5 Twenty Years with Par Values, 1946-66

Margaret De Vries
Published Date:
June 1986
  • ShareShare
Show Summary Details

The core of the international monetary system set up at Bretton Woods was a regime of fixed exchange rates, embodied in a system of par values. By 1966 this par value system, which had represented an entirely new venture in foreign exchange policy, had been in operation for 20 years. It seems appropriate, therefore, to take a look at the extent to which members had agreed and adhered to par values up to that time.


The original Articles of Agreement of the Fund required each member to agree with the Fund on an initial par value, that is to say, an exchange rate for its currency stated in terms of gold. The establishment of such a par value was basic; it was made a prerequisite for use of the Fund’s resources, and a member could be declared ineligible to use the Fund’s resources if it undertook an unauthorized change in its par value. Once a par value was agreed, it served as a firm base price for a member’s exchange rate: each member was obliged to maintain the spot exchange rate of its currency within a range of 1 percent of par. Any change from the initial par value had to be proposed by the member and (except for an initial cumulative 10 percent) approved by the Fund. A member could propose a change in the par value of its currency only to correct a “fundamental disequilibrium” in its balance of payments; the Fund on its side could not withhold its consent from a change that was necessary to correct such a disequilibrium.

The par value technique was thus one of “managed flexibility” of exchange rates—it stood between the permanently fixed exchange rates of the gold standard and fully fluctuating rates. Economists came to refer to it as the “adjustable peg” system, with par values being “pegs” that could be changed under specified conditions. Its objectives were twofold: to achieve international participation in exchange rate changes rather than let countries take unilateral action, and to maintain stable—as against frequently changing—exchange rates.


In mid-1946, as the Executive Directors, assisted by the Managing Director and staff, undertook their first deliberations, they decided to ask members to establish initial par values almost immediately, even though persuasive arguments could have been marshaled against doing so. The countries that had been devastated by World War II still had to undergo reconstruction; most countries were in the grip of serious inflation, widespread shortages of goods and services prevailed, and international trade was badly distorted.

These difficulties in the world economy, notwithstanding, the Fund was eager to get going and to make a start toward orderly exchange arrangements. On September 12, 1946 it called upon its 39 members to communicate their par values. At this stage, however, the Fund did not expect a general revision of the exchange rates. The crucial problem was to restore the export capacity of members, but this restoration was hampered more by disrupted production and transportation than by inappropriate exchange rates. In these circumstances, so long as a member’s exports continued to flow, the Fund was ready to regard the member’s existing rate as an appropriate par value. On this basis, on December 18, 1946, initial par values were announced for 32 members and also for many colonial territories.1 The other members—Brazil, China, the Dominican Republic, Greece, Poland, Uruguay, and Yugoslavia—requested more time, and the Fund agreed. In 1948 and 1949 initial par values were determined on much the same basis for the Dominican Republic, Brazil, and Yugoslavia and also for five new members (Australia, Lebanon, the Syrian Arab Republic, Turkey, and Venezuela). Greece and Uruguay, as well as Austria and Italy, which also joined in these early years, did not, however, set initial par values until much later.

Broken Cross Rates and Suspended Par Values Getting initial par values established proved easy. Implementing a par value system, on the other hand, proved to be much harder, especially since in the aftermath of World War II many foreign exchange problems inevitably arose. One problem that occurred frequently in implementing par values was that of “broken cross rates.” These were rates of exchange quoted in free markets in various places in the world, especially between the U.S. dollar and the pound sterling, that differed from the parity relationships between these currencies. For example, instead of the pound sterling being sold at the parity cross rate of $4.03 per pound, it was sold at a discount of, say, $3.65 per pound. This problem rose from the inconvertibility of currencies; in normal circumstances, where currencies were freely interchangeable, dealers in foreign exchange would take advantage of differences in rates and so keep the differences very small. The Fund attached great importance to the maintenance of cross rates because it feared that the direction of world trade that would otherwise be established would have little to do with the new pattern necessary for restoring the convertibility of major currencies.

