8 Selection and Calculation of Exchange Rates Under International Arrangements

Abstract

Two separate problems of choice have arisen, or have become intensified, in an environment of fluctuating exchange rates, and have led, therefore, to new legal rules or to modifications in legal rules formerly applicable. One kind of problem is not the choice of a currency, because in the circumstances choice of the appropriate currency is settled, but determination of the appropriate exchange rate for the currency if translation of it into another currency is necessary. Problems of this kind, and the solutions of them, are discussed in this chapter.

Problems of Choice

Two separate problems of choice have arisen, or have become intensified, in an environment of fluctuating exchange rates, and have led, therefore, to new legal rules or to modifications in legal rules formerly applicable. One kind of problem is not the choice of a currency, because in the circumstances choice of the appropriate currency is settled, but determination of the appropriate exchange rate for the currency if translation of it into another currency is necessary. Problems of this kind, and the solutions of them, are discussed in this chapter.

The other kind of problem is the choice of the appropriate currency for some purpose when two or more currencies are involved in the circumstances of the case. Of course, once the choice is made among the currencies involved, the problem of selecting the appropriate exchange rate arises if it is necessary to translate that currency into another currency. Examples of the second category of problems are discussed in Chapter 12 and in other chapters.

Some of the problems discussed in this monograph did not arise, or were less troublesome if they did arise, in the period when the par value system was in force. As there were fixed relationships among currencies—although subject to changes in the relationships and to the margins prescribed by the Articles for exchange transactions—much less turned on the choice of the appropriate currency or of the appropriate exchange rate. In the discretionary system of exchange arrangements, it often becomes necessary to specify that parties may choose the exchange rate that is to apply to their dealings or that they are not free to do so. If the parties are not free, the applicable exchange rate must be prescribed. Issues of freedom or the absence of freedom to choose the applicable exchange rate can be settled by national law or by treaty.

An example of freedom for parties recognized by national law can be cited from the Report of the English Law Commission. The Commission concluded that parties were free, and should continue to be free, to agree that the translation of any currency into another currency that becomes necessary in their dealings shall be made at the exchange rate prevailing at a particular date or to name the exchange rate that is to be applied.1

IMF’s Former Practice

Throughout its history, the IMF has had to select the exchange rate to apply in conducting its transactions and operations and has had to decide how to calculate the specified exchange rate for a particular kind of transaction or operation. The IMF has followed four different procedures in response to developments in the international monetary system.

Initially, of course, the IMF had to have a procedure when its financial activities involved currencies for which there were effective par values in accordance with the Articles. The procedure was then simple: the IMF bought and sold currencies on the basis of par values, which meant, in effect, at the parities between the two currencies directly involved when the IMF sold one member’s currency to another member and purchased from the latter member an equivalent amount of its currency. (The principle was qualified only to the extent of a modest service charge levied by the IMF for the transaction.) The IMF continued this practice even after the IMF authorized expansion of the margins to 2 percent for many exchange transactions in the territories of members.

The effect of the practice was that the IMF ignored fluctuations in exchange rates in the markets within the narrow 1 percent or the later 2 percent margins above and below parity. The procedure was simple to administer and it did not seem inequitable to members because the margins were not wide and were unlikely to give rise to serious loss when the currency purchased from the IMF was put into the market by the purchasing member. All that was necessary was a rule that the IMF had to instruct its depository to make the necessary transfer within three business days after the IMF’s receipt of an authenticated request for the purchase of a currency. At first, the value date in the IMF’s instruction was two business days ahead, but later the value date was extended to three business days, partly because differences in time meant that two business days in the United States allowed for only one business day in Europe.

The second procedure was adopted by a decision of the IMF in 1954 entitled “Transactions and Computations Involving Fluctuating Currencies.” They were the currencies for which an initial par value had not been established under the Articles or currencies for which the established par value was not being maintained in accordance with the Articles. The decision declared that it would not be applied to the IMF’s holdings of a fluctuating currency when there was no practical interest in doing so for the benefit of the IMF or members. (It will be seen that there is an echo of this reservation in current practice when only the currency of a single member is involved in a financial operation with the IMF, as, for example, when a member pays charges to, or receives remuneration from, the IMF in the member’s own currency.)

The IMF’s practice under its decision on fluctuating currencies was again a simple one. The heart of the decision was that computations in these currencies would be based on the midpoint between the highest and the lowest rate for the U.S. dollar quoted, for cable transfers for spot delivery, in the main financial center of the country of the fluctuating currency on the day specified in the decision. The choice of the specified day was related to the character of the transaction or operation. For the sale or purchase by the IMF of a fluctuating currency in exchange for another currency, the day specified in the decision was the last business day in the main financial center of the country of the fluctuating currency before the IMF instructed its depository to transfer or receive the fluctuating currency. The U.S. dollar was selected as the standard by which to determine the exchange rate of the fluctuating currency because of the assumption that the United States was maintaining the value of the dollar in relation to gold as the common denominator of the par value system. The assumption was based on the undertaking of the United States to buy and sell gold freely, for U.S. dollars, with the monetary authorities of other members for the settlement of international transactions.

The third development related to the SDR. To make the new monetary asset attractive, the IMF decided to take account of exchange rates in the markets to prevent the transferee of SDRs from providing currency to the transferor in these transactions at an exchange rate disadvantageous to the transferor. Some currencies would be at a premium above parity and others at a discount below parity within the margins that were consistent with the Articles. If a transferee of SDRs had been permitted to select the currency and to provide it at par, as the IMF did in buying and selling currencies in its transactions with members, the transferee might decide to provide a currency that was at a discount in the market.

To prevent the transferee from making this selection, the First Amendment established the principle of equal value. It has been preserved by the Second Amendment. The transferee of SDRs has to provide equal value for them no matter which member is the transferee and which currency the transferee provides to the transferor in exchange for the SDRs. Principles for determining rates of exchange in these transactions were incorporated in the Fund’s Rules and Regulations. The exchange rate in terms of the SDR if the U.S. dollar was provided in a transaction between members was equal to the par value of the U.S. dollar. If a different currency was provided from among those named by the IMF as the currencies for this purpose, the appropriate exchange rate was the “representative rate” of the currency if the IMF found that such a rate for spot delivery of the U.S. dollar could be readily ascertained in the member’s market. If the IMF found that a representative rate between the U.S. dollar and Patria’s currency (the currency provided) could not be readily ascertained in Patria’s exchange market, but the IMF found that a representative rate could be readily ascertained, in accordance with the rules, for the currency of Terra in Patria’s market, the cross rate was applied. The rate for the currency of Patria to be provided in exchange for SDRs was calculated by reference to the rate between the currency of Patria and the currency of Terra in Patria’s market and the rate between the currency of Terra and the U.S. dollar in Terra’s market. In all cases in which these and associated rules did not suffice for determining the exchange rate for a currency in the transactions discussed here, the IMF would determine the rate.2

The fourth, and so far the latest, development is the discretionary system of exchange arrangements that followed the collapse of the par value system and the validation of the discretionary system by the Second Amendment. The IMF’s current procedures are summarized in the next section. Current procedures differ from the practices inspired by the second and third developments but draw heavily on those earlier practices.

IMF’s Present Practice

The basic principles for the transactions and operations of the IMF and for maintenance of value obligations under the Second Amendment were adopted by a decision of the IMF on December 5, 1977, which came into operation on April 1, 1978, the date on which the Second Amendment became effective.3 The decision was designed to be appropriate for the circumstances of the discretionary system of exchange arrangements under the Second Amendment. The principles of the decision are that for computations by the IMF relating to a member’s currency held in the General Resources Account,

  • (a) the exchange rate on the occasion of the use of the currency in a transaction or operation between the IMF and another member is the rate as of three business days before the value date of the transaction or operation, but, if this rate cannot be used for some legitimate reason, the rate of the closest preceding day that is practicable is to be applied;

  • (b) the exchange rate on all other occasions is the rate on the basis of which the IMF is holding (accounting for) the currency.

