaica, Kenya, Malaya (Singapore and Malayan Union), Malta, Mauritius, Nigeria, Northern Rhodesia, Nyasaland, Palestine, Sarawak (British North Borneo), Seychelles, Sierra Leone, Southern Rhodesia, Tanganyika, Tonga, Trinidad, Uganda, and Zanzibar.

The determination of par values for six members (Brazil, China, the Dominican Republic, Greece, Poland, and Yugoslavia) was deferred at their request because their domestic monetary situation was not sufficiently stable to warrant the adoption of par values. The Uruguayan submission of a par value was not definite enough to permit its formal acceptance. The determination of initial par values for the French dependent territories in Indo-China and for the Netherlands Indies was also deferred.

Since imports were subject to controls by nearly all countries, the main criterion for judging the rates proposed was their effect on exports. So long as exports continued to flow, it could fairly be assumed that the prevailing exchange rate would help to attain by the end of the transitional period a “tolerable balance of payments.”1 Existing exchange rates did not seem to be impairing the ability of any country to export. Supply shortages were so acute that, despite the current exchange rates and despite the import controls, there was no difficulty in disposing of any exports that a country was in a position to offer. Difficulties of production and transport and other prevalent obstacles to a smooth supply of goods were by far the most serious impediments to an expansion of exports. The Executive Directors were also aware that to encourage a general revision of exchange rates would have been quite unacceptable to member countries. Such a revision “would have gravely dislocated exchange relationships which were already working in practice and enabling trade to be carried on.”2

For members subsequently joining the Fund, a procedure for the establishment of an initial par value is prescribed which gives the Fund the right to initiate the process. The standard membership resolution provides (paragraph 5) that the Fund is to request the member to communicate, within thirty days of receipt of the Fund’s request, a proposed par value for its currency. Within sixty days of receipt of the proposed par value by the Fund, the Fund and the member must agree on an initial par value. In practice, the period of thirty days is not introduced until the member is known to be ready to establish a par value, and the sixty-day period is extended if this is found to be necessary. Should no agreement be reached at the end of the extended period, the member will be deemed to have withdrawn from the Fund.

In the late 1940’s and early 1950’s, the Fund made persistent efforts to get original members to agree initial par values if they had not already done so. Brazil and the Dominican Republic, having postponed theirs in 1946, set them in 1948; Yugoslavia’s par value was agreed in May 1949. Most countries that joined the Fund in these years very quickly agreed initial par values. These included Australia, Lebanon, Syria, Turkey, and Venezuela in 1947, Sweden and Pakistan in 1951, Ceylon in 1952, Burma, Germany, Japan, and Jordan in 1953, and Haiti in 1954. Table 1 at the end of this chapter lists the initial par values agreed with the Fund for each year from 1946 through 1965.

Table 1.Initial Par Values Established, 1946–65
Date 1Member 2 and Rate
Dec. 18Belgium—2.28167 U.S. cents per franc
Dec. 18Bolivia—2.38095 U.S. cents per boliviano
Dec. 18Canada—100.000 U.S. cents per dollar
Dec. 18Chile—3.22581 U.S. cents per peso
Dec. 18Colombia—57.1433 U.S. cents per peso
Dec. 18Costa Rica—17.8094 U.S. cents per colón
Dec. 18Cuba—100.000 U.S. cents per peso
Dec. 18Czechoslovakia—2.00000 U.S. cents per koruna
Dec. 18Denmark—20.8376 U.S. cents per krone
Dec. 18Ecuador—7.40741 U.S. cents per sucre
Dec. 18Egypt—413.300 U.S. cents per pound
Dec. 18El Salvador—40.0000 U.S. cents per colón
Dec. 18Ethiopia—40.2500 U.S. cents per dollar
Dec. 18France—0.839583 U.S. cent per franc
Dec. 18Guatemala—100.000 U.S. cents per quetzal
Dec. 18Honduras—50.0000 U.S. cents per lempira
Dec. 18Iceland—15.4111 U.S. cents per króna
Dec. 18India—30.2250 U.S. cents per rupee
Dec. 18Iran—3.10078 U.S. cents per rial
Dec. 18Iraq—403.000 U.S. cents per dinar
Dec. 18Luxembourg—2.28167 U.S. cents per franc
Dec. 18Mexico—20.5973 U.S. cents per peso
Dec. 18Netherlands—37.6953 U.S. cents per guilder
Dec. 18Nicaragua—20.0000 U.S. cents per córdoba
Dec. 18Norway—20.1500 U.S. cents per krone
Dec. 18Panama—100.000 U.S. cents per balboa
Dec. 18Paraguay—32.3625 U.S. cents per guaraní
Dec. 18Peru—15.3846 U.S. cents per sol
Dec. 18Philippines—50.0000 U.S. cents per peso
Dec. 18South Africa—403.000 U.S. cents per pound
Dec. 18United Kingdom—403.000 U.S. cents per pound
Dec. 18United States—100.000 U.S. cents per dollar
Apr. 18Venezuela—29.8507 U.S. cents per bolívar
June 19Turkey—35.7143 U.S. cents per lira
July 29Lebanon—45.6313 U.S. cents per pound
July 29Syrian Arab Republic—45.6313 U.S. cents per pound
Nov. 17Australia—322.400 U.S. cents per pound
Apr. 23Dominican Republic—100.000 U.S. cents per peso
July 14Brazil—5.40541 U.S. cents per cruzeiro
May 24Yugoslavia—2.00000 U.S. cents per dinar
Mar. 19Pakistan—30.2250 U.S. cents per rupee
July 1Finland—0.434783 U.S. cent per markka
Nov. 5Sweden—19.3304 U.S. cents per krona
Jan. 16Ceylon—21.0000 U.S. cents per rupee
Jan. 30Germany—23.8095 U.S. cents per deutsche mark
May 4Austria—3.84615 U.S. cents per schilling
May 11Japan—0.277778 U.S. cent per yen
Aug. 7Burma—21.0000 U.S. cents per kyat
Oct. 2Jordan—280.000 U.S. cents per dinar
Apr. 9Haiti—20.0000 U.S. cents per gourde
Jan. 9Argentina—5.55556 U.S. cents per peso
Mar. 13Israel—55.5556 U.S. cents per pound
May 14Ireland—280.000 U.S. cents per pound
July 23Sudan—287.156 U.S. cents per pound
Nov. 5Ghana—280.000 U.S. cents per pound
July 17Spain—1.66667 U.S. cents per peseta
Aug. 12Libya—280.000 U.S. cents per pound
Oct. 16Morocco—19.7609 U.S. cents per dirham
Jan. 8Saudi Arabia—22.2222 U.S. cents per riyal
Mar. 30Italy—0.160000 U.S. cent per lira
Oct. 7Uruguay—13.5135 U.S. cents per peso
Mar. 29Greece—3.33333 U.S. cents per drachma
Oct. 27New Zealand—278.090 U.S. cents per pound
June 1Portugal—3.47826 U.S. cents per escudo
July 20Malaya—32.6667 U.S. cents per dollar
July 25Cyprus—280.000 U.S. cents per pound
Mar. 8Jamaica—280.000 U.S. cents per pound
Mar. 13Liberia—100.000 U.S. cents per dollar
Mar. 22Afghanistan—2.22222 U.S. cents per afghani
Apr. 17Nigeria—280.000 U.S. cents per pound
Apr. 26Kuwait—280.000 U.S. cents per dinar
June 14Somalia—14.0000 U.S. cents per shilling
Oct. 20Thailand—4.80769 U.S. cents per baht
Sept. 28Tunisia—190.476 U.S. cents per dinar
Jan. 26Burundi—1.14286 U.S. cents per franc
Feb. 10Trinidad and Tobago—58.3333 U.S. cents per dollar
Aug. 6Sierra Leone—140.000 U.S. cents per leone

The date given is that on which the initial par value became effective, usually a few days later than that on which the par value was agreed with the Fund.

Excludes nonmetropolitan territories.

The date given is that on which the initial par value became effective, usually a few days later than that on which the par value was agreed with the Fund.

Excludes nonmetropolitan territories.

Delayed parities

There were, of course, some instances of initial parities being delayed. Two illustrations of relatively short delays are provided by Finland and Austria. Both joined the Fund in 1948 but Finland did not set its first parity until 1951 and Austria not until 1953. Finland agreed as an initial par value the exchange rate which had been in effect since the devaluations of September 1949. The establishment of Austria’s parity, however, was dependent on further drastic devaluation. The rate adopted for the schilling after the war was S 10 = $1. In November 1949 Austria introduced multiple rates; these were replaced by a single rate in May 1953, at which time an initial par value of S 26 = $1 was established.

As to long delays in establishing initial parities, Italy is a case in point, and also serves to illustrate the caution exercised both by members in communicating par values and by the Fund in not applying strong pressure to do so. Italy joined the Fund in March 1947. The Membership Resolution, drafted soon after the Fund came into being, contained terms similar to those in the Articles of Agreement. In effect, Italy was declared a country whose metropolitan territory had been occupied by the enemy.3 It was specifically granted the privileges of Article XX, Section 4 (d) (ii), which provided that an enemy-occupied country could, with the permission of the Executive Board, draw on the Fund’s resources prior to establishing an initial par value.

On April 24, 1947, in accordance with its usual procedure, the Board requested the Government of Italy to communicate within thirty days a par value for the lira based on the exchange rate prevailing on March 27, 1947, the date on which Italy joined the Fund. The Italian economy was then encountering several difficulties. Prices were rising rapidly, owing to inflation, and the balance of payments was in marked disequilibrium. Production was stagnating and the rate of unemployment was high. Multiple currency practices existed: there was an official rate of Lit 225 = $1 and a free rate of Lit 585–600 = $1. A rate of Lit 410 = $1 seemed to the Italian Government at the time as the only rate which could reasonably be proposed as a par value. But this rate could not be considered definitive. Moreover, the Fund’s request was for a rate based on the rate on March 27, 1947, which was presumably already outdated.