Nor were broken cross rates the only problem facing the Fund in its early years as it tried to implement a par value system. Three members (France, Mexico, and Peru) very soon had to suspend their par values. In January 1948, France became the first nation in history that, instead of independently changing the par value of its currency, came to an international organization to ask its opinion and concurrence. The French proposal, however, involved not only a change in its par value, but also a provisional free market for certain currencies and transactions. It was to this free market that the Fund objected, on the grounds that the free market was liable to have serious adverse effects on other Fund members. France proceeded with its measures despite the Fund’s objections, thus making an unauthorized change in its par value. As a consequence, the Fund declared France ineligible to use its resources. This was the only occasion on which the Fund had so acted with reference to a member in good standing. (Czechoslovakia was also later declared ineligible to use the Fund’s resources, but as a step in a process that ended in the country being voted out of membership.) France made subsequent changes in its exchange system in October 1948 and in September 1949 which the Fund welcomed, and in 1954 France was declared again eligible to use the Fund’s resources. In December 1958 a new par value for the franc was agreed with the Fund.

Mexico’s suspension of transactions at its official rate came in July 1948, shortly after that of France. The Mexican authorities decided to let the market determine an exchange rate. Although to some extent imports were excessive, the main difficulty was capital flight. During the following year extensive consultations took place between the Mexican authorities and Fund officials, and in June 1949 a new par value was set.

Then, in November 1949 the Fund received a proposal by Peru to abandon its official par value and to permit the exchange rate to fluctuate freely until a “natural and stable level” was found. After extensive deliberation, the Fund expressed no objection to this fluctuating rate as a temporary measure, but with the understanding that Peru would remain in close touch with the Fund. (By 1966 Peru still had not agreed to another par value, but it had a unitary fluctuating rate that had remained stable for some years.)

The Devaluations of September 1949 By 1948 it was becoming increasingly evident that the initial par values set in 1946 would soon have to be adjusted. When the Fund was first formed, many officials and economists had feared that countries would resort to frequent changes in the par values, but by 1948 it had become clear that such fears were misplaced. Countries had in fact tended to lean in the opposite direction and were maintaining rates that should have been changed. Reasons for the hesitancy to alter exchange rates included fear of the inflationary consequences of exchange depreciation and concern about raising the cost of imports and encouraging capital flight, as well as skepticism, more prevalent in the late 1940s than in later years, about the effectiveness of exchange rate changes.

In September 1949, after the need for change in the exchange rates of the currencies of the Western European countries had become pronounced, there occurred the only series of sweeping devaluations of the 20-year period after World War II. The Fund concurred in proposals for changes in the par values of 13 members in this order: the United Kingdom, South Africa, Australia, Norway, India, Denmark, Egypt, Iceland, the Netherlands, Canada, Belgium, Iraq, and Luxembourg, and for the currencies of most of the then colonial areas. Generally, the currencies of these countries were devalued by 30.5 percent; the pound sterling, for example, went from $4.03 per pound to $2.80 per pound. The Fund also agreed to changes in the exchange rates of Finland, France, and Greece, for which no par values were in force.

Setbacks, 1950–56 During the early and mid-1950s, several countries joined the Fund and most of them agreed on initial par values: Burma, Sri Lanka (Ceylon), the Federal Republic of Germany, Haiti, Israel, Japan, Jordan, Pakistan, and Sweden; only a few of the new members in difficult circumstances (Afghanistan, Indonesia, Korea, and Viet Nam) did not set par values.

An important break with par values, nonetheless, occurred in September 1950, when Canada suspended its fixed rate and announced that the rate for the Canadian dollar would be permitted to fluctuate in response to market forces. The Fund, legally unable under its original Articles of Agreement to approve a fluctuating rate, recognized the particular Canadian circumstances that had caused such action, and it was agreed that Canada and the Fund would keep in close touch toward setting a new par value. With Canada added to France and Peru, there were three Fund members that had decided on a temporary abandonment of par values in favor of fluctuating rates. In addition, Italy and Thailand had never set par values and were using fluctuating rates.