The three-business-day rule referred to in (a) above may seem antiquated in view of the speed with which modern technology makes it possible to conduct transactions and operations. Central banks, however, have shown some preference for this rule, because, to cite one reason, it gives a central bank time to manage its investments and to disinvest in order to exchange balances of its currency for a “freely usable currency” on the request of the purchasing member when the balances have been sold by the IMF and the currency sold is not “freely usable” according to the Articles.4

The three-business-day rule applies to the transfer of SDRs to a member designated by the IMF to receive the SDRs and provide currency. The exchange rate is the one determined in accordance with the decision described above, namely, the rate that prevailed three business days before the value date for the transaction. The exchange rate three business days before the value date may be different from the exchange rate on the value date. To enhance the qualities of the SDR, a different rule applies to transfers of SDRs by agreement between two members. The exchange rate in these transactions is the one prevailing at the date of the agreement. The parties can agree to carry out the transaction on that day or on any business day within three business days from the date of the agreement. Under such an agreement, the exchange rate can be a more recent one than under the three-business-day rule.5

“Business days” for the purpose of the rule means business days of the IMF. However, to ensure that the IMF’s notice of a transaction is adequate, the IMF takes account of the business days in the member countries involved in a transaction. The IMF seeks, for example, to avoid designating a day that is not a business day in such a country.

In accordance with the decision, the exchange rate for the currency of a member is determined each time that the currency is used in a transaction or operation conducted through the General Resources Account with a member other than the issuer of the currency. This practice ensures that the member is treated equitably by basing the calculation of what it is to pay to, or receive from, the IMF on a recent exchange rate. The exchange rate may be substantially different from the exchange rate in relation to the SDR on the basis of which the Fund is holding and valuing a currency immediately before the transaction or operation is carried out.

As noted above, on all other occasions on which the IMF is dealing with a currency, the rate in relation to the SDR on the basis of which the IMF is already valuing its holdings of the currency is applied. The occasions are those on which the IMF is dealing with a member in the member’s own currency. A new determination of an exchange rate is unnecessary, and is dispensed with, because the member incurs no detriment and gains no benefit if in such a case the IMF’s existing “book rate” is not adjusted for a new transaction or operation so as to bring the rate up to date. The member’s relationship with the IMF on the basis of the SDR as the unit of account is the same as it would be if the IMF’s holdings of the member’s currency in the General Resources Account were adjusted to conform with a recent change in the exchange rate and the member’s transaction or operation with the IMF were carried out at the adjusted rate. Many of a member’s rights and obligations under the Articles and under policies of the IMF are affected by the ratio, based upon the SDR as the unit of account, between a member’s quota and the IMF’s holdings of the member’s currency in the General Resources Account. This ratio is the same whether or not an adjustment of the IMF’s holdings of a member’s currency held in the General Resources Account is made for the purpose of transactions and operations not involving another member.

The obligation to maintain the value of the holdings of a currency in the IMF’s General Resources Account requires determination of the value of the currency in relation to the SDR. The IMF’s technique is to determine the value of the U.S. dollar in relation to the SDR and the value of any other currency in relation to the SDR through the medium of an exchange rate for the currency against the dollar. The rate for a currency other than the dollar against the SDR is calculated through the medium of the exchange rate of the currency against the dollar because most currencies are quoted regularly in this way. This procedure makes it possible to apply a uniform standard to all currencies.

The value of the U.S. dollar in terms of the SDR is equal to the reciprocal of the sum of the equivalents in U.S. dollars of the amounts of the five currencies in the SDR basket.6 The value in terms of the U.S. dollar of the other four currencies is calculated on the basis of exchange rates determined in accordance with procedures followed by the IMF from time to time. These procedures are followed solely for the purpose of determining the value of the U.S. dollar in relation to the SDR. The procedures do not establish representative rates for the other four currencies against the dollar for other purposes. Different procedures are followed for other purposes, as explained below. The central role of the U.S. dollar in international monetary affairs is reflected in the method of valuing the dollar in relation to the SDR. The effect of this method is that the amount of the U.S. dollar in the SDR basket is tacitly assumed not to fluctuate in value for the purpose described here, although the value of the dollar in relation to the SDR does fluctuate because of changes in the exchange rates of the other four currencies against the dollar.

The procedure for determining the U.S. dollar value of the other four currencies in the SDR basket for the purpose of arriving at the value of the dollar in relation to the SDR is to take the middle rate between the buying and the selling rates for the four currencies in spot transactions in the London foreign exchange market, as reported to the IMF by the Bank of England. If the exchange rate for any currency cannot be obtained in the London market, the middle rate at noon in the New York market is used, based on buying and selling rates communicated to the IMF by the Federal Reserve Bank of New York. If an exchange rate cannot be obtained in New York, the rate used is the middle rate at the fixing in the Frankfurt market, based on buying and selling rates communicated to the IMF by the Deutsche Bundesbank. The IMF’s decision establishing this procedure7 provides for other steps to be taken if an exchange rate is not available in accordance with the steps already described. The decision ensures that there will be no gap in the procedure for arriving at necessary exchange rates.

The London market is chosen for the four currencies because it is desirable to establish the value of the U.S. dollar in relation to the SDR as early in the day as possible. If the exchange rates of these currencies against the dollar were taken for this purpose from the four national exchange markets, it might not be possible to collect exchange rates for all the currencies at the same time, because of differences in the times at which the markets are in operation. If major changes were to occur in the exchange rates already notified to the IMF for a currency or currencies before the notification of exchange rates for another currency or currencies in the basket, the effect might be incompatible with the weights of the currencies in the SDR basket. The most active period during the day in major exchange markets is normally the hours between 11:00 a.m. and 2:00 p.m. The London market has been selected not only because it is an extremely active and substantial one, but also because the IMF can make the necessary calculations before its own business day begins.

The procedure as outlined above makes it possible for exchange risks arising from use of the SDR as the unit of account by private or official parties to be covered fully by transactions in the same market at the same time. This practice would not be possible if the IMF took the exchange rate for the yen from the Tokyo market, because that market closes before the European markets open for business.

The practice as described so far establishes the procedure exclusively for determining the value of the U.S. dollar in relation to the SDR. As noted above, to enable the IMF to carry out its transactions and operations, it is necessary to establish a procedure for determining the value in relation to the SDR of the currencies of all other members, including the other four members with currencies in the SDR basket. For this purpose, a representative exchange rate between a member’s currency and the U.S. dollar is established. If the IMF is satisfied that the member has an exchange market in which a representative spot rate for the member’s currency against the dollar can be readily ascertained, that rate is chosen for determining the value of the member’s currency in relation to the SDR.

This rule8 does not apply in three cases. First, as explained already, the value of the U.S. dollar in relation to the SDR is determined by the procedure as first described above, namely, the procedure involving exchange rates in the London market. Second, if a member (Patria) has an exchange market in which the IMF finds that a representative spot rate for Patria’s currency against the U.S. dollar cannot be readily ascertained, a cross-rate is calculated for this purpose. The calculation is made through the medium of a representative spot rate against Patria’s currency, readily ascertainable in Patria’s exchange market, for the currency of a third member (Terra) that has an exchange market in which a representative spot rate for Terra’s currency against the dollar is readily ascertainable. The rates in this calculation are the representative rate for Patria’s currency against Terra’s currency in Patria’s exchange market and a representative rate for Terra’s currency against the dollar in Terra’s exchange market. Third, if the procedures as summarized in this discussion are not adequate to establish the value of a currency in terms of the SDR, the IMF will determine the rate.

A member is entitled under the Articles to choose its exchange arrangement and to determine the external value of its currency. This fact and the indisputable jurisdiction of a member to manage the exchange rate for its currency in its own exchange market are the rationale for normally taking the exchange rate for a member’s currency from its own market. The IMF, however, decides what procedure it will follow for selecting the exchange rate the IMF will apply for the purposes of the Articles. The procedure is specified by the IMF, but in consultation with the member. The object is to recognize the kind of exchange rate that is “representative” of the rate in the bulk of exchange transactions in a member’s exchange market. In practice, the IMF requests a member to describe what kind of exchange rate it considers representative of the rates for its currency, and usually the IMF accepts the member’s choice.

EMS Practice

The practice of the EMS for calculating the value of each currency in terms of the ECU closely resembles the IMF’s practice with respect to the SDR. In particular, the practice of the EMS involves the exchange rate for each currency in the market of that currency. The Basle Agreement of March 13, 1979 provides, in relation to the operation of the divergence indicator, that “the market value of the ECU in each currency shall be calculated by a uniform method as frequently as necessary and at least on the occasion of each daily concertation session among central banks.” In the accounting procedures of the EMCF for the purposes of the Very Short-Term Financing Facility, the translation of currencies into ECUs is to be effected at the daily rates for the ECU as established by the Commission’s staff on the basis of the method adopted. The method has been formulated as follows:

The central bank in each Member State communicates a representative market exchange rate for its currency against the United States dollar. The dollar has been chosen as giving the most representative rate in all financial centres. The rates are taken from the exchange markets at 2.30 p.m. They are communicated by the National Bank of Belgium to the Commission, which uses them to calculate an ECU/ EUA equivalent first in dollars and then in the currencies of the Member States. If an exchange market is closed, the central banks agree on a representative exchange rate for the currency against the dollar which is communicated to the Commission.9

Delayed Discharge of Obligations to IMF

Problems of determining the appropriate exchange rate to be applied in transactions and operations conducted through the General Resources Account arise when exchange rates are fluctuating and the discharge of obligations is delayed beyond the due date for performance. These problems have arisen in connection with obligations owed to the IMF.

