The Monetary Law of Interstate Obligations


A. Introduction


B. Determination of the Money of Account








C. Nominalism and Treaties


Application of principle


Effect of catastrophic depreciation


Post-maturity depreciation


D. The Calculation of Damages


E. Protective Clauses


F. The Payment of Interstate Debts


Place of payment


Manner of performance


The money of payment


The rate of exchange


The type of rate of exchange


G. Defences to International Monetary Obligations


H. The Effect of Exchange Control


A. Introduction


Monetary obligations arising under public international law will normally involve obligations arising between international persons (such as States or organizations created by a treaty made between States). Although the respective contexts may be very different, the types of difficulty with which the lawyer will be concerned are likely to be similar to those which arise under a domestic contract—for example, what is the money of account; what is the money of payment; what interest rate is to apply; may late payment result in liability for interest, damages, or other remedies?


But two complications must be borne in mind. First of all, international practice in this field has developed over a long period of time and it is difficult to discern any uniformity or consistency of approach; it is thus difficult to formulate with confidence any applicable rules of customary international law. Secondly, the international legal order lacks any currency of its own, with the necessary result that questions touching both the measure and the means of performance of international obligations require recourse to national currencies. The resulting interplay of international and municipal law creates problems which involve peculiar difficulties.


The absence of an international currency caused limited practical difficulty where international law could resort to the device of adopting an independently defined unit of account. Thus, during the first seven decades of the twentieth century, numerous multilateral treaties were concluded on the basis of the gold franc of 100 centimes weighing a gramme and of a fineness of 0.900,1 or the gold franc containing 65½ milligrammes of gold of a fineness of 900/1,000,2 or of European Monetary Units of account.3 In these and similar cases,4 the definition is identical with that of a national unit of account as constituted at the material time, ie the French franc or the US dollar, yet by incorporating the full definition in their text these treaties achieved more than they could have done by a mere reference to the national unit of account or the mere adoption of a gold clause; they effectively created an independent monetary system which could not be affected by purely domestic legislation. As a result of the ‘demonetization’ of gold and the volatility of its value, references in most of these cases have been replaced by references to Special Drawing Rights (SDRs)5 or European Currency Units (ECUs),6 with the result that international practice works with monetary standards which are even more susceptible to fluctuations in value.

B. Determination of the Money of Account



As noted previously,7 it may occasionally be difficult to ascertain the money of account in the context of a private law contract; if the parties have referred to ‘dollars’, does this refer to US dollars or to Canadian dollars, or to one of the several other currencies which use that name? When entering into treaties, States will usually take care of questions of this kind. In those few cases in which the money of account has not been defined with sufficient clarity, the intention of the parties will have to be deduced from the construction of the treaty or from the circumstances attending its conclusion. As in the case of private contracts, it is difficult to offer any general rules in this area. However, the following comments may offer some guidance in cases of doubt:

(a) There is no general presumption that the money of account is that of the country in which the treaty is signed. Thus, where a treaty was executed in Switzerland, the fact that France was a signatory (whilst Switzerland was not) suggested that references to ‘francs’ involved French francs, rather than the Swiss unit.8

(b) Likewise, it has on occasion been suggested that debts arising under public international law are expressed in the currency of the creditor State. However, this appears to lack any foundation and would in any event be of no assistance where debits and credits may accrue separately to the parties throughout the life of the agreement.

(c) The purpose of the payment may provide an indicator of the money of account. If the payment is intended to meet the cost of maintaining the seat of an international organization, this may tend to suggest that the money of account is the currency of the country in which the seat is located. Thus, in the preeuro era, a requirement to contribute to administrative expenses in ‘francs’ would probably have been a reference to French francs if the headquarters were located in Paris; or to Belgian francs if the seat of the organization was situate in Brussels.