Therefore, on May 23, 1947 the Government requested an extension of the usual period of ninety days for agreeing on a par value with the Fund, and on May 28 the Board informed the Government that it had extended the period indefinitely and was prepared to discuss with a representative of the Government a new date for communication of a par value.

During the following three years Italy made great progress in the achievement of economic stability, as a result of the adoption in November 1947 of a comprehensive anti-inflationary program and of financial assistance from the United States. By mid-1950 a reasonable degree of monetary stability had been attained, production had expanded beyond the prewar peak, and the increase in exports had helped to reduce substantially the balance of payments deficit. In 1949 the exchange rate for the lira, only theoretically a fluctuating rate, had been devalued from Lit 575 = $1 to about Lit 625 = $1. This rate applied to all transactions in the official market, there was no currency discrimination, and the rates of various other currencies were pegged to the lira-dollar rate.

During the 1950’s, consequently, there were frequent contacts between the Italian Government and the Fund to explore the possibility of agreement on a par value. The Board’s belief was that the setting, and defense, of a par value would help to consolidate the strength of the lira both internally and externally. On all these occasions the Government showed extreme caution and suggested that continuing uncertainties, such as the process of trade liberalization under the OEEC and the establishment of a new tariff system, counseled delay. Between September 1949 and the establishment of external convertibility for the lira on December 29, 1958, Italy maintained a unitary exchange rate of between Lit 624.60 and Lit 625.10 = $1. The Italian authorities, nonetheless, argued for the need to maintain a rate for the lira which was, at least potentially, a fluctuating one.

In September 1959 the Government asked the Fund to agree to a rate of Lit 625 = $1 for determination of the amount of Italy’s currency subscription in connection with the increase in its quota. The Board took the occasion to suggest again that a par value be instituted. A few months later, on March 15, 1960, Italy proposed an initial par value of Lit 625 = $1. Foreign trade had been liberalized. Reserves had increased fivefold since 1950. The national product was growing at an average annual rate of more than 5 per cent. It was difficult, if not impossible, to justify any further delay in setting a par value. The Board agreed with the communicated par value, expressing satisfaction that the rate prevailing for many years had at last become the par value.

The instance of Italy is by no means unique. Greece, Thailand, and Uruguay, among others, joined the Fund early on but set their par values only many years thereafter. Table 2 at the end of this chapter gives the date of membership and the date that the initial par value became effective for all countries that were Fund members in the years 1945–65.

Table 2.Dates of Membership and of Initial Par Values, December 27, 1945-December 31, 1965 1
Member 2Date of

Effective Date

of Initial

Par Value 3
1. AfghanistanJuly 14, 1955Mar. 22, 1963
2. AlgeriaSept. 26, 1963Not established
3. ArgentinaSept. 20, 1956Jan. 9, 1957
4. AustraliaAug. 5, 1947Nov. 17, 1947
5. AustriaAug. 27, 1948May 4, 1953
6. BelgiumDec. 27, 1945Dec. 18, 1946
7. BoliviaDec. 27, 1945Dec. 18, 1946
8. BrazilJan. 14, 1946July 14, 1948
9. BurmaJan. 3, 1952Aug. 7, 1953
10. BurundiSept. 28, 1963Jan. 26, 1965
11. CameroonJuly 10, 1963Not established
12. CanadaDec. 27, 1945Dec. 18, 1946
13. Central African RepublicJuly 10, 1963Not established
14. CeylonAug. 29, 1950Jan. 16, 1952
15. ChadJuly 10, 1963Not established
16. ChileDec. 31, 1945Dec. 18, 1946
17. ChinaDec. 27, 1945Not established
18. ColombiaDec. 27, 1945Dec. 18, 1946
19. Congo (Brazzaville)July 10, 1963Not established
20. Congo, Democratic RepublicSept. 28, 1963Not established
21. Costa RicaJan. 8, 1946Dec. 18, 1946
22. CubaMar. 14, 1946Dec. 18, 1946
23. CyprusDec. 21, 1961July 25, 1962
24. CzechoslovakiaDec. 27, 1945Dec. 18, 1946
25. DahomeyJuly 10, 1963Not established
26. DenmarkMar. 30, 1946Dec. 18, 1946
27. Dominican RepublicDec. 28, 1945Apr. 23, 1948
28. EcuadorDec. 27, 1945Dec. 18, 1946
29. El SalvadorMar. 14, 1946Dec. 18, 1946
30. EthiopiaDec. 27, 1945Dec. 18, 1946
31. FinlandJan. 14, 1948July 1, 1951
32. FranceDec. 27, 1945Dec. 18, 1946
33. GabonSept. 10, 1963Not established
34. GermanyAug. 14, 1952Jan. 30, 1953
35. GhanaSept. 20, 1957Nov. 5, 1958
36. GreeceDec. 27, 1945Mar. 29, 1961
37. GuatemalaDec. 27, 1945Dec. 18, 1946
38. GuineaSept. 28, 1963Not established
39. HaitiSept. 8, 1953Apr. 9, 1954
40. HondurasDec. 27, 1945Dec. 18, 1946
41. IcelandDec. 27, 1945Dec. 18, 1946
42. IndiaDec. 27, 1945Dec. 18, 1946
43. IndonesiaApr. 15, 1954Not established
44. IranDec. 29, 1945Dec. 18, 1946
45. IraqDec. 27, 1945Dec. 18, 1946
46. IrelandAug. 8, 1957May 14, 1958
47. IsraelJuly 12, 1954Mar. 13, 1957
48. ItalyMar. 27, 1947Mar. 30, 1960
49. Ivory CoastMar. 11, 1963Not established
50. JamaicaFeb. 21, 1963Mar. 8, 1963
51. JapanAug. 13, 1952May 11, 1953
52. JordanAug. 29, 1952Oct. 2, 1953
53. KenyaFeb. 3, 1964Not established
54. KoreaAug. 26, 1955Not established
55. KuwaitSept. 13, 1962Apr. 26, 1963
56. LaosJuly 5, 1961Not established
57. LebanonApr. 14, 1947July 29, 1947
58. LiberiaMar. 28, 1962Mar. 13, 1963
59. LibyaSept. 17, 1958Aug. 12, 1959
60. LuxembourgDec. 27, 1945Dec. 18, 1946
61. Malagasy RepublicSept. 25, 1963Not established
62. MalawiJuly 19, 1965Not established
63. MalaysiaMar. 7, 1958July 20, 1962
64. MaliSept. 27, 1963Not established
65. MauritaniaSept. 10, 1963Not established
66. MexicoDec. 31, 1945Dec. 18, 1946
67. MoroccoApr. 25, 1958Oct. 16, 1959
68. NepalSept. 6, 1961Not established
69. NetherlandsDec. 27, 1945Dec. 18, 1946
70. New ZealandAug. 31, 1961Oct. 27, 1961
71. NicaraguaMar. 14, 1946Dec. 18, 1946
72. NigerApr. 24, 1963Not established
73. NigeriaMar. 30, 1961Apr. 17, 1963
74. NorwayDec. 27, 1945Dec. 18, 1946
75. PakistanJuly 11, 1950Mar. 19, 1951
76. PanamaMar. 14, 1946Dec. 18, 1946
77. ParaguayDec. 27, 1945Dec. 18, 1946
78. PeruDec. 31, 1945Dec. 18, 1946
79. PhilippinesDec. 27, 1945Dec. 18, 1946
80. PolandDec. 27, 1945Not established
81. PortugalMar. 29, 1961June 1, 1962
82. RwandaSept. 30, 1963Not established
83. Saudi ArabiaAug. 26, 1957Jan. 8, 1960
84. SenegalAug. 31, 1962Not established
85. Sierra LeoneSept. 10, 1962Aug. 6, 1965
86. SomaliaAug. 31, 1962June 14, 1963
87. South AfricaDec. 27, 1945Dec. 18, 1946
88. SpainSept. 15, 1958July 17, 1959
89. SudanSept. 5, 1957July 23, 1958
90. SwedenAug. 31, 1951Nov. 5, 1951
91. Syrian Arab RepublicApr. 10, 1947July 29, 1947
92. TanzaniaSept. 10, 1962Not established
93. ThailandMay 3, 1949Oct. 20, 1963
94. TogoAug. 1, 1962Not established
95. Trinidad and TobagoSept. 16, 1963Feb. 10, 1965
96. TunisiaApr. 14, 1958Sept. 28, 1964
97. TurkeyMar. 11, 1947June 19, 1947
98. UgandaSept. 27, 1963Not established
99. United Arab RepublicDec. 27, 1945Dec. 18, 1946
100. United KingdomDec. 27, 1945Dec. 18, 1946
101. United StatesDec. 27, 1945Dec. 18, 1946
102. Upper VoltaMay 2, 1963Not established
103. UruguayDec. 27, 1945Oct. 7, 1960
104. VenezuelaDec. 30, 1946Apr. 18, 1947
105. Viet-NamSept. 21, 1956Not established
106. YugoslaviaDec. 27, 1945May 24, 1949
107. ZambiaSept. 23, 1965Not established

As 4 of the 107 countries listed in this table withdrew from membership (Cuba, Czechoslovakia, Indonesia, and Poland), the actual membership on December 31, 1965 was 103.

Excludes nonmetropolitan territories.

The date given is that on which the par value became effective, usually a few days later than that on which the par value was agreed with the Fund.

As 4 of the 107 countries listed in this table withdrew from membership (Cuba, Czechoslovakia, Indonesia, and Poland), the actual membership on December 31, 1965 was 103.

Excludes nonmetropolitan territories.

The date given is that on which the par value became effective, usually a few days later than that on which the par value was agreed with the Fund.

A variety of circumstances help to explain delays in establishing par values: chronic inflation, a multiplicity of rates, the need to experiment with devaluation. The continuous ebb and flow of opinion about fluctuating rates no doubt also contributed, at least in part. Countries that already had fluctuating rates, albeit nominally, were reluctant to give them up. Their hesitation did not stem from the fact that they gained any advantages from their fluctuating rates; in most instances rate fluctuations were smoothed out by official intervention in the exchange market. Rather, the authorities concerned believed that it would be easier to change a rate if it had not been instituted as a par value. This consideration became less important as the Fund established its authority over changes in all exchange rates.