The present practice of the IMF on the selection and calculation of exchange rates, as described above, is based on the assumption that members perform their obligations on the due date. A member is responsible if for any reason it does not perform on time. If a member gives instructions to one of its entities or a commercial bank to make a timely transfer to the IMF, but delay occurs, the IMF’s instructions governing performance remain in force. However, the amounts of currencies necessary to discharge the obligation on the basis of the SDR as the IMF’s unit of account are adjusted to accord with exchange rates prevailing three business days before the day on which the late performance takes place.

This practice of adjustment is subject to two exceptions. If the currency used to discharge the obligation has depreciated since the due date, the member performing its obligation must transfer more units of the currency to the IMF. If the member did not, it would be failing to maintain the SDR value of its transfer in accordance with the member’s maintenance of value obligation. But if the calculation made for adjustment under the three-business-day rule is less than the equivalent of SDR 5,000, adjustment is waived, unless the member that issues the currency being transferred to the IMF elects to see the adjustment made. The issuer has this option because eventually it will bear the burden of adjustment under the provisions on maintenance of value. The waiver is intended to avoid administrative costs for the IMF and members when de minimis amounts are due, even though maintenance of value is not observed in such a case.

The IMF’s practice is more complicated if the currency of payment has appreciated since the due date. A distinction is made between two situations. In the first situation, the transfer is not made within ten days from the date of the IMF’s instructions. If the member informs the IMF that the transfer will be made on a specified date after that period, the IMF issues new instructions, and the member will transfer an amount calculated on the basis of the SDR rate for the currency prevailing three business days before the new value date (namely, the date specified by the member as the one on which it will discharge its obligation). Suppose, however, that a member makes a delayed transfer after ten days from the date of the IMF’s instructions without informing the IMF of the member’s intention. The amount due is nevertheless recalculated by the IMF on the basis of the exchange rate prevailing three business days before the date of the actual transfer.

In both cases described above of transfer after a delay of ten days, the currency may appreciate. The member then transfers fewer units of the appreciated currency than the member was called upon to transfer under the original instructions. It can be said that, in a certain sense, the member has benefited by delay, or at least has been relieved of a burden, but the rationale of the practice is that the requirements of maintenance of value are being strictly observed.

The second of the two exceptions referred to earlier applies to a situation in which again the currency of transfer has appreciated after the due date but the member discharges its obligation within ten days after the IMF’s instructions. In this case, the IMF does not issue fresh instructions to adjust the amount the member must transfer so that the amount will be in accord with the new exchange rate. As a result, the IMF receives more than is necessary to maintain value in terms of the SDR three business days before the date of actual payment. As the IMF has received more than enough according to this standard, the IMF, to maintain the value of its holdings of the currency in the General Resources Account, transfers the excess to the member issuing the currency received. An oddity of the contrast between the IMF’s treatment of cases of appreciation is that the member discharging its obligation later than ten days after the IMF’s instructions gets more generous treatment than the member that makes a transfer within the ten days. In neither case, however, does the IMF profit.

The legal basis for the IMF’s treatment of appreciation when a delayed transfer is made within ten days is unclear. A possible explanation is that the IMF’s unwillingness to make an adjustment is intended to deter members from delaying settlements in order to benefit from an expected appreciation of the currency of settlement. Another rationale might be that the IMF is imposing a sanction on delinquent members. The issue then would be whether the IMF has an implied power to impose sanctions not made explicit in the Articles, but even if an affirmative answer were given to this question there would be no explanation of the different treatment of delayed settlement within and beyond the period of ten days.

The explanation of a sanction recalls the provisions of the Articles on the appreciation of a member’s currency that was floating in violation of the member’s obligations under the former par value provisions of the Articles. Article IV, Section 8 on maintenance of value under the original Articles was explicit in requiring a member to transfer more of its currency to the IMF if its currency depreciated in gold value, but did not declare that the IMF was to return an appropriate amount of the member’s currency to the member on an appreciation in gold value. Both kinds of adjustment were to be made on a devaluation or revaluation, that is, on every change in par value in conformity with the Articles. It would have been easy for the IMF to decide that the treatment of both depreciation and appreciation in nonconformity with the Articles was intended to be a sanction in both cases to deter illegality: on a depreciation the member transferred more units of its currency, and on an appreciation the member was not entitled to receive any return of currency.

The IMF did not draw this deduction, and instead considered appreciation as a casus omissus. The IMF took account of the fact that if it sold the appreciated currency at par, it would encourage requests to purchase the currency because purchasing members would want to enjoy an exchange profit.10 If the IMF sold the currency at the appreciated exchange rate, the IMF would make an exchange gain. To avoid this predicament, the IMF decided that it was bound to maintain the value of its holdings of the appreciated currency by returning an amount of the currency that corresponded to the appreciation. The IMF drew this deduction from the general language of subsection (a) of Article IV, Section 8 of the original Articles on maintenance of the gold value of the IMF’s assets: “The gold value of the Fund’s assets shall be maintained notwithstanding changes in the par or foreign exchange value of the currency of any member.” “Foreign exchange value” was interpreted to embrace appreciation as well as depreciation. With this interpretation, the IMF was in a position to sell the currency at the appreciated exchange rate without the complications that would follow if value were not maintained.

The deterrence of delay in the performance by members of their financial obligations to the IMF has become more important than ever. The desire to deter is particularly strong in a period in which arrears in the discharge of financial obligations to the IMF have accumulated and have created grave concern. In the days of the par value system when the effect of appreciation under the provision on maintenance of value was settled, nobody thought that arrears would occur.

The practice by which delayed payments are adjusted to conform with changes in exchange rates is comparable, to the extent that the IMF permits adjustment, to the movement from the breach day rule to the judgment day or payment day rule when the courts of some countries deal with overdue payments. This topic is discussed in Chapter 12 of this volume.

Practices on the selection of exchange rates similar to some of those described above apply to payments in IMF-related transactions even though transfers to the IMF are not involved. For example, the three-business-day rule applies to transfers of SDRs between a member wishing to transfer SDRs and a member designated by the IMF to receive them. If the transferee of SDRs delays the transfer of currency in exchange for the SDRs, the transferee is not entitled to pay less than the amount stated in the IMF’s instructions because of the appreciation of the currency. It would be inappropriate to subject the transferor of SDRs not only to delay but also to the inconvenience or worse of obtaining less than it expected to receive and use, for example, in the discharge of obligations.

It will be observed, therefore, that the IMF’s practices on delayed payments are not confined to maintenance of the value of holdings in the General Resources Account. The practices are designed also to ensure that the principle of equal value in transactions involving SDRs is made effective. The transferor of SDRs must receive the same value whatever the currency that is provided and whether or not it is provided on the due date.

Negotiable Instruments and Exchange Rates

The proposed United Nations Convention on International Bills of Exchange and International Promissory Notes11 is an example of a convention that will recognize the freedom of private parties to specify exchange rates but will include detailed provisions to deal with the cases in which parties have not selected a rate. Article 75 provides that an instrument must be paid in the currency in which the sum payable is expressed, but that the drawer may declare in the instrument that payment is to be made in a specified currency other than the currency in which the sum payable is expressed. In the latter case, the instrument must be paid in the currency specified for payment. The currency in which the sum payable is expressed is the unit, or the currency, of account, but may or may not be the currency of payment.

The amount payable in the specified currency if not also the currency of account is to be calculated according to the rate of exchange prescribed in the instrument. In the absence of a prescribed rate, the amount payable is calculated according to the rate of exchange for sight drafts (or, if there is no such rate, according to the appropriate established rate of exchange) on the date of maturity

  • (i) ruling at the place where, in accordance with the Convention, the instrument must be presented for payment, if the specified currency is the domestic currency of that place, or

  • (ii) if the specified currency is not the domestic currency at the place of presentation, according to the usages prevailing there.

It would seem that the established rate of exchange should not be interpreted to refer exclusively to an official fixed rate. The concept should be interpreted as broadly as possible in view of the diversity of national practices on exchange rates in the discretionary system of exchange arrangements.