Attempts at clarification have occasionally had the opposite effect. Under the terms of a 1967 treaty,9 France undertook to pay a sum of ‘163 million French francs (132 million DM)’ and Germany undertook to pay ‘43 million French francs (35 million DM)’. It would seem that the French franc was the money of account, and that the reference to Deutsche marks was solely for information purposes. The European Court of Justice had to decide a similar problem, in a case in which it had expressed a fine in terms of ‘80,000 units of account (FF 444,235.20)’. The Court found that the French franc was the money of account for the purposes of its order, and that it had merely wished to demonstrate that the amount of the fine had been derived from the unit of account used by the Community for budget purposes.10


There are relatively few cases in which an international tribunal has had to determine the money of account in the context of a debt claim. One case11 arose out of events in April 1941, where Greek vessels had taken on various cargoes from the United States and the United Kingdom—the former having been purchased for US dollars and the latter for sterling. The vessels were then diverted to ports outside Greece, where their cargoes were taken over by the British authorities for use against the common enemy. In February 1942, Greece and the United Kingdom agreed12 that the Greek Government would prevent the cargo owners from making a claim against the British Government, and that the latter would credit the Greek Government with the f.o.b cost of the cargoes. The parties were agreed that, even in respect of goods purchased in the United States, credit was to be given to Greece in terms of sterling.13 The issue was the rate of exchange at which the f.o.b. value of the goods of American origin was to be converted into sterling. Not unnaturally, the British Government contended for the pre-September 1949 rate of US$4.03 to £1, whilst the Greek Government argued for the post-September rate of US$2.80 for £1. Now, no question arose either as to the money of account or as to the money of payment, for both were admittedly sterling. The case thus involved only the question of which rate of exchange was to be employed, and this was plainly a question of the construction of the treaty concerned. The arbitrator concluded that the treaty of 1942 ‘created a single account in a single currency, and it was a credit in pounds sterling that the British Government undertook to give … to the exclusion of any other currency’. The award in this case is thus authority for the (perhaps self-evident) proposition that the determination of the money of account and the identification of any required rate of exchange are, in public international law no less than in private law, a matter of construction.14


The identification of the money of account in the context of an international obligation assumed a particular relevance in investment cases arising in the wake of the Argentine economic crisis and the resultant emergency legislation introduced in the course of 2002. In National Grid plc v Argentine Republic,15 local subsidiaries of National Grid held a concession to provide electricity in Argentina. Both the concession and the official tariffs stipulated for payment in pesos. However, the contracts provided for calculation of the tariffs by reference to US dollars. Since the remuneration intended to be earned under the concession was thus linked to a US dollar valuation, that currency (rather than the peso) had to be taken into account in determining the compensation payable by Argentina in that particular case.



The problem of determining the money of account in which unliquidated damages are to be expressed is familiar to international practice.16 In some cases, the Convention creating an international tribunal has specified the currency to be employed for the assessment of damages. Any ambiguity surrounding the precise meaning of any such treaty provision must, of course, be resolved by a process of construction. In the absence of any express directions, tribunals must search for other indications in the relevant treaty. A process of construction is still involved, albeit perhaps from a different starting point. In the Mexican arbitrations, those indications were frequently found in the provisions according to which any balance due from one to the other Government after the disposal of all claims shall be paid ‘in gold coin or its equivalent to the Government of the country in favour of whose citizens the greater amount may have been awarded’.17 In view of this provision, the United States–Mexico General Claims Commission adopted the practice of rendering awards in US dollars, apparently on the basis that this would avoid any future uncertainties with respect to the rate of exchange and was consistent with the underlying purpose of Article IX of the Convention concerned.18 And yet, this arbitrary approach necessarily also involved an exchange operation (or calculation) and gave rise to difficulties in other contexts.19 It may be added that this practice was not followed by the other Mexican Claims Commissions, all of which awarded damages in gold pesos.20