What motivated Fund concurrence in these delays was not any weakened faith in the system of par values. In part the Board concurred because of the realization as time went on that it was preferable to have parities set at realistic and maintainable levels rather than to have them established prematurely. By the late 1940’s it had become apparent that par values would not be adjusted frequently; the appropriateness, therefore, of a particular parity was of greater significance than its timing. In part, too, the Fund did not wish to jeopardize the improving Fund-member relations in the exchange rate field. As Chapter 5 indicates, the Fund had managed to bring under its jurisdiction virtually all changes in exchange rates, whether par values or not, and the Fund’s views were frequently being sought. In this environment, to attach excessive significance to the formal status of a par value at the expense of workable, even affable, member relations might be foolhardy.

Short-lived parities

Despite the pragmatism of the Fund’s policy on the establishment of initial parities, a few par values were set which quickly proved untenable. Because countries changed the multiple rates that coexisted with their par values, their par values were within a short time not meaningful. The examples of Argentina and Uruguay best illustrate this phenomenon.

The Fund agreed in January 1957 to an initial par value for Argentina. An exchange rate of M$N 18 = US$1 had been introduced in October 1955 in connection with a devaluation and as part of a major effort to liberalize and stabilize the economy. The Argentine authorities now proposed this rate as a par value. At the time, this exchange rate applied to the bulk of exports and imports (approximately 85 to 90 per cent of total trade transactions), although retention taxes applied to most exports and a free market also existed.

The staff found it difficult to make any analysis of the appropriateness of the proposed par value based on a comparison of prices, and believed that the rate of M$N 18 = US$1 had not been in effect long enough to judge whether it was an equilibrium rate. Nonetheless, the staff recommended to the Board that it should agree to the proposed par value. This recommendation was based on the expectation of the Argentine authorities that the rate would prove to be appropriate in the light of their domestic financial policies and the anticipated improvement of exports. Relying on this expectation, the Board accepted the rate, but with some misgivings.

It soon became doubtful whether the par value could be maintained. Already by May 1957, changes in the effective exchange rates for imports were being made by transfers of certain commodities to the free market, and throughout 1958 the exchange rates for exports were frequently adjusted. On January 12, 1959, the Argentine authorities introduced a single, free, exchange market for all transactions; at the outset the rate in this free market was M$N 66.60 = US$1. The par value of M$N 18 = US$1 thereafter applied only to the proceeds of exports effected before December 30, 1958, and to payments for imports contracted for before that date.4 No new par value had been suggested by Argentina down to the end of 1965, and no transactions were then taking place at the par value agreed in 1957.

The Uruguayan par value agreed in 1960 provides another example of a short-lived initial parity. In connection with a stabilization plan supported by a stand-by arrangement, the Fund in October 1960 agreed to an initial par value for the peso of Ur$7.40 = US$1. This rate was the effective one for exports of greasy wool; it resulted from the application of a retention tax of approximately 25 per cent to the sale of exchange receipts for this product at the free market rate (at that time approximately Ur$11.42 = US$1). The Uruguayan authorities believed that this rate was adequate and viable for the wool exports to which it was applicable (wool accounted for somewhat more than half of the country’s total exports). Hence, although the staff had been pressing for the adoption of the free market rate as the basis for an initial par value, it recommended that the Fund should concur in the rate proposed by Uruguay. The Board agreed. However, by January 1, 1961—only a few months later—all exchange transactions were moved to a free market with a fluctuating rate, where the prevailing rate was about Ur$11.00 = US$1.5

Criteria for accepting proposed par values

According to the Articles of Agreement (Article XX, Section 4 (b)), the criterion for Fund concurrence in a proposal for an initial parity is a negative one: the par value must not in the Fund’s opinion cause “recourse to the Fund on the part of that member or others on a scale prejudicial to the Fund and to members.” In practice, the prospects for a country’s exports have been a crucial factor in the consideration of a proposed parity. As noted above, in the selection of initial parities for original members in 1946, considerable emphasis was given to the effect on exports; 6 and for the next several years this consideration continued to be given the most weight. “So long as an exchange rate does not hamper a country’s exports, there is little to be said in present world conditions for altering it.” 7 The Fund has continued to rely heavily on the effect-on-exports test. However, once world markets again became competitive, in the 1950’s, the Fund was more inclined to accept a par value that differed from the existing exchange rate.

Since the middle of the 1950’s, members have generally not proposed par values unless they were fairly certain that the rate suggested would be acceptable to the Fund and that they could maintain that rate. The nature of the circumstances prevailing when countries have proposed parities suggests some of the criteria the Fund has developed for accepting them. One such circumstance is that exchange rates proposed as par values have commonly been in effect for a number of years. From 1957 to the end of 1965, twenty-seven countries proposed initial par values shortly after they accepted membership. Most of these countries, especially those in the sterling area, selected as initial parities the exchange rates that had been in effect since the devaluations of September 1949. Both more and less developed countries were among those that proposed as par values exchange rates that had already been in effect for some eight to fifteen years: Cyprus, Ghana, Ireland, Jamaica, Kuwait, Liberia, Libya, New Zealand, Nigeria, Portugal, Sierra Leone, Somalia, and the Sudan.

Because it was known that the Fund wished to approve as par values only exchange rates which were clearly stable, only countries which had such rates offered them as par values. Indeed, assuming that the rate proposed is appropriate for a member’s exports, stability has often been the principal factor noted in the Board’s discussions of proposed par values. This is, of course, in accordance with the Fund’s emphasis on the achievement of general economic stability. In addition, a stable exchange rate usually has had a close link with other currencies. Thus, for example, when the Fund considered an initial par value for New Zealand, the long-standing relation of the New Zealand pound to the pound sterling was recognized. Another reason why the Fund has used the test of stability for judging a proposed par value is that a stable rate provides good evidence that the authorities can maintain the par value proposed. Still a further reason for adopting the criterion of stability has been that where an exchange rate has not been stable, the correct level for a par value has been extremely hard to gauge. The difficulties of finding an appropriate rate, in the absence of an already stable rate, was evident, for example, in the Canadian experience of the early 1960’s. After the authorities had decided to give up the fluctuating rate they still delayed the institution of an effective par value, partly because of the difficulty of choosing an exchange rate satisfactory for both current and capital account. (The details of Canada’s fluctuating rate are discussed in Chapter 7.)

In addition to stability, the Fund has, since the mid-1950’s, considered the uniformity and general applicability of the exchange rate suggested as a par value. When a member proposes a parity, exchange reform has usually already been completed. Few, if any, transactions continue to be carried out at exchange rates other than the proposed one. In this sense the par value must be an effective one. Austria, Saudi Arabia, and Spain, for example, all introduced par values only after exchange reform had been made fully effective. This policy, emerging after 1955, was in contrast to that followed earlier when the Board had been eager that members should establish par values even if multiple rates coexisted with these par values.

A final criterion of the suitability of a proposed par value is that the country concerned should have a strong balance of payments and reserve position, and few restrictions. Malaysia and Nigeria, for example, had few restrictions and the former had been contemplating coming under Article VIII. Preferably, the country should be able to relax any prevailing restrictions. Thus, simultaneously with their new parities, Jamaica and Kuwait accepted the obligations of Article VIII.


Although generally the initial par values of member countries have been agreed with little controversy, some problems have attended the establishment of a few initial parities. One such instance involved a dispute in which one country believed itself to be adversely affected not by the devaluation of another country’s currency but by that other country’s abstaining from devaluation. Such a situation is, in effect, the opposite of competitive devaluation. Difficulties have also arisen in connection with the establishment of, or alteration of, par values for totally planned economies. One such instance was the establishment of a par value by Yugoslavia in 1949; another involved a change in par value by Czechoslovakia in 1953.

A controversial parity—Pakistan

A dispute about an exchange rate arose when the Fund deliberated on an initial par value for the Pakistan rupee, after Pakistan had become a member early in 1950. In September 1949 Pakistan had elected not to follow the United Kingdom and the rest of the sterling area in devaluation, and maintained the rate for the Pakistan rupee at PRs 1 = $0.30225. This was the same as the rate for the Indian rupee before the 1949 devaluation, and Pakistan proposed this rate to the Fund in 1950 as an initial par value. India’s new rate was Rs 1 = $0.21.

India refused to trade with Pakistan at its current rate of exchange and Pakistan was adamant that it would not devalue. Consequently, trade between the two countries—which had previously been extensive and essential to both—had come to a virtual halt. The balance of payments of both countries was being seriously harmed. Pakistan’s exports at the time consisted almost entirely of raw jute, most of which went to India for the manufacture of burlap. Inasmuch as India produced most of the world’s burlap, Pakistan had no other outlet for its jute. And India had no other source of supply.

Pakistan’s position was that, on the basis of price comparisons, exchange devaluation would not help to expand raw material exports; nor would it aid in diversifying exports, whether in composition or trading pattern, as industrialization progressed. Devaluation would also turn the terms of trade against imports—which were mainly consumer goods—and especially textiles from India. India’s position was that Pakistan’s exchange rate made it impossible for India’s burlap producers to operate profitably.

A special staff working party was set up in the Fund to examine the economic merits of Pakistan’s proposed par value. In September 1950 the Executive Board, at a special meeting in Paris during the Fifth Annual Meeting, discussed the problem. Not wishing to add to the contention between the two members, the Board agreed to postpone indefinitely the establishment of a par value for Pakistan, and reserved the right to fix the date when the matter would be further considered.

