CHAPTER 7 Fluctuating Exchange Rates

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

In no field of the Fund’s endeavor have principle and expediency been more thoroughly interwoven than in that of fluctuating rates. Fixed exchange rates have remained the Fund’s main principle; but temporarily, and in certain circumstances, fluctuating rates have been recognized as useful expedients.

Reconciliation between the aim of exchange rate stability and awareness of the need for flexibility of exchange rates has come by degrees: over the years it is possible to discern almost step-by-step movements in the Fund’s policy in this regard. In its very early days the Fund put primary stress on the fact that fluctuating rates were inconsistent with the Articles. Quite soon, however, pragmatic considerations began to overshadow the more technical ones, and a less rigid approach was adopted. Hoping that a few exceptions would not cause a general return to the unstable rates of the pre-Fund era, the staff and the Board recognized the possibility that unusual circumstances might justify temporary departures from the standard prescribed by the Articles.

This toleration was not entirely a mere giving in to circumstances. Already the useful possibilities of less exacting policies toward exchange rates could be envisaged. Consequently, as successful experiences mounted, the Fund’s policy gradually extended to outright advocacy of fluctuating rates for those situations where they seemed especially fitting.

Nevertheless, these shifts in the Fund’s attitude toward flexible rates from 1948 to 1965 were not at the expense of the Fund’s belief in exchange rate stability. There were two reasons why the Board could accept fluctuating rates on a number of occasions without surrendering the prime objective of rate stability. First, fluctuating rates have, for several countries, turned out to be part of the journey to a fixed exchange rate. Second, the circumstances in which fluctuating rates could meet with Fund approval have been carefully delineated. Although liberalizing its policies on flexible rates, the Fund opposed all suggestions that a general scheme of fluctuating rates be instituted.


Mexico in 1948

That members of the staff and nearly all the Directors were anxious about the status under the Articles of a country that had suspended its par value, is clear from the action taken when the first such example came up for the Fund’s consideration. In July 1948 Mexico suspended all transactions at the par value, and proposed to allow exchange transactions to take place at whatever exchange rates the market set. Although imports were to some extent excessive, the country’s payments difficulty was due not so much to the balance of trade as to capital flight.

When the Mexican proposal was considered, the legal staff made much of the point that the Fund had no authority, under the terms of the Agreement, to approve a fluctuating rate. Not wishing to disapprove the Mexican action, the Board therefore did not take any decision, on the understanding that Mexico would establish a new parity “within three weeks.” It soon became evident that a new parity would not be so quickly set, and the situation then posed for the Fund, as a new international agency, was a most difficult one. The two questions regarded as basic were: Could the Fund agree to a continued violation of the Articles? Could access to the Fund’s resources be permitted during the period in which the par value was ineffective?

Mr. Gómez (Mexico), when explaining to the Board Mexico’s request for an extension of time, argued the necessity for Mexico to ascertain the level at which a new rate could be maintained before agreeing a new parity. The Mexican Government had adopted all the strict internal stabilization measures the Fund had recommended, but more time was still required. Although it was possible that a stable exchange rate might be maintained if exchange controls were introduced, Mr. Gómez made it clear that the Mexican authorities were especially reluctant to resort to exchange controls. It was not even certain that the authorities could enforce exchange controls.

Nonetheless, Mexico did not wish to place itself in the position of violating the Articles of Agreement. Mr. Gómez suggested, therefore, that Mexico’s situation might be considered as falling within the provisions of Article IV, Section 5 (c) (iii), which entitles the Fund to “a longer period in which to declare its attitude” concerning a proposed change in par value. Recognizing that this Section might not have been intended to cover the exact circumstances of the Mexican case, he urged that the Fund and Mexico should be guided by the principles rather than by the strict letter of the Articles.

Most Executive Directors feared dangerous implications in an interpretation of the Fund’s Agreement which would permit a member to change its par value by introducing a free rate. Although the staff and the Directors considered at some length the economic feasibility of fixing a new parity, they were even more concerned about Mexico’s formal status under the Articles. Mexico was not meeting the basic obligations of Article IV, Section 3, which required members to maintain exchange transactions within 1 per cent of parity. The request for an extension of time under Article IV, Section 5 (c) (iii), amounted to a request that the Fund condone a technical violation of its Agreement during the extended period of time.

Arguments for a new parity for Mexico

In the next several months a number of staff missions went to Mexico and the situation was discussed by the Board several times. The reasons why Mexico hesitated to set a new par value were considered, and each time the Fund stressed the economic advantages to Mexico of quickly doing so—the promotion of exports, the encouragement of foreign capital inflow, the return of domestic funds held abroad, and support for the determination of the Government to take the strong fiscal and credit measures necessary to redress Mexico’s economic difficulties.

In the early part of 1949 the United States was in the midst of a recession. Hence, one issue to which special attention was given was whether a country closely affected by business conditions in the United States could determine a new par value so long as the recession there continued. Following the experiences of the 1930’s, the fear continued to prevail in the late 1940’s that fluctuations in U.S. business conditions would seriously harm the United States’ trading partners.

The Fund took the view that the fixed rate system of its Agreement had already taken into account the possibility of changing business conditions in its largest member countries. Swings in business conditions should, therefore, not be the occasion for members to break away from the par value regime. Depressions and recessions were, instead, to be relieved by use of the Fund’s resources.

What disturbed the legal staff and several Directors most was the continuation of Mexico’s anomalous formal position in the Fund. Mexico’s exchange system remained unapproved by the Fund. Accordingly, in January 1949 a proposal was made in the Board, which received general support, to write to Mexico urging the establishment of a new par value by the end of March.

Another staff mission to Mexico followed the dispatch of this communication and intensive consideration by the Board was renewed. In July 1949—one year after suspension of the initial par value—a new par value was agreed and Mexico was declared again eligible to use the Fund’s resources.