Where the tribunal can find no express or implicit guidance in the Convention from which it derives its existence, then a different approach becomes necessary. As a starting point, the award of damages should in principle be made in a freely transferable and convertible currency, since this is consistent with the objective of affording full reparation to the injured party.21 In The Wimbledon,22 Germany had wrongfully refused to allow a French vessel to pass through the Kiel Canal. The Permanent Court of International Justice determined that damages should be paid in French francs because ‘this is the currency of the applicant in which his financial operations and accounts are conducted and it may therefore be said that this currency gives the exact measure of the loss to be made good’. Similarly, in the Corfu Channel case,23 the International Court of Justice assessed damages in sterling where the claimant State (the United Kingdom) was awarded damages for the loss of a destroyer. This principle appears to be sound, both in the sense that it seeks to identify as closely as possible the losses suffered by the claimant, and in that the principle has a parallel in private law.24 Nevertheless it cannot be said that this principle has been consistently applied; furthermore, it cannot be appropriate to every case and, consequently, is not of general validity.25 Thus the value of a house destroyed in the course of a rebellion, and for the loss of which the respondent Government has to indemnify the owner, can only be assessed in terms of the currency of the country in which the property is situate.26 Similarly, when Greece wrongfully took over lighthouses operated by a French firm following the premature termination of a concession, the Arbitration Tribunal was fully justified in determining the value of the concession by calculating the annual profit in terms of the Greek drachma;27 both the income derived from the concession and the expenses incurred in earning it would have been expressed and paid in that currency.28


The various views just expressed would appear to be consistent with the International Law Commission Articles on State Responsibility (2001), although those Articles do not explicitly deal with the money of account in this context. Where a State is responsible for an international wrong, it is placed under ‘an obligation to compensate for the damage caused thereby, insofar as such damage is not made good by restitution’.29 Any such compensation ‘shall cover any financially assessable damage including loss of profits insofar as it is established’.30 This language makes it plain that a State which improperly terminates a concession agreement may be ordered to compensate the concessionaire, and that the damages must include an assessment of the value of any likely future profits over the unexpired period.31


The process of identifying the money of account may be assisted if it is remembered that compensation is awarded where (and to the extent to which) restitution is not available for some reason. In practice, monetary compensation may well be the preferred solution of the parties. It may be the only available solution in others, for example, where the claim relates to the destruction of property. But the primary status of the requirement for restitution should not be overlooked.32 Article 35 of the ILC Articles records the obligation of a State responsible for an intentionally wrongful act ‘to make restitution, that is, to re-establish the situation which existed before the wrongful act was committed, provided and to the extent that restitution … is not materially impossible’. Herein lies the clue; compensation must, as nearly as possible, place the claimant in the position in which it would have been, had restitution been possible. The compensation must be fixed so that it will correspond to the value which restitution would otherwise have provided; where appropriate, this may mean that different heads of damage may be compensated by awards in different currencies and each currency award may bear interest at different rates.33


As a result, where the claim relates to real property which has been wrongfully destroyed or confiscated, its value will initially have to be ascertained in the currency of the country in which the property was situate.34 Where the claim relates to moveable property, however, a more flexible approach seems to be necessary; it should perhaps be valued in the currency of the country in which such chattels were most frequently used. In the case of an improperly terminated concession or similar agreement, restitution would have required both parties to continue with the performance of their respective obligations; consequently, the compensation payable to the disappointed concessionaire should be calculated in the currency or currencies in which his profits would have been accrued under the contract.35 Equally, in investment cases, the tribunal will often award damages in the currency of the host State where that is specifically requested by the claimant.36


Yet there may be countervailing considerations. For example, as noted elsewhere,37 the principle of full compensation may require that the award be made in a freely transferable currency. In line with these principles, the ICSID tribunal in the Vivendi case38 awarded damages for loss of a concession in US dollars (the currency in which the investment was originally funded) rather than Argentinian pesos (the currency in which the income was to be generated). The decision was clearly influenced by the significant depreciation of the peso vis-à-vis the dollar. Similarly, where an investment had been funded in a combination of pounds, German marks, US dollars, and Ghanaian cedis, the tribunal made awards in the hard currencies but, in order to satisfy the compensatory principle discussed earlier in this paragraph, the portion of the investment made in cedis was compensated by an award expressed in US dollars.39 The conversion of the cedi amount into US dollars was effected by reference to the exchange rate as at the date of the expropriation (rather than as at the date of the earlier investment) because the Government of Ghana was not expected to underwrite the value of its own currency. At first sight, this appears to be at odds with the principle that the claimant should not be prejudiced by a currency depreciation between the date of the wrong and the date of the award.40 However, the true principle appears to be that the investor bears the risk of a currency depreciation up to the date of the expropriation or other wrongful act, whilst he is protected from such depreciation occurring between that date and the date of the award.41 The result is that the rate of exchange in effect as at the date of the wrong should generally be applied in calculating the award, at least if the claimant might have been expected to convert the compensation into his own currency on receipt. This is a logical approach, in that the claimant should not suffer further losses following the date on which the wrong occurs and with reference to which compensation should be paid.42 Where, however, a currency merely suffers ‘normal’ (as opposed to dramatic) depreciation, then this factor may be left out of account on the basis that an award of interest will offer sufficient compensation.43