The issue was settled in March 1951, when the Board agreed with Pakistan on an initial par value of PRs 1 = $0.30225. The upward movement of world prices following the outbreak of hostilities in Korea, together with an increasing demand for Pakistan’s exports, had strengthened the case for maintaining the original exchange rate of the rupee. But intrinsic to the Fund’s action was that in February 1951 the Governments of India and Pakistan had concluded a new trade and payments agreement, providing for the purchase and sale of each other’s currencies and for conversion of balances at the par value of the Pakistan rupee as declared to the Fund. Therefore, the Indian Government no longer pressed its objection to acceptance by the Fund of the par value proposed by Pakistan.

This rate for the Pakistan rupee was maintained until July 1955, when the Board concurred in a devaluation of 30.5 per cent, bringing the Pakistani rate into line with other sterling area currencies. This move followed a period in which the prices of Pakistan’s exports had been declining under the impact of increased competition in agricultural markets, while at the same time the prices of home-consumed goods had risen because of the monetary expansion induced by the financing of large capital expenditures.

An exchange rate for a planned economy—Yugoslavia

One of the questions about exchange rates which was receiving heightened attention toward the end of 1965, especially in discussions about Eastern European economies, was one with which the Fund had grappled in its early days: what is the role of the exchange rate in a centrally planned economy? This question had arisen at Bretton Woods. The delegation of the U.S.S.R. had proposed the provision which became Article IV, Section 5 (e), that a member may change the par value of its currency without the concurrence of the Fund if the change does not affect the international transactions of members of the Fund. The question of the significance of the exchange rate in a socialist economy became a practical one for the Fund in August 1947 when Yugoslavia, an original member, requested the Fund to consider a par value for the dinar. The establishment of a par value had been postponed in December 1946 because Yugoslavia had found it difficult to supply adequate basic data to the staff.

The Yugoslav authorities communicated a rate of Din 50 = $1. A report by the staff on the Yugoslav proposal commented that comparisons with the cost of living in other countries and with wholesale prices in the United States showed the dinar only slightly overvalued at this rate, but pointed out that these comparisons were of doubtful validity since supply, demand, prices, and wages in Yugoslavia were all governmentally controlled. But the staff also noted that devaluation of the dinar would not help to increase that country’s exports in the short run. The Executive Board considered the staff’s findings, but believed that more information was required before a par value could be agreed.

It was not until May 1949 that the Board again considered this matter. Meanwhile, in the course of 1948 the staff examined the significance for the Fund of an exchange rate in a state-controlled economy. Two main conclusions were reached. One was that a state-controlled economy could have a fundamental disequilibrium in the Fund sense of that term. The second was that a member could alter its par value in order to correct a fundamental disequilibrium only after consultation with the Fund; this obligation applied to all members.

When the par value for the dinar was discussed by the Board in May 1949, the Fund agreed to the communicated rate of Din 50 = $1. Mr. Kolovic (Yugoslavia), Alternate to Mr. Sucharda, argued that, although the function of the par value in a planned economy was somewhat different from its function in a free economy, it was of equal importance for either type of economy. He contended that, although in Yugoslavia imports were limited by an economic plan, in drawing up that plan the calculations of trade items were based on the purchase prices of these items expressed in terms of foreign currencies at existing parities plus distribution costs. Hence, the exchange rate played the same role for trade items in a planned economy as in a free one.

Mr. Southard (United States) questioned this line of reasoning. In his view, the real function of a par value in a centrally controlled economy was difficult to understand. The exchange rate was probably a subordinate policy instrument, the main adjustments being made through internal controls. Because of the unusual nature of the economy and the absence of the usual criteria, Mr. Southard could not regard the acceptance by the Board of the Yugoslav par value as implying that the economy was in complete equilibrium.

Mr. Sucharda (Czechoslovakia) insisted that any reservations about the par value for Yugoslavia were unnecessary. He asked that the staff make comprehensive studies so as to dispel any doubts about the Yugoslav parity.

These antipodal views were in practice fused on January 1, 1952, when the par value for the dinar was altered from Din 50 = $1 to Din 300 = $1. This change in the exchange rate not only represented a drastic devaluation but also signified a far-reaching reform of the Yugoslav economic system, a reform which gave a much greater role to the exchange rate. It had been the Fund’s view that, under the previous system, the exchange rate had not been one of the determinants of foreign trade transactions.8 Such transactions were carried out by the state foreign trade monopoly according to a plan, without much regard to the relations between prices in Yugoslavia and abroad. Now, as a result of a process of reorganizing and liberalizing the Yugoslav economic system and equilibrating Yugoslav prices, begun in 1950, a new rate came into being.

The Fund concurred in the devaluation. Explaining its action, the Board stated that the new exchange rate would help to forge a link between domestic and foreign prices, and put the exchange rate in line with the intended level at which prices were to be unified and stabilized.9

Later on, Yugoslavia was to encounter serious difficulties in synchronizing domestic and foreign prices while at the same time attempting to use relative domestic prices both as incentives to producers and as taxation of consumers. This use of the price system was made inordinately difficult because of the absence of customs tariffs in Yugoslavia. Utilization of the exchange system for too many objectives led to the emergence of a complex maze of multiple rates: by 1955 the Yugoslav exchange rate system had become one of the most complicated of any maintained by a Fund member. By 1961, however, following a series of steps closely worked out between Yugoslavia and the Fund, this rate system had gradually been unified.

A dispute with Czechoslovakia

The issue of the role of the exchange rate in a totally controlled economy came up again in 1953. But this time the circumstances were entirely different. On a minor level, the Czechoslovak proposal in 1953 was not for the establishment of an initial par value but for a change in its par value. On a major level, the Fund’s relations with the member were in great contrast. Whereas cordial and friendly relations between Yugoslavia and the Fund had developed very early, the Fund’s relations with Czechoslovakia had by 1953 become extremely stiff.

On June 2, 1953, the State Bank of Czechoslovakia sent a cable informing the Fund of a change in the par value of the koruna that had become effective the day before. On July 1, the Managing Director, Mr. Rooth, notified the member that since it had not consulted the Fund on this matter, the item would be placed on the agenda of the Board for July 20, 1953, that the member was entitled to send a representative, and that the Fund would wish information on the reasons for the lack of consultation. Czechoslovakia replied that it would be represented by its Ambassador in Washington.

At the meeting on July 20, the representative of Czechoslovakia took the position that his Government’s action was covered by Article IV, Section 5 (e), of the Fund Agreement. Since Czechoslovakia conducted the overwhelming majority of its international transactions in foreign currencies, the change in the par value of the koruna did not and could not in any way affect international transactions.

Mr. Southard emphasized that consultation with the Fund was to take place before changes in the par values were made, and that such consultation was more than a matter of form. In his view, there was no basis for arguing that changes in par values which members claimed did not affect international transactions were not subject to the obligation of advance consultation. No such exception was stated in the Fund Agreement; on the contrary, advance consultation was required even where the Fund had no right to object (Article IV, Section 5 (c) (i)).

A salient question was whether the change in the exchange rate did in actuality affect international transactions. Unless the rate had some international effect, why was it being changed? The Executive Directors took the view that, before the Fund could judge the impact of the alteration in the par value on international transactions, it would have to have information on how the country’s foreign trading system operated. Were any of the outstanding debts which were altered by the monetary reform owed to foreign countries or their residents? At what exchange rates were the koruna clearing balances of foreign countries converted from old to new korunas? At what rates were other koruna debts to foreign countries converted? What were the total balances, both in korunas and in foreign currencies, between Czechoslovakia and other countries before the exchange rate was changed, how were they distributed between countries, and how were they affected by the revaluation of the rate?

The representative of Czechoslovakia forwarded these questions to his Government. Temporary postponements of the date set for further consideration by the Board were granted, and it was the middle of September 1953 before Czechoslovakia was ready to reply. The representative of Czechoslovakia then read a prepared statement in which he argued that the economy of his country was very different from that of other Fund members and that the change in par value did not affect international transactions. The gist of his contention was that the Czechoslovak balance of payments was the result of a national economic plan. The balance of payments expressed the country’s foreign trade plan; equilibrium in the balance of payments was always maintained. The principal motives for the change in par value were internal. The change in no way affected either the current or the capital transactions of other members of the Fund. The Czechoslovak representative thought it was significant that some Fund members—in their payments agreements with Czechoslovakia—had already recognized the new par value. He further supplied brief information on the questions asked by the Fund but did not provide basic background material.

Meanwhile Mr. Southard had suggested that action be taken under Article XV, Section 2 (a), to make Czechoslovakia ineligible to use the Fund’s resources. He noted the several occasions on which the member had failed to supply adequate information to the Fund, and pointed out that in its financial relations with the Fund Czechoslovakia was using an unauthorized exchange rate which reduced the number of korunas due to the Fund. However, by October 1953 the issue of the par value was just one of several controversies between Czechoslovakia and the Fund. From September until mid-December 1953, the Directors repeatedly deliberated the par value issue per se. Representatives of Czechoslovakia were present at a series of meetings. Their contention continued to be that, as the change in the par value was not intended to correct a fundamental disequilibrium, and did not affect international transactions, it had not been necessary for Czechoslovakia to consult the Fund. The representatives discussed detailed items of the balance of payments accounts in an effort to prove that the change in parity had not affected these accounts. They explained how foreign trade was conducted in Czechoslovakia. However, such important information as data on the levels and composition of exports and imports, or quantitative estimates for the balance of payments and for national income, could not, they insisted, be revealed for security reasons.

The Executive Directors were unable to accept the arguments put forward by Czechoslovakia. The staff had shown that diplomats and tourists were affected by the new rate. That being the case, Czechoslovakia was not justified in claiming the right to alter its exchange rate without Fund approval. The contrast with the position taken by Yugoslavia, also a socialist economy, on the relevance of its exchange rate was especially noted.

On December 16, 1953, the Board took the following decision:

  • The Government of Czechoslovakia on June 1, 1953, changed the par value of its currency, the koruna, and subsequently informed the Fund that concurrence of the Fund with this change was not required because Czechoslovakia had taken the action in accordance with the provisions of Article IV, Section 5 (e). Having considered the arguments offered by Czechoslovakia, and such information as was made available, the Fund concludes that the change of par value by Czechoslovakia does not come under Article IV, Section 5 (e).