Partly because of these important deviations from the par value system and partly because from time to time exchange crises emerged in other currencies, a few economists, such as Professors Milton Friedman, James Meade, and Gottfried Haberler, began to advocate the use of fluctuating exchange rates not only for a particular currency but also for most currencies. In effect, they argued for the substitution of general flexible rates for the par value system. The Fund, therefore, in its Annual Report, 1951, emphasized the reasons for its firm continued belief in the doctrine of par values.2 Experience had shown that stable exchange rates were the objective of virtually all Fund members; even those members with nominally fluctuating rates had in practice generally stabilized them over long periods. Moreover, there was no such thing as a “natural” level for an exchange rate; the proper rate depended, in each instance, upon the entire range of policies followed by both the member concerned and its trading partners. Fluctuating rates might also induce large speculative capital movements. In addition, fluctuating rates in individual member countries required a fixed frame of reference—that is, fixed rates for major currencies—which would be lost if the rates for major currencies also fluctuated.

The Fund also recognized, however, that there might be occasional and exceptional cases where a member concluded that it could not maintain any par value for a limited period. In these circumstances, although the Fund could not give approval to the withdrawal of support of a par value, if it found the member’s arguments persuasive, it would accept them, even while it was working with the member toward the eventual re-establishment of a par value.

Multiple exchange rates were another problem that arose increasingly during the early and mid-1950s.3 Many developing members gradually introduced multiple rates and then shifted more transactions to exchange rates other than their par values. Hence, in several instances, the par values that had been established were no longer “meaningful”—that is, the bulk of the member’s transactions were no longer conducted at par values. This situation involved a dual problem—how to unify multiple exchange rates and how to attain new par values at a later date. To help solve this dual problem, Fund officials had, since 1947, been evolving various procedures and policies. In mid-1957, they felt that the time was ripe for the Fund to intensify its efforts in this direction. As part of this intensification, the Fund stressed with its members that complex systems of multiple rates might be simplified and eventually a new realistic par value achieved by temporary use of a fluctuating rate.


In the early 1950s, progress toward a universal system of fixed par values thus received a setback. It was, however, only a temporary setback. In the second half of the decade progress toward attaining a par value system was resumed. By the end of 1958, following the establishment of external convertibility (i.e., convertibility of currencies when held by foreigners), the Fund was able to report that the problem of broken cross rates for Western European currencies no longer existed. In addition, members that had had difficulties in maintaining or achieving par values had gradually introduced or reintroduced them. France—without a par value for 10 years—set one in 1958. Italy, after 13 years of Fund membership, and Greece, an original member, agreed on initial par values in 1960 and 1961, respectively. In May 1962, after nearly 12 years with a fluctuating rate, Canada returned to the par value system with a new par value of 92.5 U.S. cents per Canadian dollar.

Most of the par values that were readjusted in 1949 were still being maintained at the end of 1966. Problems had arisen for the Canadian dollar and for the currencies of some countries with multiple rates. And when France set its par value in 1958 another 14.9 percent depreciation was undertaken, and in March 1961, 5 percent appreciations for the deutsche mark and the Netherlands guilder were made to lessen the expansionary pressures in the German and Dutch economies. Except for these situations, exchange rates by 1966 had been generally stable, especially for the main currencies, since the end of World War II; this stability was an important factor in the expansion of dollar earnings and in laying the basis for the restoration of convertibility of the Western European currencies.

Moreover, half of the 44 new members that joined the Fund from 1957 to 1966 also set initial par values: Burundi, Cyprus, Ghana, Ireland, Jamaica, Kuwait, the Libyan Arab Jamahiriya, Malawi, Malaysia, Morocco, New Zealand, Nigeria, Portugal, Rwanda, Saudi Arabia, Sierra Leone, Somalia, Spain, the Sudan, Trinidad and Tobago, Tunisia, and Zambia.

A steady wave of elimination of multiple exchange rates also commenced in the late 1950s as many members moved to abolish their multiple exchange rates. Several of them (Finland, Israel, Spain, and Turkey) devalued their currencies, eliminated multiple exchange rates, and set new par values in a single stroke. Others—Costa Rica, the Islamic Republic of Iran, Jordan, Nicaragua, the Philippines, Saudi Arabia, Thailand, and Yugoslavia—proceeded by stages, some first established unitary fluctuating rates and then stabilized them. In due course they all attained par values. Argentina, Bolivia, Korea, Lebanon, Paraguay, and Peru attained unitary fluctuating rates. And by 1966, the exchange rates of some of these members, Bolivia, Lebanon, and Peru, for example, although technically floating, had been fairly stable for a number of years.