The Belgian proposal

Preoccupation with the formal status under the Articles of members with fluctuating rates was short-lived. In September 1949, at the time of the general European devaluations, the Belgian Government requested the Fund to agree to its abandoning its existing par value and delaying establishment of a new one until an appropriate rate could be determined. The Executive Directors raised no objection to the plan, there being a general feeling that the Fund had to recognize the difficulties facing Belgium. Nonetheless, deliberation took place over the precise wording of a decision. The Fund could not agree to the Belgian proposal because it had no authority to approve the adoption of fluctuating rates. There was the further question as to how far the Fund should go in condoning Belgium’s action. The problem seemed simple: if Belgium found that circumstances made a fluctuating exchange rate imperative, the Fund should not try to prevent the introduction of that rate. However, it was quite another matter for the Fund to lend moral support to a fluctuating rate by formally expressing “approval” or “no objection.” Mr. Rasminsky (Canada), for example, was disturbed at the change in the attitude of the Directors toward fluctuating exchange rates. Previously, the Fund had held strictly that fluctuating rates were illegal under the Articles; now the Board was considering a decision in which the Fund would agree to the need for a fluctuating rate, or at least would not object. In his view, the Fund should only note the Belgian proposal, treat it as a situation of force majeure, and refrain from imposing any sanctions against the member.

Mr. Tansley (United Kingdom) agreed with Mr. Rasminsky. He was especially uneasy lest the decision be taken as an invitation to other members to have recourse to fluctuating rates for indefinite periods. Moreover, he thought that any intimation of approval would be unfair to those members who had taken the risks of deciding on a new par value.

Because of these considerations, the Board discussed imposing a one-month time limit, but decided that such a limit would be unwise. A draft decision in which the Fund agreed to Belgium’s plan was withdrawn. Two other alternatives—one in which the Fund raised no objection to Belgium’s proposal, and a second in which the Fund merely noted the Belgian measures—were formally voted upon. The first of these two won, by a vote of 53,600 to 38,490. The final decision was, therefore:

  • 1. Belgium has informed the Fund that it proposes to abandon the existing par value of the Belgian franc, beginning September 20, 1949, and, until the situation in the monetary field can be properly assessed so as to make possible a decision on the appropriate level for a new par value, no new fixed par value would be established; instead a free market for the U.S. dollar would be established in Brussels for the time being. Cross rates between the dollar and other currencies would be maintained and the normal operation of payments agreements would not be affected. The firm resolve of the Belgian Government would be to propose a new fixed par value for the Belgian franc as soon as circumstances warrant it. In the meantime the member would keep in close contact with the Fund concerning developments in the free market. The Fund has been requested to concur in the plan as an appropriate step in the light of existing circumstances….
  • 2. The Fund recognizes that the plan proposed by the Government of Belgium is an appropriate step in view of the exigencies of the situation. The Fund therefore raises no objection to the plan and takes note of the intention expressed by the Belgian Government to consult with the Fund and to propose a new parity as soon as circumstances warrant.

Two days later, on September 21, 1949, Belgium, fearful of the dangers of a fluctuating rate, proposed a new par value and the Fund agreed. The decision quoted above was revoked.

Peru turns to fluctuating rates

That the Fund’s views on fluctuating rates had been considerably modified was further evidenced in November 1949, when Peru suspended all transactions at the par value. The country had had an exchange system consisting of three separate markets, the rates in one of which were fixed on the basis of the par value. It now proposed that the market with fixed rates should be abolished and all transactions should be conducted in two markets, one for trade and the other for nontrade transactions. Although the rates in both markets would fluctuate, they were expected to remain close to each other.

The Peruvian authorities explained that the weakening of demand abroad had created problems for Peru’s metal exports. In addition, the accumulation of more sterling than could be used under Peru’s import restrictions had caused difficulties for exports to the sterling area. The authorities contended that the parity rate had thus become unrealistic but that it was impossible for the time being to determine the level at which a new rate should be fixed. A prestabilization period was needed.

In a preparatory paper, the staff, guided by the less rigid view toward fluctuating rates that the Board had taken in the Belgian instance, suggested that the Fund should merely note the Peruvian proposal and take no further action. The staff reasoned that, since a system of unitary, fixed rates was fundamental to the Fund Agreement, the Fund should do no more than recognize the circumstances in which a fluctuating rate was introduced and refrain from applying sanctions. It was to be understood that Peru could not use the Fund’s resources. The staff suggested that a time limit should be set within which a new par value should be proposed; otherwise, the Peruvian authorities might not have sufficient incentive to take the difficult steps necessary to stabilize the internal economy.

Several Directors, including Mr. Southard (United States), wished to go further in support of Peru’s action, and suggested that the Board should commend Peru’s new rate system as an effective step toward realizing the Fund’s objectives. Observing that the exchange system was essentially unified and that the new rates seemed realistic, they viewed Peru’s action as a substantial improvement on the complex multiple rate systems of many other countries.

As described in the preceding chapter, the Fund was becoming increasingly aware at this time that multiple rates would be difficult to abolish. Many countries had complex exchange systems based on par values which were clearly overvalued. By contrast, Peru’s simple dual market system with two fluctuating rates seemed much closer to a unitary, realistic exchange rate.

Emergence of a case-by-case approach

Not all Executive Directors shared this favorable view of Peru’s fluctuating rates. Recalling that a fluctuating rate had been under discussion only two months earlier when proposed by Belgium, some Directors insisted that the Fund’s action on Peru should not be taken as introducing a new general policy on fluctuating rates; as in the field of multiple rates, so with fluctuating rates, a preferable solution lay in a policy which considered each country’s situation on its merits.

The Board accepted this view and, on the understanding that the purpose of Peru’s fluctuating rates was the establishment of a unitary exchange system on a more appropriate level, decided not to object to the use of fluctuating rates as a temporary measure. It asked the Peruvian Government to remain in close consultation with the Fund in order to fix a new par value when circumstances permitted. Meanwhile, the Board considered it urgent that Peru should adopt appropriate fiscal and credit policies in order to restrain inflationary forces and assure internal financial stability, and instructed the staff to submit a complete review of Peru’s problems within six months.