The Convention contains rules for determining the applicable exchange rate if the instrument is dishonored by nonacceptance or nonpayment. In the case of nonacceptance, the rate of exchange is the one included in the instrument; and if there is no such rate, the holder has the option of choosing between the rate of exchange ruling on the date of dishonor or on the date of actual payment. An option of this kind is undoubtedly inspired by ideas of restitutio in integrum and probably also the preference that a nonperforming party rather than the innocent party should bear the burden of any financial disadvantage that results from the failure. The allocation of such burdens is discussed more extensively in connection with the Miliangos doctrine in Chapters 12 and 15 of this volume.

If an instrument is dishonored by nonpayment, a comparable rule applies. The amount payable is to be calculated on the basis of the exchange rate included in the instrument; and if there is no such rate, the holder has the option of choosing between the rate of exchange ruling on the date of maturity and the rate ruling on the date of actual payment.

Paragraph 4 provides that nothing in Article 75 prevents a court from awarding damages for loss suffered by the holder by reason of fluctuations in exchange rates after dishonor by nonacceptance or by nonpayment. Paragraph 4 means that the availability of damages in these circumstances is left to the courts. It can be assumed that national laws differ on the availability of damages for loss occasioned by the nonpayment of a debt and changes in exchange rates, and that the negotiators of the convention have regarded their own laws on the subject as entrenched and properly beyond the reach of the Convention.

Paragraph 5 of Article 75 sets forth a rule that explains what is meant by the exchange rate ruling at a certain date in any case of dishonor. The rate is the one ruling, at the option of the holder, at the place where the instrument must be presented for payment or at the place of actual payment. The holder, therefore, has a double option: he can choose between two places in determining the market in which the rate is to be selected, and he can choose between the rates ruling at the place of his choice on two different dates. The option of place may be confined to cases of dishonor, because paragraph 1(b) prescribes the exchange rate ruling at the place where the instrument must be presented for payment when payment is made in accordance with the terms of the instrument. If this interpretation is correct, the option of place is again an example of penalizing a nonperforming party through the medium of fluctuating exchange rates.

Article 76, paragraph 1 declares that nothing in the Convention prevents a contracting party from enforcing exchange control regulations applicable in its territory and regulations relating to the protection of its currency, including regulations it is bound to apply by virtue of international agreements to which it is a party.12 If, by virtue of Article 76, paragraph 1, an instrument drawn in a currency that is not the currency of the place of payment must be paid instead in the domestic currency of that place, the amount payable is to be calculated according to the rate of exchange for sight drafts (or, if there is no such rate, according to the appropriate established rate of exchange) on the date of presentment ruling at the place where the instrument must be presented for payment in accordance with the Convention.

If the instrument referred to in the preceding paragraph of this discussion is dishonored by nonacceptance, the amount payable is to be calculated, at the option of the holder, at the exchange rate ruling on the date of dishonor or on the date of actual payment. If such an instrument is dishonored by nonpayment, the amount is to be calculated, at the option of the holder, according to the rate ruling on the date of presentment or on the date of actual payment. The options of place and date of exchange rate in these cases of dishonor, in contrast to the options under paragraphs 4 and 5 of Article 75, seem not to be motivated by the desire to penalize, because dishonor under Article 76 is the consequence of exchange control or other regulations over which the obligor has no authority.

Treaties on Investment

The exchange rate ruling on the date of actual payment as an option under the Convention deserves notice. This exchange rate has received increasing favor in treaty law as the appropriate rate for numerous purposes in the discretionary system of exchange arrangements, in which the fluctuation of exchange rates is widespread. The Miliangos decision is part of a similar trend in national law and may have been an influence in promoting the trend in international law.

The same exchange rate has greater importance than ever under treaties for the reciprocal promotion and protection of investments. Treaties of this kind provide that for transfers from the country of investment the exchange rate shall be “the exchange rate applicable on the date of transfer” in the country in which the investment was made,13 or “the rate of exchange prevailing at the time of remittance.”14 Sometimes the treaty provides that the applicable exchange rate shall be the “prevailing rate of exchange on the date of transfer with respect to current transactions in the currency to be transferred.”15 The rate under such a formulation applies whether the transfer is of a current or a capital character in economic analysis or under the definitions in the IMF’s Articles. The formulation prevents the use of a special rate for capital transactions if there is one in an exchange system because of the assumption that if there is such a rate it is likely to be more unfavorable for the transferee.16

In some instances the formula in a treaty requires that transfers “shall be made without delay and shall be freely transferable at the official rate of exchange prevailing on the date used for the determination of value.”17 The reference to an official rate of exchange may mean a fixed rate or a rate quoted by the monetary authorities, but may also mean market rates consistent with the exchange arrangement chosen by a contracting party.18 Finally, all formulations should mean the exchange rate prevailing at the date of actual transfer whether or not the transfer has been delayed, and whether the delay has been proper or improper.

IMF’s Advice to Other Organizations

It is not uncommon for treaties to declare that the procedure for selecting the appropriate exchange rate for some purpose under the treaty shall include consultation with the IMF. The advice of the IMF may be binding, or, particularly if the treaty is the constitutive treaty of another organization, the treaty will provide that the organization must take account of the IMF’s advice but is not necessarily bound by it. For example, whenever it is necessary for the purposes of a loan agreement or a guarantee agreement to determine the value of a currency in terms of the SDR under the Agreement establishing the International Fund for Agricultural Development (IFAD), the value is to be calculated in accordance with Article 5, Section 2(b) of that treaty:

For the purposes of this Agreement, the value of a currency in terms of the Special Drawing Right shall be calculated in accordance with the method of valuation applied by the International Monetary Fund, provided that:

  • (i) in the case of the currency of a member of the International Monetary Fund for which such value is not available on a current basis, the value shall be calculated after consultation with the International Monetary Fund;

  • (ii) in the case of the currency of a non-member of the International Monetary Fund, the value of the currency in terms of the Special Drawing Right shall be calculated by the Fund [IFAD] on the basis of an appropriate exchange rate relationship between that currency and the currency of a member of the International Monetary Fund for which a value is calculated as specified above.

Clause (ii) respects the sensitivities of nonmembers of the IMF by not requiring consultation with the IMF on the exchange rate for a nonmember’s currency, but the exchange rate is nevertheless related to the currency of a member of the IMF for which a value has been calculated in accordance with clause (i).

More recently, the problem of determining the appropriate rate of exchange for the purposes of a treaty arose in the negotiation of the Convention Establishing the Multilateral Investment Guarantee Agency (MIGA). Article 9 of the Convention is formulated as follows:

Whenever it shall be necessary for the purposes of this Convention to determine the value of one currency in terms of another, such value shall be as reasonably determined by the Agency, after consultation with the International Monetary Fund.

MIGA and Exchange Rates

Under Article 17, MIGA takes decisions on the payment of claims to the holder of a guarantee in accordance with the contract of guarantee and the policies of MIGA. Contracts of guarantee must require holders of guarantees to seek, before a payment is made by the Agency, such administrative remedies as may be appropriate under the circumstances, provided that the remedies are readily available to the holders under the laws of the country of investment (the host country).

The rights and obligations of MIGA and insured investors are set out in General Conditions of Guarantee for Equity Investments. The subject of Article 16 of the General Conditions is the concept of a reference rate of exchange for translating the currency of the host country into freely usable currency when it becomes necessary to determine the amount of compensation that MIGA is obligated to pay to an insured investor. The reference rate is the effective average rate of exchange applied by the central bank of the host country on the date of loss for the exchange of the domestic currency into the guarantee currency. (The term “central bank” includes any other body with regulatory authority over foreign exchange.) To reduce the chance of dispute between MIGA and the insured investor and the need for arbitration, a descending order of categories of exchange rates is specified, and the exchange rate to be applied by the central bank must respect this order if on the date of loss the central bank has not freely provided the guarantee currency or has applied multiple exchange rate categories with respect to conversions into that currency. The order of exchange rates does not involve any idea of imposing a financial disadvantage on anyone. The order is fixed, and there is no option to depart from it.