It was especially important to the Fund not only that members establish par values but also that the par values established be “realistic.” They should be appropriate for the current balance of payments position and for the payments prospects of the member concerned. Rather than hasten the establishment of initial par values that might be inappropriate, the authorities of many members took the view that the establishment of realistic par values would be encouraged if the Fund permitted access to its resources to new members prior to their establishing initial par values. Hence, in June 1964 the Board of Governors approved an amendment to the membership resolutions of those members that had not agreed on initial par values. This amendment permitted these members, if the Executive Board so agreed, to draw on the Fund’s resources before they had set an initial par value. The Executive Board also decided that this provision should be incorporated in future membership resolutions. Enabling members to draw on the Fund before they had set par values proved exceedingly wise. By drawing on the Fund, members put themselves in a position where they were able to set an initial par value, an objective of the Fund. Thus in the next 2½ years, 6 of the 29 members concerned agreed to initial par values.


Ironically, the effective attainment of a par value system and of stable unitary exchange rates coincided with renewed attacks in the early 1960s by several economists on the way in which the par value system was working. Their complaint was that par values were not altered sufficiently and frequently; “adjustable pegs” were rarely adjusted. As a result, exchange rate changes were no longer an important instrument of balance of payments policy and a serious balance of payments adjustment problem arose.

By the mid-1960s the number of economists, especially in university circles, who were advocating more widespread use of floating exchange rates had grown considerably. Other economists were suggesting widening the range of the margins permitted under the par value system—something like 5–7 percent either side of par value instead of the existing 1 percent. A few economists suggested use of a “crawling peg”—automatic adjustment at periodic intervals, specified in advance.

The system of par values, of course, continued to have many academic defenders. In addition, Fund officials and most central bankers continued to favor—strongly—fixed exchange rates. Financial authorities argued that fixed rates provided a firm foundation for international trade and capital movements, whereas fluctuating rates created great uncertainties. Other arguments in favor of fixed rates were that national monetary authorities were forced to integrate their policies; that fixed exchange rates had to be defended by anti-inflationary policies; and that they therefore helped to secure domestic financial stability and that they eliminated the danger of competitive depreciation. Fund officials also had two additional arguments. First, the process of international collaboration in financial matters would be far more difficult under fluctuating rates. Second, the experience of most Fund members with fluctuating rates had been that their financial authorities preferred to stabilize the exchange rates of their countries, if at all possible. Thus, despite the serious weaknesses with the par value system that were becoming apparent in the 1960s, the trend toward more stable rates and eventually par values, as described, was indeed evident.

For all these reasons, when the Group of Ten agreed in 1963 to study proposals for reform of the international monetary system, changes in the exchange rate system were explicitly ruled out.4 A press release of the Group of Ten thus stated that: “The underlying structure of the present monetary system—based on fixed exchange rates and established price of gold—has proven its value as the foundation for present and future arrangements.” (Later, after severe strain, the system of par values did finally collapse in early 1973. These developments are described in Parts Two and Three.)

Note: This article was published earlier in a slightly different form in Finance & Development (Washington), Vol. 3 (December 1966), pp. 283–89. A longer version can be found in Margaret G. de Vries, “Fund Members’ Adherence to the Par Value Regime: Empirical Evidence,” Staff Papers, International Monetary Fund (Washington), Vol. 13 (November 1966), pp. 504-32.

Belgium, Bolivia, Canada, Chile, Colombia, Costa Rica, Cuba, Czechoslovakia, Denmark, Ecuador, Egypt, El Salvador, Ethiopia, France, Guatemala, Honduras, Iceland, India, the Islamic Republic of Iran, Iraq, Luxembourg, Mexico, the Netherlands, Nicaragua, Norway, Panama, Paraguay, Peru, the Philippines, South Africa, the United Kingdom, and the United States.

Pp. 36–41.

Multiple exchange rates are discussed in the article on pp. 21–29.

Ministers of finance and governors of central banks of the ten industrial members that agreed to lend to the Fund under the General Arrangements to Borrow. (See entry under October 24, 1962 in the Chronology.)

    Other Resources Citing This Publication