The requirements of the Articles once again

During the foregoing discussion of the Peruvian proposal, the status under the Articles of a country with a fluctuating rate was again reviewed. This time some Directors considered it possible for the Fund to approve a fluctuating rate. They argued that Sections 6 and 8 of Article IV envisaged the eventuality that some members would have to abandon existing par values without immediately fixing new ones. Section 6 provided for sanctions against members who fixed par values over the Fund’s objection, but not against members who abandoned par values without moving to new ones. Section 8 provided for interim payments to maintain the value of the Fund’s assets by countries without par values.

The General Counsel did not agree. His view was that, although Article IV, Section 6, did give evidence that the Fund Agreement envisaged situations in which a member might operate temporarily without a par value, these situations would remain unauthorized and thus inconsistent with the Agreement. The member would be required to agree with the Fund on a par value within a reasonable period or the sanctions provided in Article XV, Section 2, might have to be invoked.

Also in November 1949, the Board considered the procedure for dealing with the exchange system of Thailand, which had a fluctuating rate as part of a multiple rate system. A staff mission was then in Thailand; and the Thai authorities, presumably prompted by its presence, had cabled the Fund requesting authorization for its then prevailing policies, and for a change in one of its rates. Mr. de Largentaye (France) questioned whether a member that had no par value but had a fluctuating rate needed to secure the Fund’s approval for changes in that rate. He did not accept that the Fund was to be the judge of whether circumstances required a change in the fluctuating rate. If it had been intended by the drafters of the Articles that the Fund’s approval was required, Article XIV would have so specified. Accordingly, he abstained from any decision on the change in the Thai rate. The decision taken was in fact quite limited. In view of the lack of information at the time on Thailand’s exchange system, the Fund did not object to the continuance of Thailand’s existing policies, pending the return of the staff mission and consideration of its report.

Continued deferment of a par value for Peru

At the end of 1950, the Board took up a first request from Peru to defer establishing a new par value. The country’s two fluctuating rates had been close together and had remained stable, but in the Government’s view a new parity could not yet be fixed. The Board discussed setting a time limit within which Peru must take action on a par value. Mr. Southard questioned the desirability of this. He thought that the Fund should not take an inflexible attitude toward such cases, particularly since most members seemed convinced of the advantages of stable exchange rates and the use of fluctuating rates did not seem especially contagious. The most important consideration was to find amicable ways of genuine consultation with members having such arrangements, so that the Fund could periodically review with them the feasibility of establishing a formal parity. Mr. Crick (United Kingdom) similarly questioned a proposed six-month deadline. Hence, the decision was merely to the effect that the Fund would from time to time review with Peru its exchange situation, with a view to determining whether action respecting a new parity would be feasible.

In 1960, by combining its two markets, Peru introduced a unitary fluctuating rate. This rate was still in existence on December 31, 1965, having remained, or been held, stable at 26.8 soles per U.S. dollar. In the intervening years, the Fund had permitted Peru to draw on the Fund’s resources and to assume the obligations of Article VIII—that is, to be regarded as a country which has a convertible currency.1


Why Canada turned to a fluctuating rate

A further example of the Fund’s acceptance or toleration of a fluctuating rate in exceptional circumstances is afforded by Canada. At the end of September 1950, Mr. Rasminsky informed the Fund that Canada had decided for the time being to suspend the par value of the Canadian dollar and to allow its foreign exchange value to fluctuate in response to market forces. The main purpose was to restrain a heavy inflow of capital, especially of speculative capital and mainly from the United States. This inflow, which had become especially great in 1950, was adding to the money supply and tending to depress interest rates, thus augmenting inflation. The object of the Canadian proposal was thus in contrast to most other exchange rate adjustments, which are intended to rectify an unfavorable balance of trade and to check an outflow of capital. In the view of the Canadian Government, it was impossible to determine in advance with any reasonable assurance what new rate level would be appropriate.

When the Canadian proposal was received, the staff recommended that one or both of two alternatives to the contemplated fluctuating rate should be considered: (1) the absorption of additional reserves by open market operations; (2) the introduction of additional controls to restrain the inflow of capital. In the staff view, a basic principle of the Fund was that exchange rates should not be used as an instrument of domestic monetary policy. The rationale of the Canadian proposal was the hope that a small appreciation of the exchange rate would have a minor effect on the current payments position but would have an important restraining effect on the capital inflow; the absorption of the residual dollar inflow could be handled by open market operations.

The staff, however, believed that it was unlikely that small changes in the exchange rate would have important effects on capital flows. Continued fluctuation of the rate would encourage and even justify the expectations of speculators. The Canadian problem was a fundamental one: capital movements had recurrently led to pressures on the exchange rate. The Canadians should act on the assumption that the world-wide instability leading to pressures on its exchange rate would continue and might even grow worse.

Mr. Rasminsky explained to the Directors that the suggestions made by the staff had already been submitted to the Finance Minister for consideration and had been referred to the Canadian Cabinet. The Canadian authorities, however, had rejected these alternatives as inadequate. As for open market operations, the Government could not take on additional debt of an unpredictable amount. And there were technical and political difficulties in controlling capital movements: direct controls would impose a tremendous, if not impossible, administrative burden; they might create hostility toward American capital; and, in addition, the Canadian Government was opposed to exchange controls of any kind.

The Fund’s immediate response

The decision which the Executive Directors took at the time was that the Fund, unable under its Articles of Agreement to approve a fluctuating rate, recognized the exigencies of the situation and took note of the intention of the Canadian Government to remain in consultation with the Fund and to re-establish an effective par value as soon as circumstances warranted. This decision reflected the position of the majority of Directors that Canada should, in the light of its special circumstances, be permitted to try a fluctuating exchange rate for a short period. If, at the end of that period, the Canadian Government had to conclude that the alternatives were either to declare a new par value at parity with the U.S. dollar or to restore the old par value and impose effective capital controls, at least there was a better prospect of reaching general agreement that no other alternatives existed. At the same time, the language of the decision was not as forthright as in the Belgian case in 1949; it made no reference to “an appropriate step.”