The descending order consists of five categories, formulated as follows:

  • (a) first, the exchange rate category generally applied by the central bank on the Date of Loss to earnings remitted on account of foreign equity investments;

  • (b) second, the exchange rate category generally applied by the central bank on the Date of Loss for purposes of servicing private foreign debt;

  • (c) third, the most depreciated (that is, requiring the greatest amount of Local Currency per Guarantee Currency unit) exchange rate category applied by the central bank on the Date of Loss for the sale of Guarantee Currency monies to private residents of the Host Country;

  • (d) fourth, the most representative clearing rate on the Date of Loss legally used by commercial banks or any other private market in the Host Country; and

  • (e) fifth, the clearing rate used on the Date of Loss outside the Host Country in the most active market for conversion of Local Currency into the Guarantee Currency.19

The reference rate of exchange is the average of the high and the low exchange rates for the applicable exchange rate category on the date of loss. Any payment by MIGA is computed net of all charges and expenses due in case of conversions and transfers by or on behalf of the guarantee holder under the prevailing laws, regulations, and business practices of the host country. In the IMF’s practice, the effective, in the sense of the true, exchange rate is regarded as the rate that includes certain charges and expenses.

If MIGA is unable to determine the reference rate of exchange according to the practice described above, or if the guarantee holder objects within 30 days of being notified of MIGA’s determination, the reference rate of exchange must be determined, with the approval of the guarantee holder, by applying the order of exchange rates quoted above to any freely usable currency other than the guarantee currency. If MIGA does not obtain such approval within 30 days from the date of request for approval, either party may request the Managing Director of the IMF to designate an expert to make a final decision on the applicable reference rate of exchange. The IMF has authorized the Managing Director to perform this service.

It is noticeable that the choice among the prescribed categories when the host country applies multiple rates of exchange does not depend on the IMF’s approval of the multiple currency practice that gives rise to the multiplicity of exchange rates. Under Article VIII, Section 3 of the IMF’s Articles, a member is required to avoid multiple currency practices or discriminatory currency arrangements unless approved by the IMF or authorized by the transitional arrangements of Article XIV of the IMF’s Articles.20 An exchange rate can be the result of a multiple currency practice and constitute a discriminatory currency arrangement at the same time. The freedom of members to choose their exchange arrangements under the discretionary system of exchange arrangements has not absolved members from their obligations under Article VIII, Section 3.

The English case of Lively Ltd. and Another v. City of Munich,21 however, can be cited for the proposition that whatever may be the status of exchange rates under the IMF’s Articles as between the IMF and a member, market exchange rates should be applied when necessary to do justice between contracting parties:

The present issue is concerned with a rate of exchange applicable to a commercial transaction; it is not concerned with treaty obligations by governments to the IMF or interse.22

The case suggests also that a particular rate of exchange can be applicable under a treaty to commercial contracts without depriving the contracts of their commercial character.

If the IMF were approached by MIGA for its advice on exchange rates in a particular case, it is improbable that the IMF would fail to mention the inconsistency of one or more of the rates with the Articles. The IMF should not refuse its services, however, because the applicable rate was deemed to be inconsistent with the IMF’s Articles. MIGA would have to decide what to do if the IMF stated that an exchange rate was inconsistent with the IMF’s Articles. MIGA might decide to apply the principle of Lively Ltd. and Another v. City of Munich.

MIGA’s categories of exchange rates illustrate the importance of precision in international arrangements not only about the kind of exchange rate that is to apply for some purpose but also the date of the selected exchange rate.23 The regulations of MIGA provide that contracts of guarantee shall specify the currency into which conversion is guaranteed; and also the basis and the date of the exchange rate or rates to be applied in calculating a claim. Normally, the rate will be the one prevailing in the host country on the date on which the host government denies, or is deemed to have denied, conversion or transfer at the exchange rate specified when the guarantee was issued.24 Under this practice, the applicable rate of exchange will be the one prevailing on the date of loss as defined and not on the date of actual payment. The latter date is the practice prescribed by treaties on the reciprocal promotion and protection of investments discussed above. Furthermore, as noted already, the investor does not have an option to choose the exchange rate more favorable to him on the model of the proposed United Nations convention on negotiable instruments.

Disputed Exchange Rates

P. Dumortier Frères SA and Others v. Council of the European Communities,25 decided by the European Court of Justice, illustrates the disputes that can arise under international arrangements if there is no clear specification of the applicable exchange rate. In an interlocutory judgment of October 4, 1979,26 the court ordered the EC to pay the applicants amounts equivalent to certain production refunds in respect of the period August 1, 1975 to October 19, 1977, because the refunds had been abolished without legal justification. The court also awarded interest at 6 percent from the date of the interlocutory judgment. The parties were to agree on, and inform the court of, the amount of compensation to be paid, and to present their views to the court if they failed to agree. The parties agreed on the amount of compensation payable in terms of the European unit of account (EUA), but did not agree on the date of the exchange rate to be applied in calculating the equivalent in French francs or in ECUs when the ECU replaced the EUA.

The applicants maintained that the appropriate date was the date on which the court delivered its interlocutory judgment. The Community authorities argued that the appropriate dates were the various dates of production in respect of which refunds had been improperly withheld. The applicants’ main arguments were threefold. First, the interlocutory judgment had awarded compensation, and compensation was governed by principles common to the laws of member states. One principle was that the date on which interest begins to run is the same as the date for assessing damages. The interlocutory judgment had awarded interest from October 4, 1979 and not from the dates when the refunds should have been paid. October 4, 1979 must be understood to be the date as of which the right to compensation was established and the amount of damages had to be calculated. Second, the problem resembled the translation of foreign currency into national currency when damage has been suffered in the foreign currency as the currency of account. Third, French francs were payable in France, and the payment had to be in accordance with French case law, under which the exchange rate prevailing on the date of judgment is applied for translating into French francs the amount of damage assessed in a foreign currency.

The Council contended that the interlocutory judgment had awarded compensation in units of account equivalent to the refunds that should have been paid monthly during the period August 1, 1975 to October 19, 1977. The judgment implied that translation of the units of account into French francs had to be made as of the various dates of production, so as to put the applicants into the same financial position they would have been in if the refunds had been paid. (The effect of such a decision would be less favorable to the applicants because of subsequent changes in exchange rates.) Furthermore, if the date of judgment were applied, the effect would be unfair because of the differences in exchange value among currencies that had occurred before that date. The deutsche mark had been revalued in relation to the unit of account, but the French franc had been devalued. German producers would lose, but French producers would profit, in terms of their national currencies if exchange rates prevailing on the date of judgment were adopted as the solution.

The Council argued also that the right to compensation existed before the judgment and was not created by it. The function of the judgment was only to establish precise criteria for the assessment of compensation. The rate of interest awarded by the court was compensation for the failure to make timely payments.

The court held that the interlocutory judgment had not awarded arrears of refunds. It had awarded compensation for liability incurred by the Community because the refunds had been abolished without proper cause. The compensation was equivalent to the amount of the refunds, but only as a basis for calculating the compensation and not as the refunds themselves. The award of interest from the date of the judgment showed that the court intended to assess the damage as of that date. The court held that translation into the national currencies of all producers, wherever established, had to be made by applying the exchange rates prevailing at October 4, 1979.

The Advocate General offered information about how damages would be awarded for noncontractual liability on the basis of principles common to the laws of the member states of the Community.27 He found a fairly clear tendency in favor of the principle that damages must be determined with reference to the date of judgment. This principle was based on the logic of restitutio in integrum. In both civil and criminal matters, the French Court of Cassation (Commercial Chamber) had decided in some cases that the rate of exchange as at the date of payment, and in other cases as at the date of judgment, must be applied. By contrast, the Conseil d’Etat had held in relation to the noncontractual liability of public authorities that compensation expressed in foreign currency must be translated into the national currency at the rate of exchange prevailing when the obligation arose, namely, when the damage was caused. Italian courts must take into account a loss of value suffered before judgment is rendered. In English law, the rate of exchange prevailing at the date of actual payment applies. The laws of Belgium, Germany, and the Netherlands observe the same principle. The laws of Denmark and Ireland do not take account of variations in exchange rates after the date when damage occurs.

The Advocate General took a view on what would constitute the discriminatory treatment of undertakings that differed from the one advanced by the Council. His initial premise was the necessity to respect the Community’s legal principle of equal treatment, which he took to mean equal purchasing power in the circumstances of the case. If the Council’s argument in favor of the dates when refunds should have been made were adopted, undertakings receiving French francs in France would receive less purchasing power than undertakings receiving deutsche mark in Germany. The reason for this discrepancy was that since the date or dates for which the Council contended the franc had been devalued and the deutsche mark revalued. The compensation awarded at the date of the court’s judgment in this proceeding could give the assurance of equal purchasing power only if compensation was based on exchange rates at that same date.