The establishment of an initial par value is, by Article XX, Section 4 (c), a prerequisite to the use of the Fund’s resources. Hence, countries without initial par values are, in the absence of any further Fund decision or action, ineligible to draw upon the Fund. Prima facie, this deprives members of financial support from the Fund even when they are, for example, contemplating extensive exchange reforms or coping with a temporary balance of payments problem.

Thailand queries its position

Exceptions to the above provisions were possible to members whose metropolitan territory had been occupied by the enemy. Such a member was granted certain privileges under Article XX, Section 4 (d). According to subsection (i) of this Section, the ninety-day period prescribed by Article XX, Section 4 (b), could be extended. Subsection (ii) further provided that:

Within the extended period the member may, if the Fund has begun exchange transactions, buy from the Fund with its currency the currencies of other members, but only under such conditions and in such amounts as may be prescribed by the Fund.

The privileges of Article XX, Section 4 (d), had been automatically accorded to original members whose metropolitan territories had been occupied by the enemy. When Italy became a member in 1947, the Membership Resolution incorporated references to Article XX, Section 4 (d), thus making explicit the possibility that Italy could draw on the Fund prior to agreeing a par value. The question of drawings for countries without parities was queried rather forcefully at the Annual Meeting, 1952, by the Governor for Thailand, Prince Viwat. The Membership Resolution for Thailand, which became a member in 1949, had contained no provision to the effect that the privileges of Article XX, Section 4 (d) (ii), be accorded to that country. Therefore, the Executive Directors were not authorized to permit exchange transactions with Thailand under that provision. Furthermore, paragraph 5 of the Resolution stated categorically that the member “may not engage in exchange transactions with the Fund before the thirtieth day after the par value of its currency has been agreed.”

This particular clause relating to exchange transactions with the Fund had appeared for the first time in 1948 in the Membership Resolution for Finland and had subsequently been repeated for Austria and Burma, as well as for Thailand; the same paragraph also occurred in the Membership Resolution for Indonesia, which was, in 1952, in the process of joining the Fund. The staff explained to the Board that the wording for those countries joining the Fund after Italy had been changed because it had been cumbersome to draft membership resolutions in such detail as that for Italy. Now, in 1952, Thailand was asking whether there were any circumstances which would prevent Thailand from enjoying the privileges of Article XX, Section 4 (d) (ii), and, if so, whether these obstacles could be removed by any action of the Fund.

Prince Viwat, in September 1952, pointed to the nature of the balance of payments difficulties of Thailand. His country’s deficits, when they occurred, seemed to exemplify the classic textbook case for use of the Fund’s resources. Thailand was a large exporter of rice and normally had little balance of payments problem. But since the successful production of rice depended upon the quantity of rainfall, he expressed Thailand’s potential predicament in 1952 in four words—“no rain, no rice.” 10 Any balance of payments difficulty would be temporary and likely to be self-correcting, since a two-year failure of the monsoon was a rarity. Therefore, he asked, should a country which is creditworthy and has a good record as a debtor, if faced with a temporary and self-correcting balance of payments difficulty, be barred from using the Fund’s resources simply because it has not been able to declare a par value? The Managing Director replied that he hoped that countries in positions similar to that of Thailand would soon be able to agree with the Fund on par values for their currencies. Should an emergency develop, however, the Fund and the member would endeavor to find a practical solution for the problem.11

The question of the need for par values prior to Fund drawings was carefully considered by the Board beginning in mid-1953. Mr. Rajapatirana (Ceylon), Alternate to Mr. Yumoto (who had been elected by Thailand) called attention to Prince Viwat’s statement of several months before and urged the Directors to reconsider whether the relevant paragraph of the Membership Resolution for Thailand was the correct condition to be imposed. In his opinion, an injustice had been done to Thailand and should not be perpetuated because that country, owing to circumstances beyond its control, could not comply with the technicality of declaring a par value. (These circumstances concerned cross rates, which are discussed later in this chapter.)

Action considered for a few countries

The staff’s first recommendations on this subject were considered by the Board in June 1953. The conclusion of the staff was that, under the Articles of Agreement and paragraph 5 of the Membership Resolution for Thailand, the Executive Directors could not permit Thailand to buy exchange from the Fund before a par value for its currency was agreed. However, the Board of Governors could authorize the Executive Directors to permit exchange transactions with Thailand, or any other member in similar circumstances, on terms substantially similar to those prescribed in Article XX, Section 4 (d) (ii), notwithstanding any provision of the original membership resolutions for such countries.

A draft text of a resolution for adoption by the Governors was therefore submitted by the staff to the Executive Board for consideration. This draft resolution covered Thailand and Burma, and would cover Indonesia if it accepted membership under the terms of its Membership Resolution. For additional new members the staff proposed that the problem be handled in the drafting of the membership resolutions.

When this draft resolution came before the Board, a few Directors insisted that the communication of a par value should be retained as a requirement for drawings. But several Directors believed that new members without par values should have at least the same privileges for drawings as those members which, while having agreed par values with the Fund, were not necessarily maintaining them.

Mr. Southard went further and wondered whether the staff suggestions were too limited. He questioned the wisdom of the distinction, implicit in the staff approach, between enemy-occupied countries and those not so occupied; the draft resolution would, moreover, provide a solution only for Thailand, Burma, and possibly Indonesia. There still remained the problem of Uruguay, an original Fund member without an initial par value, which could not be accorded the privileges of Article XX, Section 4 (d) (ii), since its territory had not been occupied by the enemy in World War II. The Board thereupon agreed to postpone discussion of the draft resolution pending a further review by the staff.

General action considered

The next discussions by the Executive Board weighed the pros and cons of action on this issue for a great number of countries. Directors recognized that the Board of Governors had the legal authority to provide in a membership resolution for exchange transactions between the Fund and a new member prior to agreement on a par value, even in those instances where the member had not been occupied by the enemy within the meaning of the Articles. But several Directors were concerned that any general action might grant new members which had not been occupied by the enemy a privilege which some original members did not have. Several Directors recognized that with the passing of time there was a diminution of the original significance of occupation by the enemy. However, they doubted the wisdom of proposing at this time any action by the Governors which would be generally applied to all new members; such action would amount to discrimination against original members that had not yet set par values.

On the other hand, some Directors noted that there was only one original member, namely Uruguay, that did not have the right to purchase exchange from the Fund because it had not agreed a par value with the Fund.12 Perhaps some solution might be found for that case. These Directors stressed that the Fund already seemed to have a new category of members not mentioned in the Articles, namely those which did not make their par values effective. They suggested that there seemed to be no point in requiring new members to declare a par value as a prerequisite to using the Fund’s resources if original members did not live up to their obligations under Article IV, Section 3, and yet remained eligible to engage in transactions with the Fund.

In line with this reasoning, the staff then proposed a broad approach, viz., that the Board of Governors should be asked to give the Executive Directors discretion to engage in transactions on an ad hoc basis before a par value was agreed, when the circumstances of a particular country justified assistance from the Fund. If this approach was too comprehensive, the staff recommended as a minimum that a proposal be made to the Board of Governors covering new members whose territories had been occupied by the enemy.

Although two Directors favored the general approach suggested by the staff, most had difficulties with any broadly applying resolution. Universal action would grant the privileges of Article XX, Section 4 (d) (ii), to countries never occupied by the enemy, thereby granting greater privileges to some new members than were given to original members. A clear-cut distinction between old and new members raised many delicate points which would provoke lengthy debate in the Governors’ Meeting.

Agreement for Thailand and Indonesia

Accordingly, the Board went back to considering a limited action which would apply only to Thailand and Indonesia. (Burma, in the meantime, had agreed an initial par value.) No new category of members would thus be set up. The Board, aware of the economic situations in Thailand and Indonesia, acknowledged that the establishment of par values might not be possible for some time. It was clear that special action for Thailand and Indonesia could not be regarded as discriminatory against Uruguay, and that such special action for two countries would not set a precedent.

In August 1953, the Executive Directors approved a report to the Board of Governors and two draft resolutions amending the Membership Resolutions for Thailand and Indonesia.13 The draft resolution for Thailand read as follows (that for Indonesia was similar):

  • Resolved:
  • That paragraph (5) of Resolution No. 3-4 of the Board of Governors on Membership for Siam [Thailand] be amended to read as follows:
  • (5) That Siam may not engage in exchange transactions with the Fund before the thirtieth day after the par value of its currency has been agreed in accordance with (4) above and its subscription shall be paid in full before such thirtieth day; provided, however, that at any time before such thirtieth day the Executive Directors are authorized to permit exchange transactions with Siam on the same terms and conditions as those prescribed by Article XX, Section 4 (d) (ii), of the Articles of Agreement.

These two resolutions were passed by the Board of Governors in September 1953.14 The amended paragraph 5 has served as an archetype for later membership resolutions.

Agreement for all new members

The Governors’ resolutions on Thailand and Indonesia were precursors to much broader action ten years later. The question of the need for initial par values prior to drawings again came up in 1962–63 following the admission to Fund membership of 24 new African countries. The circumstances of these countries suggested that many of them might not be able to establish initial par values for some years. Would they, on that account, not be eligible to draw on the Fund? When, in 1964, the Executive Directors considered this question for a second time, they were no longer disturbed at the idea that countries might use the Fund’s resources prior to the establishment of par values, and thus were able to accept a general change in the Fund’s policy that they had rejected eleven years earlier.

The first step taken to make new members eligible to use the Fund’s resources without waiting for the establishment of initial par values was to reduce the waiting period between the time when a member country had agreed to a par value with the Fund and had paid its subscription and the time when it was eligible to use the Fund’s resources. The standard provision in membership resolutions had been a waiting period of at least thirty days.