A few Directors again commented that possibly the Articles had been somewhat deficient: they made no provision for the case of a country which could not immediately propose a new par value. It was the view of Mr. Beyen (Netherlands), for example, that where a country could not propose a new parity, the Board could not ask the government to take the responsibility of doing so; the Fund had to give the member the benefit of the doubt in instances of this kind, since the responsibility for the action was the government’s.

The Chairman, however, stressed that the position taken by Mr. Beyen called for very definite qualification because, if too widely interpreted, it might amount to a negation of one of the main principles of the Fund. He could not agree that the Fund was obliged to accept any proposal submitted by a member, on the theory that the responsibility for the proposed action and its effect were the government’s alone. The Fund also had a responsibility to all its members, and each member in turn had a responsibility to other members.

Debates over the Canadian rate

For the next decade, until Canada returned to the par value system, there were intensive discussions among the staff and in the Board about the Canadian rate. The discussions in the Board began only a few months after the introduction of the fluctuating rate. In April 1951, in the course of a review by the Board of the Canadian economy, Mr. Rasminsky defended the fluctuating rate. He stressed that the excessive capital inflow had been reduced and import and exchange restrictions had been liberalized. Mr. de Largentaye insisted, however, that the Canadian Government could have used alternative arrangements to deal with the capital inflow. Mr. Melville (Australia) believed that the Canadian example, and the fact that the Fund had not objected, had made it difficult for the Fund to refuse other requests for fluctuating rates.

In another review in February 1952, several Directors inquired what economic factors, if any, prevented Canada from establishing a par value and whether the fluctuating rate had not had greater disadvantages than a fixed rate even for a member in Canada’s circumstances. Mr. Rasminsky replied that the Canadian experience had been that a fixed rate of exchange produced wide swings in capital movements: foreign capital came in when the fixed rate looked too low and tended to leave when the rate looked high. The Canadian Government believed that, where foreign capital movements were large, a fluctuating rate controlled them better than a fixed rate. A fluctuating rate discouraged oscillations in capital flows because of its own fluctuations. Moreover, the size of the capital flows was not enhanced by further speculative capital movements in anticipation of a change in a pegged exchange rate.

Several Directors were concerned that the Canadian problem was not a temporary one, and that the fluctuating rate, also, was not temporary. Therefore, Mr. Stamp (United Kingdom) argued that the Fund should review the general policy at issue, with emphasis on the possible result if the practice of exchange rate fluctuation spread more widely.

Mr. Rasminsky argued, in rebuttal, that Canada’s current circumstances were abnormal. Moreover, Canada’s adaptation to these circumstances had contributed to its attainment of two of the three main Fund objectives—the elimination of exchange restrictions and the establishment of a convertible currency with no discrimination. Some Directors supported his position. They believed that the world’s exchange situation had been abnormal for some time and in such circumstances fluctuating exchange rates were of great help to countries, particularly in permitting the removal of restrictions.

By 1956 the Fund had come to regard Canada’s relative success with a fluctuating rate as reflecting the uniqueness of that country’s circumstances. Canada had a trade deficit with a large capital inflow. There was confidence in the Canadian dollar because of the fiscal and credit policies being followed; Canada was relatively free of restrictions and had a convertible currency. Moreover, the institutional background led many to regard as natural a parity for the Canadian dollar somewhere near that of the U.S. dollar. Close interdependence between short-term capital movements and movements of the exchange rate had caused capital flows on the whole to be equilibrating rather than disturbing. Finally, the exchange rate fluctuated by only about 3–5 per cent, despite the absence of intervention by the authorities except to maintain an orderly exchange market.

For all these reasons, Canadian trade and normal capital transactions had not lost the important benefits commonly associated with exchange rate stability. The Canadian example was not a precedent, for the circumstances of other countries were quite different.

At the crossroads

Thus, from 1950 until about 1956, Canada’s fluctuating rate had achieved the primary objectives for which it had been established. Massive inflows of long-term capital had continued and had provided net additional savings with which Canada had been able to finance a high rate of investment and growth. Yet the potentially disequilibrating effects of these long-term capital inflows on the balance of payments had been sufficiently offset by current account balances and compensating short-term capital movements. Exchange stability had been preserved. Short-term capital movements appeared to have responded much more to changes in the exchange rate than to relative interest rates in Canada and the United States.

After 1957, however, Canada began to encounter several difficulties. A very large current account deficit emerged and persisted; at 3–4 per cent of the gross national product, it was among the largest in the world. The rate of growth of the economy had slowed down measurably, and unemployment was high—7½ per cent by mid-1961. By then disenchantment with the way the Canadian rate had been operating since the late 1950’s was discernible both among Canadian authorities and in the Fund.

The Canadian authorities attributed the country’s economic reverses in large part to the now unduly appreciated level of the exchange rate; the inflow of capital had caused the rate to reach a point at which it acted as a stimulus to imports and as some deterrent to exports. In presenting his budget to Parliament on June 20, 1961, the Minister of Finance announced several new policies, including a switch in exchange rate policy. Whereas the Government previously had had a neutral attitude toward the level of the exchange rate, there was now to be some direct official intervention in the exchange market. Canada was thus to have a managed flexible rate.

Describing the Canadian policy to the Board in the middle of 1961, Mr. Rasminsky recognized that the shift in the exchange rate policy of the Canadian Government might create some special problems for the Fund. Other Fund members might well have anxiety about the degree to which the Canadian Government would devalue the Canadian dollar, but would not have an opportunity to sanction, or object to, the rate that emerged. Mr. Rasminsky sought to reassure the other Fund members by stating that Canada did not intend to operate its rate policies so as to cause competitive depreciation. The intention of the Canadian Government was not to determine the level of the exchange rate in the market but rather to reduce the inflow of capital and the size of the current account deficit.