Averaging Exchange Rates

Two forms of averaging are resorted to because of the fluctuation of exchange rates. One form shows the behavior of the exchange rate of a denominator over a period of time. The motive for averaging of this kind is likely to be concern that the exchange rate of the denominator on a particular date may be untypical of behavior over a period. For this reason, the view is held that it would be undesirable to apply the exchange rate prevailing on a particular date. The other form of averaging shows the exchange rate behavior of a group of currencies by means of a weighting process. Composite denominators such as the SDR and the ECU are examples of the latter form of averaging. The first form of averaging is likely to be chosen in order to compare average exchange rates at different times. The second form of averaging can serve this same purpose, but the purpose may be to apply the average exchange rate of the group of currencies at one particular date.

The basic rule for conducting the IMF’s transactions and operations is that the exchange rate shall be the rate as of three business days before the value date of the transaction or operation. It has never been suggested that the rate of exchange should be other than the rate on a single day. The main reason for applying the rate on a single day is that many of the transactions are assimilated to the normal spot exchange transactions that are conducted through exchange markets. Furthermore, a member purchasing currency from the IMF may want to exchange it in the market for another currency. If the purchase from the IMF had to be made at an average rate, that rate might not reflect current market conditions, and the member might find that the rate at which it could exchange the purchased currency was an unfavorable one. Another reason for the practice is the principle of the original Articles that a member purchasing the currencies of other members from the IMF had to have a need to make immediate use of the currencies to make payments. It was considered inadmissible for a member to purchase currencies from the IMF for addition to the member’s reserves. Although the Second Amendment includes a broader definition of “need,” the rule that the exchange rate as of a single day is the appropriate rate remains unchallenged.

For the purposes of a variety of other treaties, however, the technique of averaging exchange rates over a period has seemed more suitable. Normally, the activities conducted under these treaties do not include transactions that resemble spot exchange transactions.28 When the exchange rates of currencies are fluctuating, to use the rate on a single day may seem inequitable, because the rate for a currency on any one day may be unrepresentative of the trend in exchange rates for the currency over whatever period seems appropriate. The exchange rate may have been developing, or may be expected to develop, along a path, from which, however, there may be brief divagations. Another, although related, reason may be that an average exchange rate is desirable because it will produce a rate more stable than any exchange rate prevailing on a single day.

Averaging has been adopted in two kinds of cases. In one, the averaging results in a datum that is fixed in value on the basis of the exchange rates over a single period in the past. In the other kind of case the averaging results in a changeable datum because the average is calculated over a movable and more recent period. In both kinds of cases, the length of the period as well as the initial and terminal dates may be controversial.

A recent example of the first kind has been cited already. MIGA defines the Agency’s authorized capital stock by reference to the average value of the SDR in terms of U.S. dollars for the period January 1, 1981 to June 30, 1985. The value calculated in this way is US$1.082, and it does not vary with subsequent changes in exchange rates. The reasons for the choice of the opening and closing dates of the period have been mentioned already. It was noted also that the length of the period had been controversial, with proposals ranging from six months to eleven years. The long-term trend of the exchange rate may seem more suitable to some negotiators because the result of the calculation is going to be a datum that will be in prolonged or permanent use. Those preferring a shorter-term trend may think it more suitable for the single or the occasional application of it as the basis for making payments such as subscriptions. Another justification of a shorter period may be the prospect that an average over a long period may have little relationship to current exchange rates and be even less of an indicator of future exchange rates.

It has been seen that the Agreement Establishing the EBRD adopts a form of averaging somewhat similar to that of MIGA. The original authorized capital stock of the EBRD is denominated in the ECU. Each member has the option to pay its subscription in ECUs, U.S. dollars, or yen. If the dollar or the yen is chosen, the payment is to be made on the basis of the average exchange rate of the chosen currency in terms of the ECU for the period from September 30, 1989 to March 31, 1990. The average dollar value for this period was $1.16701. The proposed treaty was signed in late May 1990, so that it would have been possible simply to include this datum instead of a formula in the text.

Averaging has been a feature of the calculation of contribution shares in the replenishment of IDA’s resources. For the seventh replenishment, for example, it was agreed that the total would be equivalent to $9 billion. Donors reached agreement on their percentage shares in this amount, which were then calculated for each in U.S. dollars. A donor had the option to express its share in various units of account, including the donor’s national currency. It was agreed that for the tabulation of contributions, expressed for all donors in U.S. dollars, national currency, and SDRs, the dollar amounts would be translated into national currency by using a six-month average of monthly exchange rates against the U.S. dollar over the period June 10 to December 9, 1983. The Board of Governors of IDA approved the resolution relating to the seventh replenishment on August 6, 1984 for commitments to be made during the fiscal years 1985 through 1987.

For the eighth replenishment, it was agreed that the U.S. dollar equivalent of the shares accepted by donors, applied to a total of $11.5 billion, would be translated into amounts of national currency using a six-month average of daily exchange rates (for business days only) against the dollar over the period March 1 to August 29, 1986. For donors choosing to denominate their obligations in the SDR, the equivalent of the U.S. dollar amounts would be translated into SDRs using a three-month average of daily exchange rates (for business days only) against the U.S. dollar over the period June 2 to August 29, 1986. The report on this replenishment does not explain why a shorter period of averaging was chosen for the amounts denominated in SDRs than for the amounts denominated in national currency. The explanation may be that a longer period was chosen for national currencies because the dollar-SDR exchange rates were less variable than the exchange rates between the currencies of participants and the dollar. The Board of Governors of IDA adopted the resolution on the eighth replenishment on June 26, 1987, authorizing new credit commitments for the period July 1, 1987 to June 30, 1990.

The ninth replenishment approved by IDA’s Board of Governors on May 8, 1990 amounts to SDR 11.68 billion for new credit commitments in the period from July 1, 1990 to June 30, 1993. The replenishment share of each member that has the option to denominate its share in SDRs or in the member’s national currency has been calculated by applying an average of the daily exchange rates, for business days, of the currency against the SDR during the period May 1 to October 31, 1989. For the class of members with shares that can be denominated in the SDR alone, averaging to arrive at the equivalent of a share in national currency is unnecessary.

The averaging techniques for MIGA and IDA described above have some common features. The exchange rates were averaged over a brief period of a few months. The periods of averaging were not close to, and would become progressively remote from, the periods in which the undertakings would be discharged. The amounts resulting from the technique were fixed and not subject to subsequent adaptation.

In some instances, the period of averaging may be much shorter than any of those mentioned already. For example, the Convention on Civil Liability for Oil Pollution Damage Resulting from Offshore Operations, 1976 permits the operator of certain installations to limit his liability for oil pollution. Liability is limited to specified amounts of SDRs translated into the currency of the contracting party in the territory of which the operator establishes a limitation fund. The fund is calculated on the basis of the average value of the currency in terms of the SDR during the 30 days preceding the date on which the fund is constituted. The amount of the fund is fixed in this way and is not adjusted because of subsequent changes in exchange rates.

The Agreement Establishing the Asian Clearing Union provides an example of averaging over a short but constantly moving period, so that the results of the calculations change with fluctuations in exchange rates over the short term. The Asian Clearing Union is a facility for settling payments for current international transactions on a multilateral basis within the area of the United Nations Economic and Social Commission for Asia and the Pacific (ESCAP). The unit of account of the Clearing Union is the Asian monetary unit (AMU), which is defined as equivalent to 1 SDR, although the Board of Governors of the Union is authorized to change the method of valuation of the AMU. The purpose of this authorization is to provide a safeguard if the SDR ceased to exist, the IMF’s method of valuing the SDR was considered unsatisfactory for the Clearing Union, or the IMF stopped publishing exchange rates in terms of the SDR for the currencies of participants in the Clearing Union.29

The rate of exchange of a participant’s currency in terms of the AMU is based on averaging. The rate is calculated by multiplying (1) the middle rate between the participant’s buying and selling spot rates for its intervention currency and (2) the rate of the intervention currency in terms of the SDR. The exchange rate applicable for the first accounting period of a month is the average of the daily exchange rates from the twelfth day of the previous month to the twenty-sixth day of that month. For the second accounting period of a month the rate is determined in the same way from the twenty-seventh day of the previous month to the eleventh day of the current month. (The General Manager, however, may modify this formula, but he must base the modification on the daily exchange rate.) If the exchange rate of the currency of any participant varies by more than percent of the rate last communicated for accounting purposes, the General Manager must immediately communicate the revised rate. That rate is then effective from the second day following the day on which it is communicated to the end of the current accounting period. The rate for the next accounting period is the average of the daily rates from the day of the effective revision to the twelfth or the twenty-seventh day of the month, as the case may be.30 Obviously, both brevity of the period for averaging and proximity of the period to the date of settlement were considered desirable for a clearing facility.