Mr. Saad (United Arab Republic), who had been elected by Afghanistan, in a discussion of a proposed par value for Afghanistan in January 1963, made the suggestion that there be an “acceleration of eligibility” by reducing this period to one day. In March 1963 the Board agreed to eliminate the delay altogether and recommended to the Board of Governors that the change should apply both to future membership resolutions and to all past ones. This was agreed by the Governors in April 1963.15 The countries for whom membership resolutions were retroactively amended in this way were Cameroon, Central African Republic, Chad, Congo (Brazzaville), Dahomey, Gabon, Guinea, Ivory Coast, Jamaica, Korea, Kuwait, Laos, Liberia, Nepal, Niger, Nigeria, Sierra Leone, Somalia, Tanganyika, Togo, Tunisia, Upper Volta, and Viet-Nam.

On April 22, 1964, in a series of three decisions, a much more sweeping policy was formulated. First, the Executive Directors recommended to the Board of Governors an amendment of the membership resolutions for all those members that had paid their subscriptions but had not yet agreed initial par values with the Fund, viz., Algeria, Burundi, Cameroon, Central African Republic, Chad, Congo (Brazzaville), Congo (Leopoldville), Dahomey, Gabon, Guinea, Ivory Coast, Kenya, Korea, Laos, Malagasy Republic, Mali, Mauritania, Nepal, Niger, Rwanda, Senegal, Sierra Leone, Tanganyika, Togo, Trinidad and Tobago, Tunisia, Uganda, Upper Volta, and Viet-Nam. This amendment authorized the Executive Directors to permit exchange transactions with any of these members prior to establishment of an initial par value, under such conditions and in such amounts as might be prescribed by the Executive Directors. The Board of Governors approved the amendment on June 1, 1964.16

The second decision was to the effect that the provision permitting transactions with new members prior to the setting of par values should be incorporated in all future membership resolutions, and the Directors agreed on guidelines for implementing this provision.

In the third decision, the Board emphasized that the two foregoing decisions should not be interpreted as a sign of weakened emphasis on the par value system. It was agreed that, when considering requests to draw by members that had not established par values, the Fund would be guided by the purposes of the Articles. Moreover, the Fund would encourage members to follow policies leading to the establishment of realistic exchange rates and to the adoption at the earliest feasible date of effective par values, and would take into account the efforts being made to achieve these objectives.

However, the Fund would give the overwhelming benefit of the doubt to requests for exchange transactions within the gold tranche. Further, members might expect that requests for drawings would be met where they were made in accordance with the decision on compensatory financing of shortfalls in export receipts.17 A further condition for transactions with a member without an initial par value was that the member should complete the payment of its subscription on the basis of a provisional rate of exchange agreed with the Fund.18

Why had such comprehensive action finally been taken? First, the terms on which Fund resources would be made available were becoming more liberal. At about this time the Fund was increasing the availability of the gold tranche to members in general. Also, in 1963 the Fund had introduced facilities for compensatory financing of export fluctuations and wished to permit the use of these facilities by members who, although otherwise eligible, had not declared par values.

Second, the staff presented several arguments in favor of permitting exchange transactions before par values were agreed. The use of the Fund’s resources by members suffering temporary payments difficulties could encourage them to follow policies leading to the establishment of adequate exchange rates and, ultimately, to realistic par values. Moreover, allowing members to draw on the Fund before they had established par values would make it unnecessary for them to propose possibly unrealistic par values in order to become eligible to draw. The staff argued that unrealistic par values should be avoided because, once set, they were often adjusted only after severe damage had been done to the economy of the country concerned. Also, the acceptance by the Fund of unrealistic par values for some members might discourage other members from adopting appropriate ones, or make it difficult for them to do so.

Third, to preclude transactions with new members which had not established initial par values, while permitting transactions with other members which did not use their par values or whose par values were not realistic, might also be regarded as unfair.

Finally, the staff found little logic in the practice of withholding the right to use the Fund’s resources from countries that had joined the Fund after Indonesia and Thailand and which had also been occupied by the enemy, for example, Korea and Viet-Nam.

On the other side of the argument, the staff recognized the need to bear in mind the risk of weakening the authority of the Fund in the field of par values. Any concession might be interpreted as a departure from the Fund’s general policy of encouraging members to establish realistic and effective par values. It might also lead to the relaxation by some members of their efforts to establish such par values at an early date.

At the Board meeting at which these staff recommendations were discussed, the Directors focused on the question: If general action was taken permitting drawings by countries without parities, would the principle of fixed par values, so important to the Fund, be threatened? The Directors indicated a strong awareness of the likelihood that the establishment of par values in many developing countries might take a long time. They agreed that where a developing country could establish an initial par value at an early stage in its membership, it should do so, their reasoning being that where countries had reached an appropriate stage of development par values were desirable. At the same time, some Directors voiced the thought that the establishment of a par value could not be regarded as a precondition of economic progress, and in certain circumstances might even be detrimental.

Both the Managing Director and Mr. Saad reminded the Board that the proposed third decision was really a reaffirmation of the Articles. Stress was placed on the statement in the draft that the Fund would “encourage members to follow policies leading to the establishment of realistic exchange rates.” The purpose of the decision would be to avoid any impression that the Fund was no longer giving the same pre-eminence as in the past to the par value principle.

Directors also agreed that the impact of the proposed decisions was that exchange transactions should not generally be permitted prior to the establishment of an initial par value, but only in appropriate cases. The Board recorded its obligation to make a careful and individual approach to requests for drawings, considering the particular aspects of each application and not losing sight of the purposes of the Articles of Agreement, in which the establishment of an initial par value was a basic principle.

Mr. Kiingi (Uganda), Alternate to Mr. Kandé, emphasized that while the adoption of these decisions would make it possible for the 29 member countries which had not yet established par values, and any countries which later joined the Fund, to participate fully in the major benefits to be derived from Fund membership, this possibility did not mean that the African countries would not try to establish realistic par values whenever possible.

Another problem which concerned the Directors as they reached the decisions mentioned was that the Board might be more lax in granting requests to use the Fund’s resources from the new members affected by the decisions than requests from old members, some of which also did not have par values, or did not have effective par values. For these reasons, some Directors suggested that it might be desirable later to review this departure from past Fund policy. The Managing Director emphasized that no discrimination between old and new members was intended.

Par values and stand-by arrangements

The Directors have agreed to stand-by arrangements with several member countries which had not established par values or did not have effective par values. The first such member was Peru, in February 1954.19 This arrangement was approved by the Directors even though, in the view of the staff, Peru’s economic circumstances seemed favorable for setting a par value and its current exchange rate appeared stable.


The Fund’s policies toward the setting of new par values by members that had suspended parities agreed earlier, or had made them nominal by subsequent changes in their exchange systems, have been as pragmatic as its policies toward initial par values.

Par values and multiple rates

In the Fund’s early years, it was not prepared to see initial par values remain unaltered while members made changes in their de facto exchange rates. Consequently, many countries with multiple rates not only established initial par values despite the multiplicity of their exchange rates; they also altered those par values when they changed their exchange systems. In effect, countries made an effort to keep their par values in line with their current exchange rate structures. For example, until 1956 Bolivia from time to time introduced new parities as it made alterations in its multiple rate system. Similarly, Ecuador adjusted its par value in 1950 when making changes in the rate structure. An extreme example is offered by Paraguay, which changed its par value once in 1951, twice within eight months in 1953–54, and again early in 1956—all of these par values being accompanied by a continued multiplicity of exchange rates. The logic of changing par values in these instances was that changes in the exchange system represented genuine steps toward more realistic rate patterns; hence par values should be moved in line with the improved exchange rate structures.

After the mid-1950’s, however, as the difficulties of final elimination of multiple exchange rates became more apparent, there was less resetting of par values in line with frequent adaptations of multiple rates. Most changes in multiple exchange rate systems, and even unifications of multiple rates, took place without the establishment of new par values, either immediately or even for some years thereafter. Only after the new system had proved itself to be enduring did new par values tend to be proposed.

Paraguay again provides an interesting example. In 1957, it replaced its multiple rate system with a fluctuating rate of exchange. Subsequently, although the guarani has depreciated by some 50 per cent, Paraguay has not declared a new par value. Similarly, several other countries which had eliminated their multiple rate systems and had had stable, unitary rates for many years, still had not instituted new par values by the end of 1965. These included Bolivia, Lebanon, and Peru. Still others—for example, Venezuela—had considerably simplified their exchange systems but had not yet set a new parity.

Urging par values

Where conditions have been considered eventually to have become appropriate, the Fund has urged countries to set par values—either initial parities or realigned parities in accordance with their de facto exchange rates. Thailand provides an example of such action by the Fund. After Thailand had eliminated multiple rates in favor of a fluctuating rate, and after the fluctuating rate had been stabilized, the Fund urged the Thai authorities to set a par value: from early in 1960 until late in 1963, when a par value was agreed, several statements were made in Board decisions on Article XIV consultations to the effect that the Fund would welcome the establishment of a par value.

Par values and convertibility status

The facts that some countries did not establish an initial par value, and that others did not establish a new parity after their earlier ones had become outdated, raised a further question: Should such countries be permitted to undertake the obligations of Article VIII? Canada, which had suspended transactions at its par value in 1950, had come under Article VIII in 1952. At that time, however, the question of the need for a member to have an effective par value prior to assuming Article VIII status had not been pointedly discussed. In 1960 this question was one of several considered by the Board at length.

The preparatory paper by the staff had suggested that countries assuming Article VIII status should conduct their international transactions through a uniform rate system based on a par value agreed with the Fund. This paper outlined various categories of countries which should probably not assume Article VIII status. One such category was of countries which had few or virtually no restrictions but where considerable uncertainty prevailed concerning balance of payments prospects. Doubt was cast on the ability of such countries to cope with a deterioration of the balance of payments without resort to restrictions. Countries which had been unable to establish effective par values were included in this category, on the ground that avoiding restrictions was not a sufficient test of the strength of a country’s external position; what was needed as well was a demonstration that the country’s exchange rate could remain stable.