The assessment of the Canadian exchange rate situation which the Fund staff made differed from that of the Canadian authorities. The staff agreed with the authorities that the period prior to 1956–57 had shown the usefulness of the fluctuating rate, and that the post-1956 experiences had been more disturbing than equilibrating. But the explanation offered by the staff for what had gone wrong since 1956–57 differed from that of the authorities.

In the staff view, the primary cause of the unsatisfactory results of the fluctuating exchange rate after the first few years was not any change in the mechanism of the Canadian balance of payments, but the effect of governmental policy on that mechanism. Governmental policy had resulted in the emergence of a wide differential between Canadian interest rates and those of the United States. A spread in which interest rates in Canada were higher than those in the United States had attracted liquid funds into Canada. The inflow of these funds kept the exchange rate appreciated during a period when basic trade and long-term capital flows that were normally motivated would have tended to push it downward.

The staff pointed out that after 1956–57 the size and composition of the long-term capital inflow into Canada had changed markedly. While direct investment inflows had continued on a more or less steady and generally upward course, they had been supplemented by huge and erratic movements of portfolio capital. Various types of capital flowed in unevenly at different times. Many of these flows seemed to have been highly sensitive to divergences between monetary and credit conditions in Canada and the United States. By the late 1950’s capital movements which were induced primarily, or solely, by interest differentials were much larger than they had been in the early 1950’s. Another factor which had made the fluctuating rate increasingly difficult to handle by 1960 had been the mounting uncertainties as to the prospects for the Canadian economy and as to the Government’s economic policies. The staff concluded, therefore, that “Canada should re-establish an effective par value as soon as circumstances permit.”

The foregoing report by the staff was before the Board when, in February 1962, it undertook the first consultation with Canada under Article VIII. At this consultation, the uneasiness of most of the Executive Directors concerning the continuance of the fluctuating rate was readily manifest. The Board did not press for an immediate re-establishment of a par value. But many Directors questioned whether Canada still had a plausible case for continued noncompliance with the Fund Agreement. What economic situation would enable a par value to be restored? Were not Canada’s special circumstances, in effect, permanent? Did not Canada’s large reserves seem sufficiently ample to enable the authorities to defend a fixed rate? How should the Fund assess the declared intention of the Canadian authorities to bring the rate down to a lower level? How far would the rate actually be depreciated? How could depreciation be achieved without endangering exchange rate stability, if large fiscal deficits and low interest rates were permitted?

The Directors observed that Canada presented a rare example of an exchange depreciation being undertaken largely to stimulate the growth of national income rather than to influence the current balance of payments. They also expressed considerable concern over the possibility that Canada’s internal policies were too expansionary.

From Canada’s experiences the Directors drew conclusions regarding fluctuating rates in general. Had not Canada’s experience demonstrated that it was difficult and potentially harmful for a country to conduct a monetary policy isolated from that of other countries? Even a flexible exchange rate had not given Canada freedom in its internal monetary policy. Had not another danger of flexible rates also been revealed—that feeling safe under the shelter of a flexible rate and deprived of the symptoms of movements in reserves, the authorities tended to disregard the external repercussions of their domestic policies? It was significantly noted that Mexico, like Canada a neighbor of the United States and a recipient of heavy inflows of U.S. capital, had also found a fluctuating rate unworkable.

Several Directors, therefore, supported the conclusions of the staff that Canada should re-establish an effective par value as soon as circumstances permitted. Some of the Directors were fairly specific that the new par value should be introduced as soon as the Canadian Government’s policies were clearly established and the authorities had reached a conclusion as to the appropriate level for the new parity.

On May 2, 1962, in the midst of exchange difficulties, Canada gave up its fluctuating rate and proposed a new par value to the Fund. The authorities stated that the new par value was necessary to correct a fundamental disequilibrium. Moreover, they expressed their awareness, deriving from the discussions at the meetings of the Board in July 1961 and February 1962, of the widespread international concern at Canada’s deviation from the established exchange system.

Having decided to re-establish a par value, the Canadian authorities faced the question of selecting a rate. Mr. Rasminsky explained that a par value at the current rate of 95 U.S. cents per Canadian dollar was thought to be too high to eliminate uncertainty in the market; the Government had already lost substantial reserves in defending that rate. Alternatively, a rate of 90 U.S. cents per Canadian dollar was rejected as being lower than the Canadian economy required. Furthermore, Canada could not count on international acceptance of such a depreciated rate. Consequently, the authorities proposed a new rate of 92.5 U.S. cents per Canadian dollar.

The Directors, agreeing with the new par value, welcomed the return of Canada to the par value system.


With the acceptance by the Board during the 1950’s of Canada’s fluctuating rate, the formal objections of the Fund to fluctuating rates had quietly come to an end. But yet another hurdle had had, in the meantime, to be surmounted. This concerned the valuation of the Fund’s holdings of currencies of countries where fluctuating rates prevailed. Article IV, Section 8, of the Agreement requires countries to maintain the gold value of the Fund’s assets. Hence, in the event of a depreciation in its par value, a member has to pay additional local currency to the Fund. Was there not a need for such payments to be made by countries without par values and with fluctuating rates? If payments were required, what rate of exchange should be applied? The staff pointed out that a similar difficulty covering the exchange rate to use for computations would arise if another member wished to draw from the Fund the currency of the country with the fluctuating rate.

These questions became a matter for consideration and decision by the Board in 1954. At that time members with fluctuating rates included not only Canada but also Lebanon, Peru, Syria, and Thailand. Even if only for the purposes of the auditors, some rules were essential. In addition, it was imperative to preserve the gold value of the Fund’s assets. And it was desirable to ensure that the Fund was not used in lieu of the outside market for short-term arbitrage gains.