It must not be thought that the technique of averaging exchange rates is applied only in order to arrive at rates of exchange for the direct settlement of financial transactions. For example, averaging is involved in the Guidelines for decisions under Article II:6(a) of the General Agreement on Tariffs and Trade (GATT), a provision discussed more fully in Chapter 9. The Guidelines provide for a comparison between the import-weighted average exchange rate of a contracting party’s currency during the period of six months preceding a prescribed date and a similar average exchange rate during the period of six months preceding the contracting party’s request for adjustment of bound specific duties under the provision. The Guidelines are not based on a comparison of the exchange rates at the end of the two periods in order to avoid the effect of transitory fluctuations. The periods for averaging, however, are limited because prolonged periods could have too substantial a moderating effect on recent fluctuations.

A second example shows that the IMF may accept the case for averaging when its operations and transactions are not directly involved. The IMF’s decision on quinquennial reviews and possible revision in the method of valuation of the SDR includes an element of averaging. The amounts of the five currencies in the SDR basket are determined on the last working day preceding the effective date of a revision. The determination is to be made in such a way that, at the average exchange rates during the period of three months ending on the date referred to, the shares of the currencies in the value of the SDR correspond to the percentage weights for the currencies arrived at in accordance with the prescribed criteria and the quinquennial period to which they apply.

In addition, the determination of the amounts of currencies is to be made in such manner that the value of the SDR in terms of currencies will be the same immediately before and immediately after a revision becomes effective. This rule means that the amounts of currency units initially determined in accordance with the prescribed criteria and average exchange rates are subject to small equiproportional adjustments in order to achieve the necessary equality. This limited equality avoids an abrupt change in the value of the SDR in terms of currencies for all entities for which the SDR is the unit of account.31

Averaging and SEC

A complex problem of selecting the appropriate exchange rate and the method of averaging can be cited from the practice of the Securities and Exchange Commission (SEC) of the United States. New and amended Rules and Regulations32 were published on December 6, 1982 to regulate the currency in which the financial statements of foreign private issuers of securities offered in the United States must be presented in registration filings with the SEC under the Securities Act of 1933 of the United States. A registration statement normally must contain consolidated income statements for the five most recent years. The staff of the SEC had understood that the primary financial statements were to be presented in the currency of the domicile of the foreign private issuer and that supplemental financial statements were to be presented in U.S. dollars for the convenience of residents of the United States.

The supplemental statements (“convenience translations”) made it necessary to select the exchange rate between the foreign currency and the U.S. dollar for the purpose of the translation. The exchange rate had been the rate in effect at the date of the most recent balance sheet included in the filing. The same exchange rate had to be used for all periods covered by the consolidated income statement. This method was considered satisfactory while the par value system was in force, because there was a good chance that the same rate (or, more properly, the same par value) had prevailed during the whole period covered by the financial statements.

The practice was criticized after the par value system was no longer in force and exchange rates were fluctuating. Critics33 discussed four possible methods for selecting and applying the exchange rate or rates for convenience translations:

  • (1) the single exchange rate as of the date of the most recent balance sheet, and application of the rate in translations for the whole of the most recent five years;

  • (2) a single composite exchange rate based on the average of all exchange rates during the five years or during the last of these years, and application of the rate in translations for the whole five years;

  • (3) five exchange rates reached by averaging exchange rates for each of the five years, and application of each average rate in translations for the year to which it related; and

  • (4) the exchange rate as of the date of the most recent balance sheet, and application of the rate in translations for the last year as the only year for which there would be translations.

Some critics, discontented with all possible methods, proposed the elimination of convenience translations.

On December 2, 1981 the SEC issued for public comment proposed new rules and amendments on the choice of foreign currency and exchange rate.34 The SEC explained that because of the disappearance of par values for currencies it was necessary to take new steps to minimize the risk of misleading American potential investors. The foreign currency for convenience translations should be the one that best depicted the effect of the financial activities of a foreign entity in the economic environment in which the entity realized its cash flows. Normally, that currency would be the one in which the primary books and records were maintained, in which the entity conducted most of its business, and in which the financial statements were reported in the country of the entity’s domicile.

Fluctuating and volatile exchange rates, as well as sustained high inflation, had made it likely that the use of a single exchange rate for convenience translations would distort performance data included in the primary financial statements. In recognition of these conditions, deviations from method (1) above had been permitted in a few cases in the practice of the SEC. Methods (3) and (4) had been used in some instances, and also a variant of method (4) in which the exchange rate at the end of each year had been used in translating the data for that year. Some private issuers had favored method (3) because it acted, in part, as an inflation deflator. The Commission thought that exchange rates were determined by many factors, and not only by relative rates of inflation. Adjustment for inflation should be dealt with separately from convenience translations. Furthermore, convenience translations, whatever method was used, obscured or rendered more difficult the analysis of an entity’s performance and of the trends in performance.

The SEC thought that each of the possible methods of translation had advantages and disadvantages, but that no single method provided a comprehensive solution, and all might produce distortions in performance data for a particular period or in trends as reflected in the primary financial statements. The Commission’s concern was heightened because American investors rely heavily on the U.S. dollar statements.

Under the Rules and Regulations published on December 2, 1982, a foreign private issuer must present its primary financial statements only in the currency of the country in which the issuer is incorporated or organized, but a different currency may be used if all the following conditions are met:

  • (a) The different currency is the currency of the primary economic environment in which the operations of the issuer and its subsidiaries are conducted. Normally, the currency will be that of the environment in which the issuer primarily generates and expends cash. (The practice of linking or indexing transactions to a particular currency does not mean that this currency is the reporting currency.)

  • (b) There are no material exchange restrictions or controls relating to the different currency.

  • (c) The issuer publishes its financial statements in the different currency for the benefit of all of the issuer’s shareholders.

The currency in which the financial statements are prepared must be disclosed prominently on the face of the statements. Dollar-equivalent financial statements or convenience translations are not to be presented. Nevertheless, an issuer is permitted to present a translation in respect of the most recent fiscal year and of any subsequent period, using the exchange rate as of the most recent balance sheet included in the filing, provided, however, that an exchange rate as of the most recent practicable date is to be used if the rate is materially different. (The solution of the exchange rate problem is a combination of the approach mentioned above that eliminates convenience translations and a proviso in favor of the fourth method, subject to modification for material changes in the exchange rate since the date of the most recent balance sheet.)

If the financial statements (1) are denominated in the currency of a country that has experienced cumulative inflationary effects exceeding a total of 100 percent over the most recent three-year period and (2) the statements have not been recast or otherwise supplemented to include information on a constant currency or current cost basis prescribed or permitted by appropriate authoritative standards, the issuer must present supplementary information to quantify the effects of changing prices on the issuer’s financial condition and the results of its operations.

Selected financial data must be disclosed in all filings, including:

  • (i) in the forepart of the document the exchange rate for the U.S. dollar, as of the latest practicable date, of the foreign currency in which the financial statements are denominated;

  • (ii) a history of exchange rates for the five most recent years and any subsequent interim period for which financial statements are presented, setting forth the rates for period-end, the average rates, and the range of high and low rates for each year; and

  • (iii) if equity securities are being registered, a five-year summary of dividends per share stated in both the currency in which the financial statements are denominated and U.S. dollars based on the exchange rates at each respective date of payment.

For the purposes of this rule, the rate of exchange is the buying rate at noon in New York City for cable transfers in foreign currencies as certified for customs purposes by the Federal Reserve Bank of New York. The average rate is the average of the exchange rates on the last day of each month during the year.

The Commission’s rules mean that, as convenience translations can he misleading, primary financial statements must be in the appropriate foreign currency without supplemental convenience translations. Nevertheless, information must be supplied about exchange rates of the foreign currency against the U.S. dollar. The information includes the averaging of these exchange rates over an extensive period. But the SEC seems to have assumed that the exchange rate on a particular recent day may be of equal or greater interest to investors than the average of exchange rates over a long period.35

1

Paragraphs 3.18–19, 6.2(4)(b).

2

See Joseph Gold, Special Drawing Rights: Character and Use, IMF Pamphlet Series, No. 13 (Washington: International Monetary Fund, 1970), pp. 35–47, 50–54.

3

Executive Board Decision No. 5590-(77/ 163), December 7, 1977, effective April 1, 1978, Selected Decisions, Fifteenth Issue, pp. 228–29.

4
See Article V, Section 3(e). Under Article XXX(f):

“A freely usable currency means a member’s currency that the Fund determines (i) is, in fact, widely used to make payments for international transactions, and (ii) is widely traded in the principal exchange markets.”

5

Rule P-6(a) of the IMF’s Rules and Regulations.