As the Board considered the possible requirements of Article VIII status, some European Directors expressed the conviction that the establishment of an effective par value was an essential qualification for a country moving to Article VIII. Mr. Southard, however, although wholeheartedly supporting the par value system, argued that the Fund should not put undue emphasis on this point. He agreed that where a country seemed to have no justifiable reason for not having a par value, the Fund should try to convince the member that it should establish one. But he also believed that the Board should recognize that there were instances where it was to a member’s advantage to have a fluctuating rate.

Other Directors also believed that it was desirable for the Fund to keep open the possibility that a country could come under Article VIII even if that country had no par value. They recognized that many countries which did not have par values were, nonetheless, relatively free of restrictions. These countries maintained dual exchange rates on the premise that they could keep capital movements unrestricted provided that a free exchange rate was applied to those transactions. Although in the 1950’s many countries that had dual exchange rates also had par values, countries that established par values in the 1960’s usually had first unified their exchange systems.

The question of whether a country coming under Article VIII should have a par value came to the fore in a specific instance in February 1961. Ten European countries and Peru had indicated to the Fund that they intended to accept the obligations of Article VIII; but Peru’s situation was somewhat different from that of the others in that it did not have a par value and, in fact, had a fluctuating rate. Some Directors believed that the significance of Peru’s move to Article VIII was reduced by the fact that its exchange rate situation was irregular. They stressed that fulfillment of the Fund’s purposes required the removal of two obstacles, variations in exchange rates and restrictions on current transfers. The attainment of Article VIII status, in their view, should be indicative of economic stability, which should be reflected in a fixed rate of exchange. Several Directors, however, were especially gratified that Peru could join the list of Article VIII countries together with the European countries. Mr. Klein (Argentina), who had been elected by Peru, argued that a country could not always fix a parity without at the same time adopting a number of restrictive regulations. Some of the other countries then moving to Article VIII had at least some restrictions, while Peru did not.

The Board decided to note Peru’s intention of accepting the obligations of Article VIII from February 15, 1961, the same action as was taken for the other ten countries then assuming Article VIII status.


Once an economic issue has been closed, the passions that it once engendered are difficult to recapture or even to imagine. Such is the situation with regard to the Fund’s actions in connection with cross rates. The existence of “broken” or “disorderly” cross rates was the source of repeated discussions between the Fund and its members and among Board members from 1946 until about 1952–53, after which the prevalence of broken cross rates diminished. By 1958, most of such problems had disappeared, largely because the general economic environment which had occasioned them had changed.

Causes and the Fund’s early position

Broken cross rates are rates of exchange between two currencies which differ from the parity relationships between these currencies. If, for example, in 1948 in a free market in a third country, the U.S. dollar sold for 10 pesos and the pound sterling for 35 pesos, the cross rate between dollars and sterling would have been $3.50 rather than the official $4.03. In the late 1940’s and in the early 1950’s, even after the devaluation of sterling, broken cross rates occurred almost everywhere that limited free market rates or fluctuating rates prevailed.

These broken cross rates resulted from the inconvertibility of currencies. In normal circumstances, where currencies are freely interchangeable, their values in terms of each other are as a rule the same in exchange markets throughout the world. Any temporary divergence which may appear is quickly eliminated by arbitrage operations. However, when currencies are inconvertible, exchange dealers are prevented from carrying out the necessary arbitrage transactions. If the exchange control authorities do not cooperate to maintain agreed exchange rates, the rate for each currency is affected by the special circumstances of local supply and demand or by unilateral controls, and the resulting cross rates may have little relation to the parities agreed with the Fund.

Deviations from an orderly pattern of cross rates in the late 1940’s and early 1950’s were part of the abnormal circumstances of the times. Not only were currencies inconvertible, but transactions in exchange markets were very limited. Most central banks fixed exchange rates unilaterally and based their official quotations on parities, maintaining a small and usually uniform spread between their buying and selling prices for each currency. There was little scope for even minor movements of exchange rates. Exchange markets for transactions in the same currency in different countries were not interrelated.

Countries that had free markets in which broken cross rates existed usually had an excess supply of inconvertible currencies—some of which had accumulated during the war years—and a shortage of convertible currencies (mainly U.S. dollars). Broken cross rates helped these countries to encourage imports requiring inconvertible currencies and to make it more profitable to export to the dollar area than to countries with inconvertible currencies.

On the other hand, given the world shortages of goods and the buyers’ markets then prevailing, the trade effects produced by broken cross rates were adverse for the countries whose currencies were sold at implicit discounts. The discounted rate, while making it easier for the country with an inconvertible currency to sell goods to countries in which cross rates were broken, made it more expensive to buy from them. In the circumstances of the late 1940’s, the latter consideration was the more important one: shortages of supplies so limited exports that an exchange rate inducement was considered of no value. Instead, countries were eager to obtain an abundant flow of imports as cheaply as possible.

The United Kingdom was particularly concerned about broken cross rates for sterling. When, for example, the Italian authorities indicated to the Fund in 1947 that Italy would have a broken cross rate between sterling and the U.S. dollar because of its fluctuating rate, the British Government made a formal protest to Italy.

Partly, the United Kingdom was troubled by the adverse effects of broken cross rates, wherever they existed, on confidence in the sterling rate. But the unfavorable implications for British imports were also of consequence. In addition, broken cross rates induced traders to buy British goods for resale in the United States and thus diverted part of the dollar receipts which would otherwise have accrued to the United Kingdom and other countries in the sterling area. And broken cross rates diminished the incentive of countries in whose markets broken cross rates prevailed to seek to earn dollars directly.

The Fund Agreement contains no express provisions regarding cross rates. But as a matter of interpretation and policy, the Fund’s approaches at the outset were based on an insistence on orderly cross rates. The Fund recognized the adverse consequences for some members of broken cross rates, and also had various legal arguments against these rates. The General Counsel’s view was that to comply with the Fund Agreement a country with a single rate should not have cross rates differing from parity. Where such cross rates resulted from multiple rates, they were discriminatory. And, clearly, the general philosophy of the Fund Agreement made discriminatory practices undesirable.

In addition, in formulating policy on cross rates, the Fund used a broader test: What would be the implications of broken cross rates for the quick restoration of the convertibility of currencies? It was the Fund’s view at the time that disorderly cross rates distorted trade relations and made future convertibility even more difficult to attain. Changes in the direction of trade in response to the movements of disorderly cross rates were likely to be determined by short-run considerations and local peculiarities which had little to do with the fundamental influences by which a new pattern of world trade should be determined. Hence, the Fund attached great importance to the maintenance of orderly cross rates based on parities (1) because of its Articles, (2) as a means of protecting its members, and (3) as a way to encourage the appropriate reorientation of trade on a multilateral basis.20

Debates over policy

By March 1949 a few Directors, while recognizing the importance in the long run of orderly cross rates, had begun to question the extent to which the Fund should continue to insist on orderly cross rates in the circumstances then prevailing. Mr. Tasca (United States), Alternate to Mr. Southard, pointed out that in many cases the only alternative to breaking the cross rates was overt discrimination through quantitative restrictions against imports from the United States. In his view, therefore, the Fund could press for parity cross rates only as part of a larger program looking toward the elimination of all currency discrimination and inconvertibility. When after the devaluations of September 1949 a few countries continued to have problems in maintaining orderly cross rates, some Directors again queried whether the Fund should press the matter. These Directors were beginning to believe that some revision of Fund policy might be desirable. The issues raised are exemplified by the cross-rate problem of Thailand.

In March 1950 Thailand, which had a multiple exchange rate system, proposed a new exchange market in which broken cross rates would prevail. The argument was that broken cross rates in markets in other Far Eastern countries made it impossible for Thailand to maintain parity cross rates. The cross rates in these other markets fluctuated a great deal; however, by buying and selling dollars in the free market, the Thai authorities could compensate for these fluctuations, hold the free market rate stable, avoid large changes in the cost of living, and maintain a fairly satisfactory payments position. The staff thought that within eighteen months, or earlier if the sterling-dollar cross rate improved, Thailand might establish a par value.

The majority of Directors acknowledged that any attempt by Thailand to maintain orderly cross rates would result in Thailand’s reserves ending up in Hong Kong or other Far Eastern free markets. Certainly the system proposed would put little pressure on sterling. In the future, should the position of sterling in the Far East improve sufficiently, Thailand would, after unification of its rates, be in a good position to establish a par value and maintain regular cross rates. If, on the other hand, the sterling cross-rate problem was not solved for a considerable time, the arrangement proposed should still make it possible for Thailand to look forward to establishing a par value, even though a disparate cross rate would remain. These Directors reasoned that no member should be expected to undertake a stabilization of cross rates which was beyond its ability to achieve, nor should the objective of orderly cross rates be allowed to prejudice a member’s efforts to achieve and maintain an appropriate parity.

On the opposite side were Directors who had grave misgivings about any change in the Fund’s policy on cross rates. They contended, for example, that a broken cross rate in Thailand might afford the country an unfair competitive advantage over other areas, such as Latin America, exporting similar products. Mr. Crick (United Kingdom), Alternate to Sir George Bolton, was especially concerned about the prospect that the volume of transactions taking place at broken cross rates would increase considerably. To support his concern, Mr. Crick cited the major discounts for sterling which continued to prevail in markets other than the one in Thailand—those in Peru, Lebanon, and Hong Kong.

Because of these conflicting views, the Board, in 1950, 1951, and 1952, undertook a general reassessment of the Fund’s cross-rate policy. The staff paper prepared for these discussions suggested that the Fund should not sanction any exceptions to the obligations of orderly cross rates except in situations where it was clearly evident that a particular problem faced by members had no acceptable solution other than a break in cross rates. In general, Fund policy would continue to be not to approve of broken cross rates, but for the first time the Fund would recognize some exceptional cases.