With little debate the Board decided that for currencies with a single fluctuating rate, computations were to be based on the mid-point between the highest and lowest exchange rates for the U.S. dollar for cable transfers for spot delivery in the main financial center of the country of the fluctuating currency on specified days. If a mid-point could not be determined in the main financial center of the country of the fluctuating currency, rates quoted in New York could be substituted.2

The Directors had, however, some difficulty in deciding upon the scope of application of these rules. How should they be applied to currencies with multiple exchange rates, some of which might fluctuate? The Directors from Latin America stressed that it was vital that the Fund should not embarrass a country which was struggling to maintain a par value. Such embarrassment might result from the Fund’s computing an effective rate other than the par value for use in Fund transactions in the currency of that country. Hence, a preamble to the decision on computed exchange rates explicitly specified that multiple rate systems were not to be covered by these special rules, and that the Fund would not determine computed rates except where there was a practical interest for the Fund or its members to do so.

These rules have the practical purpose of facilitating the operations of the Fund. They make it possible for the Fund to engage in transactions in the currencies of members with fluctuating exchange rates on an equitable basis and to make the computations required by the Fund Agreement—that is, to facilitate the periodic revaluations of the Fund’s holdings of fluctuating currencies, as well as their valuation for the purpose of actual transactions. Other members of the Fund are thus not precluded by the fluctuations in the rate for a member’s currency from purchasing it from the Fund, nor is a member whose currency fluctuates necessarily deprived of its right to purchase the currencies of other members. These rules become operative in any given case only after the Fund has decided to apply Article IV, Section 8, to its holdings of a fluctuating currency.

In July 1954, this general decision on computed rates for fluctuating currencies was made applicable to Canada and Peru. Canada, of course, had had a single fluctuating rate since September 20, 1950, and this rate was used as the basis for the rate computed for Fund transactions.

Peru had had two exchange markets since November 1949; the par value of 6.50 soles per U.S. dollar no longer applied to any transactions. The exchange rates in the two markets fluctuated freely and there was only a small spread between the rates. The Fund’s holdings of Peruvian soles had been adjusted under Article IV, Section 8 (b) (ii), in April 1951. This adjustment had been made on the basis of a rate of 15 soles per U.S. dollar, a rate which had been proposed by Peru and accepted by the Board “subject to further adjustment in the event of a significant change in the foreign exchange value of the currency or when a new par value for the sol is established in agreement with the Fund.” The rate in the certificate market—more important than the free market as measured by the volume of transactions—had been adopted as the rate for purposes of a stand-by arrangement with the Fund approved in February 1954. Paragraph 2 of the stand-by arrangement had specified that “currencies drawn from the Fund shall be used only for the support of the certificate market rate.” Paragraph 19 of the arrangement had specified that the rate of exchange to be used for drawings and repurchases would be the rate determined from time to time under Article IV, Section 8. In computing this rate for the currency of Peru, the mid-point stated in the general rules was to be that of the rate in the certificate market.

Computations under the general decision have been made only for countries where no transactions any longer take place at the par values. At the end of 1965, for example, computed effective rates applied only to Argentina, Bolivia, Brazil, Chile, Colombia, Paraguay, and Peru.3

The question of a country’s formal status in the Fund if it had no par value, while quiescent, was not entirely dead in 1954. Discussing the application to Canada of the decision on computed rates, Mr. Saad (Egypt) said that he had serious difficulty with the Fund’s adopting rules which in a sense helped to perpetuate a member’s technical violation of the Articles. However, the decision passed without objection by other Directors.

Subsequently, two technical amendments were made to the decision on the determination of rates for transactions in currencies of countries with fluctuating rates.4 The first was introduced in August 1961 when the Fund decided that it should sell gold to replenish its holdings of certain currencies. Because the currencies of some countries with fluctuating rates were involved—for example, the Canadian dollar—the earlier decision on transactions and computations involving fluctuating currencies was amended to cover such sales of gold by the Fund. The second amendment was made in December 1961 when the Board took a decision on the General Arrangements to Borrow. At that time, the decision on computed rates for fluctuating currencies was amended to cover such borrowing and the repayment of borrowing in fluctuating currencies.


The gradually evolving policies in connection with the fluctuating rates of Mexico, Peru, and Canada illustrate one major direction in which the Fund’s attitude toward fixed rates was not immutable. In these instances, the Fund, in effect, tolerated fluctuating rates; later, it went somewhat further.

Beginning in 1956, the Fund has supported programs which, in conjunction with exchange reforms and stabilization plans, have included a fluctuating rate. To some extent Peru’s fluctuating rate, described above, had been introduced as part of an exchange reform. But even before reform, Peru’s exchange system had been fairly simple, inflation had not been massive, and the devaluation involved was modest. In the late 1950’s, however, the technique of using a fluctuating exchange rate to effect reform of a complex exchange system became widely used.

The circumstances

Many of the less developed countries in which there has been prolonged inflation have found it necessary to change their exchange rates frequently. Several of these countries, for various reasons, have instituted fluctuating rates. Countries undertaking stabilization of their economies are frequently uncertain about the effects of the internal measures being adopted. Future movements in prices and wages have been difficult to estimate. In these circumstances, countries have been unable to determine in advance an appropriate level for the exchange rate. Where a combination of restrictions and multiple rates exist, it is difficult to ascertain even the average effective exchange rate being applied. Determination of an equilibrium rate is virtually impossible. In still other instances, a new par value cannot be determined until after a new tariff system has been established. For these reasons, as exchange reforms have been undertaken, several countries have instituted single fluctuating rates.

The changing Fund policy

Bolivia was among the first, in 1956, to effect exchange reform via a fluctuating rate. Its example was soon followed by Chile, Paraguay, and Argentina. At first, the Fund merely agreed to these fluctuating rates. But as more instances occurred, it was clear that the Fund had, in effect, adopted a general policy of permitting fluctuating exchange rates as a temporary instrument of exchange depreciation and liberalization of restrictions.