6

Rules 0-1 and 0-2 of the IMF’s Rules and Regulations.

7

Executive Board Decision No. 6709-(80/189) S, December 19, 1980, Selected Decisions, Fifteenth Issue, pp. 353–54.

8

Rule 0-2(b) of the IMF’s Rules and Regulations.

9

European Communities, Official journal of the European Communities (Luxembourg), No. C 69/4–5 (March 13, 1979); Commission of the European Communities, European Economy (Luxembourg), No. 3 (July 1979), p. 101.

10

If a repurchase obligation that accrued in the appreciated currency was calculated on the basis of the par value of the currency, the repurchasing member would have been unhappy because of the exchange loss it would suffer. The consequence then might be defaults in the discharge of repurchase obligations.

11

United Nations, General Assembly, A/RES/43/165, January 17, 1989. See Gerold Herrmann, “Background and Salient Features of the United Nations Convention on International Bills of Exchange and International Promissory Notes,” University of Pennsylvania journal of international Business Law, Vol. 10, No. 4 (1988), pp. 517–77.

12
Compare Article VII, paragraph 1 of the Agreement between the Philippines and the United Kingdom for the Promotion and Protection of Investments, done at London, December 3, 1980 and entered into force January 2, 1981 (International Legal Materials (Washington), Vol. 20 (1981), pp. 327–28):

“Each Contracting Party shall in respect of investments permit nationals or companies of the other Contracting Party the free transfer of their capital and of the earnings from it, subject to the right of the former Contracting Party to impose equitably and in good faith such measures as may be necessary to safeguard the integrity and independence of its currency, its external financial position and balance of payments, consistent with its rights and obligations as a member of the International Monetary Fund.”

13

Ibid., Vol. 24 (1985), p. 541.

14

Ibid., Vol. 20 (1981), p. 328. See also ibid., Vol. 29 (1990), p. 342. Under Egypt’s Law No. 230 for 1989, however, the re-export of invested funds “shall be effected in five equal annual instalments, at the highest declared rate of exchange within their value at the time of liquidation or disposal as the case may be,” but in some circumstances a transfer in one total amount is possible (ibid., Vol. 29 (1990), p. 249). Under Article 5(2) of the Treaty between Germany and the U.S.S.R. on the promotion and reciprocal protection of investments, transfers “shall be effected in accordance with customary banking procedures without unwarranted delay, but no later than three months from the date of application, at the exchange rate in effect on the date of the transfer” (ibid., p, 358, and see also p. 373). For provisions on the exchange rate applicable to other aspects of foreign investment in various countries, see ibid., pp. 265, 284, 292.

15

Ibid., Vol. 21 (1982), p. 937.

16

See Ministere Public v. R. Lambert (Case 308/86) [1989] 3 C.M.L.R. 649.

17

International Legal Materials (Washington), Vol. 21 (1982), p. 403.

18
See, for example, Article 6(2) of the Agreement between Belgium-Luxembourg and the U.S.S.R. of February 9, 1989 on the promotion and reciprocal protection of investment (International Legal Materials (Washington), Vol. 29 (1990), p. 304):

“The transfers prescribed in paragraph 1 are effected at the exchange rate applicable on the date of transfer and by virtue of the exchange regulations in force in the area in the State in whose territory the investment was made.”

Under Article 5 of the similar Agreement of July 4, 1989 between France and the U.S.S.R., it is provided, however, that (ibid., p. 326):

“The transfers envisaged in the preceding paragraph shall be realized without delay at the prevailing market rate of exchange on the date of transfer.”

19

MIGA, General Conditions of Guarantee for Equity Investments, Article 16.5 (January 25, 1989); see International Legal Materials (Washington), Vol. 28 (1989), pp. 1233–81 for the model contract of guarantee and the General Conditions of Guarantee for Equity Investments.

20

If the practice is authorized by Article XIV, the approval of the IMF is required under Article VIII, Section 3 only if the member changes the practice or clarifies that it does not regard the practice as necessary for balance of payments reasons.

21

[1976] 3 All E.R. 851. Joseph Gold, The Fund Agreement in the Courts, Volume II (Washington: International Monetary Fund, 1982), pp. 236–41.

22

[1976] 3 All E.R., at p. 862.

23

See, for example, Andean Code on the Treatment of Foreign Capital, etc., Article 9 (International Legal Materials (Washington), Vol. 27 (1988), p, 981); Treaty of Amity and Cooperation in Southeast Asia, Article VII (ibid., p. 614).

24

Section IV, paragraph 1.28 of MIGA’s Operational Regulations, International Legal Materials (Washington), Vol. 27 (1988), p. 1235.

25

Joined Cases 64 and 113/76, 167 and 239/78, 27, 28, and 45/79 [1982] ECR 1733.

26

[1979] ECR 3091.

27
The second sentence of Article 215 of the Treaty of Rome provides that:

“In the case of non-contractual liability, the Community shall, in accordance with the general principles common to the laws of the Member States, make good any damage caused by its institutions or by its servants in the performance of their duties.”

28

Note the practice by which the equivalent in ECUs of the deposits of U.S. dollar holdings contributed by a member state of the Community in return for ECUs is valued according to the market rate prevailing two working days prior to the value date for the quarterly renewal and adjustment of the deposit. The number of ECUs provided in return for deposits of gold, however, is determined by the average of the prices, translated into ECUs, recorded daily at the two London fixings during the previous six calendar months. But some safeguard was adopted against a currently unrepresentative price for gold as the result of averaging. The average price of the two London fixings on the penultimate working day of the period preceding an adjustment is applied if it is less than the average price over the period.

In Nachi Fujikosthi Corporation v. E.C. Council (Case 255/84) [1989] 2 C.M.L.R. 76, the European Court of Justice decided a number of questions concerning the Community’s antidumping rules. The normal value and the export price were expressed in different currencies, and the comparison between them had to be made by translating the yen into the relevant Community currencies in order to calculate the dumping margin. The court rejected the exporter’s argument that the comparison should have been calculated on the basis of purchasing power parities. The court rejected this contention on the ground that the effect of imports at dumped prices on Community industry had to be assessed in the light of the official exchange rates on the basis of which international transactions took place. The court approved the averaging of these exchange rates for the yen by the Community’s institutions in calculating the dumping margin (pp. 11819).

29

Compare, for example, the Treaty for the Establishment of the Economic Community of Central African States, done at Libreville, October 9, 1983. Article 8, paragraph 1 of Protocol VIII on the Clearing House for the Community provides that all transactions of the Clearing House shall be expressed in Community units of account as determined by the Exchange and Payments Committee. According to paragraph 2: “When the Committee considers that the economic situation makes it impossible to use fixed exchange rates for defining the unit of account, the effective exchange rates on the currency markets concerned shall be used.”—International Legal Materials (Washington), Vol. 23 (1984), p. 992.

30

Basic Documents of Asian Regional Organizations, Vol. 6, ed. by Michael Haas (Dobbs Ferry, New York: Oceana Publications, 1979), pp. 174–201. Settlements are made in U.S. dollars or in “any other mutually acceptable currency” at the exchange rate between the AMU and the currency of settlement on the last day of the accounting or settlement period (ibid., pp. 200–201).

31

Executive Board Decision No. 6631-(80/145) G/S, September 17, 1980, Selected Decisions, Fifteenth Issue, pp. 351–52.

32

Federal Register (Washington), Vol. 47, No. 234 (December 6, 1982), pp. 54766–68.

33

For example, Roy L. Brooks, “Currency Translations in the Registration Statements of Foreign Issuers,” The Business Lawyer (Chicago), Vol. 35 (January 1980), pp. 435–54.

34

Federal Register (Washington), Vol. 46, No. 231 (December 2, 1981), pp. 58507–11.

35

In the tax practice of the United Kingdom, the appropriate rate of exchange for the translation of foreign currencies to or from sterling is normally the rate of exchange prevailing at the time when conversion occurred or when the relevant transaction resulted in a realized gain or loss. “However, there are some cases when this would clearly be inconvenient or impractical, and the Revenue will accept the use of average or closing rates.”—Jill C. Pagan, Taxation Aspects of Currency Fluctuations (London: Butterworths, 1983), p. 10. Ms. Pagan states that each year the Inland Revenue issues a list of average exchange rates against the pound sterling for numerous currencies. The average for a year is calculated both to December 31 and March 31 (ibid.). See also Organization for Economic Cooperation and Development, Tax Consequences of Foreign Exchange Gains and Losses, Issues in International Taxation, No. 3 (Paris, 1988), paragraph 82, p. 28.