After consideration, the Board concluded that a general decision on cross rates would not be desirable; rather, each case should be determined on its merits. The Board’s public statement ran as follows:

  • The area within which broken cross rates threaten to distort the normal flow of trade is now much narrower than it was two or three years ago. The cross rates of greatest significance are the sterling-dollar rates, and the increased relative strength of sterling has been the basic reason for the improvement. There are, however, still a number of countries, including Hong Kong, Peru, Syria, Lebanon, and Thailand, where cross rates have not been brought under control, and countries confronted with special exchange problems are still sometimes disposed to look for an easy solution to practices that involve broken cross rates.21

Thereafter it was increasingly recognized, by both the Board and the staff, that broken cross rates for inconvertible currencies could not be avoided in free exchange markets. The only way to eliminate them was to abolish the free markets; yet most free markets were regarded as desirable both by the country concerned and by the Fund.

The extent of broken cross rates subsequently decreased further. In 1953, under the EPU agreement, eight European countries—Belgium, Denmark, France, Germany, the Netherlands, Sweden, Switzerland, and the United Kingdom—set up a multilateral arbitrage system. These countries allowed their banks (when authorized) to buy from and sell to authorized banks in the other cooperating countries any of their currencies in exchange for the currency of any of the other cooperating countries. The central banks concerned undertook to intervene at stated margins, usually ¾ of 1 per cent either side of parity. Cross rates between the eight currencies involved henceforth became approximately the same in all eight markets. After the establishment of external convertibility for sterling and other major European currencies in 1958, the U.S. dollar also came into the circuit of multilateral arbitrage facilities. Hence, there were no longer any broken cross rates involving European currencies and the U.S. dollar. Earlier that year the Fund had specifically noted the marked decline in broken cross rates for sterling and other key currencies.22 Thus, although some broken cross rates have continued to prevail in free markets, they no longer constitute a serious issue for the Fund.


Once countries have set par values, they are obliged under the Articles to keep their spot exchange rates within a margin of 1 per cent on either side of parity. If two countries fix margins for their currencies in terms of the U.S. dollar, the cross rate between their two currencies can, under the Articles, vary up to 2 per cent from that calculated on the basis of their parities. After European countries introduced external convertibility (in 1958), although they did not in fact allow their currencies to move all the way to the 1 per cent margin—stopping at ¾ per cent—the possibility of an illegal cross rate still arose. This was because the margins from par for the exchange rates between EPU currencies in the market were the sum of the margins prevailing for each currency against the U.S. dollar. In the extreme case, if one currency moved to the maximum permitted premium against the U.S. dollar while another currency moved to the maximum discount, the rate of exchange between the two currencies would move to a point approximately 1.5 per cent from parity. Because of this potential margin of 1.5 per cent, the central bank whose currency was involved in the differential would find it useful to buy and sell interchangeable currencies at exchange rates related to their quotations in foreign exchange markets. Central banks would thereby be engaging in transactions at exchange rates of more than 1 per cent from parity.

In recognition of such situations, and to avoid the possibility of excessive rate movements, the staff recommended to the Board that countries should fix the margin vis-à-vis the currency which they selected within 1 per cent and then ensure that the cumulative margins on either side of parity in relation to other convertible currencies would not exceed 2 per cent. This approach, the staff believed, would in general meet member countries’ needs; any exceptional problems could be dealt with by the Fund on an individual basis.

When the Executive Board considered this recommendation, most Directors agreed with the staff’s analysis. The question arose, however, as to the legal basis for the Fund to allow larger margins than those prescribed by Article IV, Section 3 (a). Was the Fund departing from its Articles either by amendment or by interpretation? The view of the General Counsel was that the proposed margins were technically multiple currency practices and could be approved by the Board under Article VIII, Section 3, which states that “no member shall engage in … any discriminatory currency arrangements or multiple currency practices except as authorized under this Agreement or approved by the Fund.” Moreover, and even more important, the exchange systems in which these rates might develop already met or went far to meet the standards laid down in Article IV, Section 4 (a). Therefore, the General Counsel contended that the Fund could declare rates within a 2 per cent margin to be consistent with the Fund Agreement.

A few Directors noted that if rates outside the 1 per cent margin were legally approved under the authority of the Fund to deal with multiple currency practices, there might be some question that, like multiple currency practices, these rates were approved “on a temporary basis.” But the decisive element, in the Board’s view, was that the decision on margins about to be taken could not be interpreted as a violation of Article IV, Section 3. In any event, this decision could always be reversed later.

With these considerations in mind, the Board took the following decision on July 24, 1959:

  • The Fund does not object to exchange rates which are within 2 per cent of parity for spot exchange transactions between a member’s currency and the currencies of other members taking place within the member’s territories, whenever such rates result from the maintenance of margins of no more than 1 per cent from parity for a convertible, including externally convertible, currency.23


Quite early in the Fund’s existence, the question of the relation of forward to spot exchange rates arose. That the Fund was obliged to see that forward exchange rates bore a proper relation to spot rates was immediately recognized by both the staff and the Board. Provision had been made in Article IV, Section 3 (ii), for the Fund to determine what constituted a “reasonable” spread between spot and forward rates.

In 1946 the Board discussed a proposed section for the Rules and Regulations to the effect that agreed parities should not be circumvented by rates for forward exchange transactions. However, it was decided that it was not necessary for the Board to adopt a special regulation concerning forward transactions.

During the early 1950’s concern about the relation of forward to spot rates was somewhat intensified. Spreads in forward rates from spot rates began to emerge after free markets in forward exchange had been re-established in a number of countries. When the Exchange Restrictions Report was discussed in the Board in 1953, attention focused on exchange arrangements which had recently gone into effect in some countries. These arrangements limited the purchase and sale of certain currencies at spot rates. As a result, transactions in forward markets had increased at the same time that there was little, if any, official intervention to support forward exchange rates.

While the Directors agreed that the Fund should not at the time attempt to take a position with respect to forward exchange arrangements, they recognized that an increase in forward transactions might constitute a rather significant development. Traders would naturally deal in the forward market if they could obtain more favorable exchange rates for certain currencies than the rate within 1 per cent of parity available in the spot market. Possibly forward market arrangements could even put in jeopardy the principle of orderly cross rates. It was noted that discounts for certain currencies already existed in the forward exchange markets of some countries.

This problem was examined further during the course of 1953. The specific question of how the Fund was to determine a “reasonable” spread between a forward and a spot rate had arisen during the 1953 consultations under Article XIV with Belgium, in connection with the forward rate for Egyptian pounds. In first analyzing the factors responsible for a spread between spot and forward rates, the staff related premiums (or discounts) of forward rates over spot rates to interest rates. According to this analysis, at any given time the forward rate of exchange between two currencies is determined by the conditions of supply and demand in the forward market. But if, for example, the premium of the forward rate over the spot rate exceeds a certain margin, it becomes profitable for a bank to undertake interest arbitrage—that is, to sell the foreign currency forward and cover the exchange risk by buying the currency involved at the spot rate and investing that currency abroad. Therefore, under conditions of free competition, interest arbitrage will narrow down the forward premium (or discount) until it equals the difference in short-term (open market) interest rates in different countries.24

Wider spreads between forward and spot exchange rates resulted from limitations on access to spot markets, expectation of devaluation, or intervention in the forward market by monetary authorities. The Fund would, so the staff thought, have to judge instances of wider spreads in relation to the particular factor responsible for such a spread. For example, if an excessive spread between a forward rate and a spot rate resulted from controls on spot transactions, the Fund might treat the forward rate as a multiple currency practice.

The problem of spreads between forward and spot exchange rates diminished in the late 1950’s as external convertibility was established. Subsequently, the staff undertook studies of the implications of government intervention in forward markets.25 Later, as the world became more concerned about the role of interest rates in stimulating short-term capital movements and hence in facilitating balance of payments adjustment, the staff also examined the scope for movements of short-term funds in response to changes in forward rates.26 But in the absence of the need for policy decisions, these analyses, while of interest to the Board, were not generally discussed.


Camille Gutt, The Practical Problem of Exchange Rates, p. 4.


Annual Report, 1947, p. 27.


Resolution No. 1-6; Summary Proceedings, 1946, pp. 49-50.


Ninth Annual Report on Exchange Restrictions (1958), pp. 23–31; Tenth Annual Report on Exchange Restrictions (1959), pp. 29–33.


Twelfth Annual Report on Exchange Restrictions (1961), pp. 360–63.


“Statement Concerning Initial Par Values,” December 18, 1946, Annual Report, 1947, pp. 70–71.


Annual Report, 1948, p. 23.


Annual Report, 1952, pp. 64–65.




Summary Proceedings, 1952, p. 65.


Ibid., p. 124.


China was also an original member without a par value, but had not paid its subscription (also a prerequisite to the use of the Fund’s resources).


Summary Proceedings, 1953, Appendix VI, Annex A, pp. 120–22.


Resolutions 8-4 and 8-5, Summary Proceedings, 1953, pp. 96–97.


Resolution No. 18-2, Summary Proceedings, 1963, p. 227.


Resolution No. 19-8, Summary Proceedings, 1964, p. 260.


See below, pp. 403, 421.


E.B. Decision No. 1687-(64/22), April 22, 1964; below, Vol. III, p. 243.


See below, p. 471.


See, e.g., Annual Report, 1948, pp. 26–27.


Annual Report, 1951, p. 48.


Annual Report, 1958, p. 127.


E.B. Decision No. 904-(59/32), July 24, 1959; below, Vol. III, p. 226.


For a detailed exposition of forward exchange theory, see Paul Einzig, A Dynamic Theory of Forward Exchange.


S.C. Tsiang, “The Theory of Forward Exchange and Effects of Government Intervention on the Forward Exchange Market,” Staff Papers, Vol. VII (1959–60), pp. 75–106; J. Marcus Fleming and Robert A. Mundell, “Official Intervention on the Forward Exchange Market,” ibid., Vol. XI (1964), pp. 1–19.


William H. White, “Interest Rate Differences, Forward Exchange Mechanism, and Scope for Short-Term Capital Movements,” Staff Papers, Vol. X (1963), pp. 485–503.

    Other Resources Citing This Publication