This policy was made manifest in the Board’s reaction to a proposal put forward by the Philippines in 1962. At this time the Philippines, which had had a fixed exchange rate, introduced a fluctuating one. When presenting to the Board the Philippine proposal for exchange reform and request for a drawing from the Fund, Mr. Saad (who had been elected by the Philippines) emphasized that the proposed fluctuating rate was clearly in line with the Fund’s policy. The Government of the Philippines was resorting to a fluctuating rate just as had other countries as a means of eliminating restrictions on payments and of finding a realistic level for the exchange rate. The proposed change in the exchange rate represented the culmination of a program, which the Philippines had been following for the previous two years, of steadily simplifying its exchange system and of reducing reliance on quantitative restrictions. The Philippine authorities regarded the fluctuating rate as temporary and looked forward to the establishment of an appropriate new par value.

The Directors agreed that the proposed exchange reform, except for a few transitional features, followed a plan with which they were already familiar. They would have preferred that the Philippine authorities move to a single fixed rate system, but they recognized the circumstances which necessitated delay in introducing a new par value. These circumstances included the persistence of a great deal of inflationary pressure, a change in government administration, and a need to revise the tariff. The Managing Director, Mr. Jacobsson, while reasserting that a floating rate was not, in general, desirable, emphasized the difficulty which a country with slender reserves experienced in adopting a fixed rate without imposing exchange restrictions. Were restrictions to be instituted, a black market would inevitably emerge.

Some Directors questioned how capital movements, both in and out of the Philippines, would react to a fluctuating exchange rate. The effect of the new rate on capital was important, especially because capital outflows had been partly responsible for the deterioration in the balance of payments. These Directors hoped that the proposed exchange reform would help to curb capital flight and even to stimulate capital repatriation. They favored a fixed rate, which in their view was more likely to make capital movements normal. Therefore, they placed a great deal of emphasis on the intention of the Philippine authorities to use the fluctuating rate only temporarily.

Experiences with temporary fluctuating rates have, in general, proved quite successful. Several such rates were eventually stabilized and furnished the basis for new par values. Indeed, by 1962–63 the Fund had even begun to initiate suggestions for fluctuating rates as temporary ways of altering exchange rates and of achieving exchange reforms. Several countries were advised—even urged—to take this road to depreciation and exchange reform. Where countries introduced greater rate flexibility, the Fund welcomed the action.

Appropriate conditions for a fluctuating rate

The questions that have come up as the Fund has considered individual fluctuating rates reveal the difficulties of deciding the circumstances in which fluctuating rates are appropriate. Many of these questions have concerned reserves and reserve policy. What is the level of reserves required before an exchange rate can be maintained? When should the central bank enter the market in order to accumulate reserves? Do not central bank purchases in the market to repay short-term liabilities differ from purchases for reserves? (The former may be treated as a genuine market force, while the demand for foreign exchange for reserve accumulation represents a policy decision of the authorities rather than a market force.) May it not be best to accumulate reserves by taking advantage of special occurrences in the exchange market rather than by exerting a continuous downward pressure on the rate?

Another group of questions has concerned the reactions of capital movements to a fluctuating exchange rate. For example, is a fluctuating rate appropriate where underinvoicing and smuggling as well as an excessive outflow of capital have been a problem? The use of fluctuating rates in circumstances of continuous inflation has given rise to still other questions. For, if internal prices are not stabilized, flexible rates are necessary to prevent overvaluation. Yet under conditions of unstable internal prices, the rate depreciates regularly; this depreciation in turn contributes to continued instability of internal prices, and also possibly induces speculative capital movements.

One situation in which a unitary fluctuating exchange rate has been found unworkable is that in which a particular exchange rate is required for a given commodity. When Colombia, for example, attempted to introduce a fluctuating rate in 1957–58, it was found to be inappropriate because of the need for a special exchange rate for coffee exports.

Pegging of fluctuating rates

The Fund’s view has been that, once a fluctuating rate is introduced, the authorities should let the rate change freely in response to market forces, intervening only to maintain orderly market conditions. The Board has, therefore, often warned against pegging of fluctuating rates. Similarly, the Board has suggested that, where two markets exist, sales of exchange by the central bank in the free market should not result in a spread of fixed size between the official rate of exchange and that in the free market. The fear of the Board has been that intervention by the central bank could lead to an undue loss of foreign exchange reserves as well as establish an inappropriate exchange rate. Where the member has drawn on the Fund, the Board has been concerned lest the Fund’s resources be used to peg a potentially unrealistic exchange rate.

In some countries where a flexible exchange rate policy is considered essential, certain “balance of payments tests” have been developed and have become part of stabilization programs.5 The object of these tests is to ensure that an exchange rate will be maintained which conforms to the basic trends in the economy. A level at which the country’s foreign exchange reserves are to be maintained during a stated period is prescribed. The country undertakes that if there is a danger that reserves will fall below this level, it will take appropriate action, for example, by allowing the exchange rate to fall. Even so, since most countries have been reluctant to see their exchange rates depreciate excessively, the Fund has frequently had to call upon countries with flexible rates not to stabilize prematurely.


Nonetheless, the Fund’s ultimate objective in cases of exchange reform has been to create the conditions for the restoration of a stable and unified exchange rate. The fluctuating rate has been regarded as a temporary means to an end. The Fund has also continued to regard a general system of fixed rates and institutionally agreed par values as decidedly superior to a system of fluctuating rates.

As early as 1951 the Fund gave several reasons for its belief that a system of fluctuating rates would not be a satisfactory alternative to the par value system. To those who advocated allowing rates to find their “natural” level it replied that there was no such thing as a natural level for the rate of exchange of a currency. The “natural” rate differed in each case, as it was dependent upon the economic, financial, and monetary policies followed by the country concerned, and by other countries with whom it had important economic relationships. In addition, whether a given exchange rate was at the correct level could be determined only after there had been time to observe the course of the balance of payments in response to that rate.

Furthermore, implicit in the arguments for fluctuating rates was the assumption that some major currency would remain stable as a point of reference. The limited number of countries that had used flexible rates had done so within the framework of fixed parities for the currencies of well-nigh all the industrial countries. Chaos would result if a substantial number of rates were allowed to fluctuate.6

By 1958 the Fund’s views were based on its widening experiences. As in the 1930’s, so in the 1950’s, fluctuating rates had been found to be inappropriate where outflows of capital were excessive or underinvoicing and smuggling were troublesome. Fluctuating rates often induced additional capital inflows which were prompted by speculation on exchange rate movements. It had also become clear that countries preferred to make adjustments in their exchange rates in some manner other than through fluctuating rates, so as to minimize speculation. Most countries seemed to recognize, in practice, the importance of orderly exchange arrangements and exchange stability. Even in countries where the authorities were not prepared formally to stabilize the exchange rate on the basis of a realistic parity, de facto stable rates, the Board observed, were often maintained for long periods of time.7

In its Annual Report for 1962 further examples were cited of the Fund’s postwar experience with fluctuating exchange rates.8 The conclusion was that a short period of fluctuation might be a means of reaching a rate which could subsequently be maintained. But in other circumstances fluctuating rates gave rise to difficulties. In exceptional circumstances, such as the large capital inflows into Canada in the early 1950’s and into Peru for brief periods, a fluctuating rate might have an upward tendency. The more general experience, however, was that if a country did not clearly and quickly adopt a monetary policy aimed at stability, the movements of its fluctuating rate were likely to be oscillations not around a stable value but around a declining trend.

If an exchange rate fluctuated widely, it might be expected to depreciate over time. Any circumstance leading to a temporary depreciation of the rate raised the domestic currency cost of imports and directly and indirectly led to price increases. These increases encouraged demands for higher money wages, at least some of which were met. Hence, depreciations were likely to induce increases in domestic costs. On the other hand, institutional rigidities limited reductions in money wages, so that exchange appreciations were unlikely to lead to any significant lowering of domestic costs. Therefore, a fluctuating rate might be expected to encourage a rising trend in domestic costs, which in turn put downward pressure on the rate.

Pegging of a fluctuating rate had still other disadvantages. A fixed rate subject to frequent changes, or a temporarily pegged fluctuating rate, was likely to encourage destabilizing speculation which would create even more serious problems than those which arose under a system of freely fluctuating rates.

It was concluded therefore that it was an illusion to expect a fluctuating rate to ease the problems facing monetary authorities. On the contrary, by eliminating the rallying point of the defense of a fixed par value, a fluctuating rate made it necessary for the authorities to exercise greater caution in determining monetary policy.

In 1964 the Deputy Managing Director summarized the Fund’s position as follows:

Some economists argue the case for freely fluctuating rates of exchange as an instrument for balance of payments adjustment. But financial officials in most countries today—even in those developing countries which temporarily have fluctuating rates—have been convinced by experience that greater flexibility than that sanctioned by the Fund’s Articles of Agreement would be undesirable and impractical. Fluctuating rates create great uncertainty for traders and investors and set up stresses in the financial and economic relationships within a country and in its international position. Moreover, far from being simple, the problems of managing a flexible rate are no less complex than those which arise in maintaining an effective par value, and in addition present their own mixture of financial, economic, and political difficulties. Indeed, countries which have endeavored to conduct their affairs on the basis of a fluctuating rate have found it extremely hard to let the exchange rate perform its intended functions. It also is generally true that exchange rate policy under a system of flexible rates tends to become much more a matter of unilateral action, which would make the whole process of international collaboration in financial matters far more difficult.9

Thus, while the Fund’s policy as regards fluctuating rates became more flexible, the Fund continued, strongly, to favor a system of fixed exchange rates. Nonetheless, events on the horizon forewarned that the debates concerning the merits of fluctuating rates would, shortly, be intensified. Among the suggestions being made in the mid-1960’s for ways to improve the system of par values was that fluctuating rates, at least for major currencies, should be substituted for par values. When, in 1964 and 1965, one large industrial country after another began to be plagued by balance of payments stresses for which there seemed no adequate solutions, the advocates of fluctuating rates were able to muster additional adherents.

Changes in fixed rates seemed ruled out because the authorities of the major countries concerned found it difficult politically to alter the exchange values of their currencies. Further, in the absence of adequate world gold stocks, most countries had come to hold sizable amounts of the currencies of the major countries as foreign exchange reserves. Therefore, among other effects of exchange rate alteration, the authorities in charge of the currencies used as reserves were concerned with the implications of changes in exchange rates for the reserve holdings of all trading nations. Tighter monetary and fiscal policies had already been introduced in most of the large industrial nations; the imposition of still more stringent internal measures to alleviate or reverse external deficits was considered contrary to the domestic objectives to which these countries were committed. Stricter credit or tax measures might jeopardize employment; reduced public expenditure might mean that urgent domestic programs would have to be postponed or foregone. Some countries had reintroduced, or enacted for the first time, controls on capital movements, but they shunned controls on current transactions.

All these difficulties with alternative solutions strengthened the arguments for enhanced flexibility of exchange rates. Nevertheless, the monetary authorities of all the large industrial nations, and especially the central bankers, as well as the officials of the Fund, continued to view as untenable any general system of fluctuating rates.


See p. 76 above.


E.B. Decision No. 321-(54/32), June 15, 1954; below, Vol. III, p. 222.


Schedule of Par Values, February 15, 1966.


E.B. Decision No. 1245-(61/45), August 4, 1961, and E.B. Decision No. 1283-(61/56), December 20, 1961; below, Vol. III, p. 224.


See below, pp. 506–10.


Annual Report, 1951, pp. 36–41.


Annual Report, 1958, p. 20.


Annual Report, 1962, pp. 62–67.


Speech by Frank A. Southard, Jr., Dallas, Texas, March 27, 1964, in International Financial News Survey, April 3, 1964, pp. 115–16.

    Other Resources Citing This Publication