It should not, however, be overlooked that this discussion is concerned with the calculation of compensation, ie with the money of account. The money in which such compensation must be paid (ie, the money of payment) is considered below.44



Questions touching the payment of interest may arise in a variety of ways. A treaty may explicitly provide for the payment of interest on amounts owing under its terms.45 The money of account will plainly be a matter of construction in such a case; in the absence of any explicit statement, it may perhaps be inferred that interest accrues in the currency in which the relative principal sum is outstanding.


Where a State seeks compensation in respect of an internationally wrongful act of another State, then, under the terms of Article 38 of the International Law Commission’s Articles:

(a) interest shall be awarded when this is necessary to ensure full reparation;

(b) if interest is awarded, then it should run from the date when the relative principal sum should have been paid until the date on which it is actually paid; and

(c) the rate of interest and the mode of its calculation shall be fixed with a view to ensuring full reparation for the claimant.46


In general terms, a claimant State is entitled to interest if the compensation awarded to it is assessed (say) by reference to the value of property as at a date which precedes the award, but not if the value is assessed as at the date of the award itself.47 Where interest is payable, it will generally be necessary for the rate to reflect the likely cost to the claimant of borrowing the relative principal amount with effect from the date on which it ought to have been paid, for only in this way can the principle of full reparation be satisfied.48 The funding cost will usually have been met in the currency in which the relative principal amount was owing, so that the money of account for interest will ‘follow’ the money of account for the primary claim. There is, however, a lack of consistency in the approach adopted as to the place by reference to which the rate of interest is to be ascertained. Some cases have adopted the rate of interest prevailing in the territory of the debtor State,49 whilst others have awarded the claimant a rate of interest reflecting the return on commercial investments in his home country.50 In an investment case, it has been observed that international tribunals have a significant degree of latitude in determining the appropriate rate of interest, taking into account the money of payment, the situation of the respective parties, and all other circumstances.51


Although not directly relevant to the ‘money of account’ issue, it may be noted that, in line with developments under English domestic law,52 international tribunals have begun to recognize that simple interest is not sufficient compensation for being kept out of money for an extended period, and that compound interest should generally be awarded as a matter of course.53

C. Nominalism and Treaties

Application of principle


Treaties providing for the payment of a sum of money have become a matter of almost daily occurrence; for many centuries, treaties of peace have imposed obligations of monetary indemnity upon the vanquished party;54 in more modern times, the TFEU creates many financial obligations, and examples could be multiplied. The present section is therefore concerned with the principle of nominalism and its application to liquidated obligations found in a treaty.


It has been noted elsewhere that the principle of nominalism applies only to liquidated obligations;55 it has no general application to claims which sound in damages. This distinction must apply (for the same reasons and with equal force) in the context of obligations arising under public international law.


This proposition does, however, require further examination. Is there any basis upon which the principle of nominalism could be set aside as a result of some peculiar characteristic of treaties or the identity of the parties which may enter into them? For example, given that the parties will be sovereign States, should it be assumed that monetary references were intended to be insulated from the vagaries of national legislation, such that a gold clause or similar protective provision should be implied? It is suggested that there was never any basis for the implication of any such clause; had the parties so required, they could have stated it expressly. There is certainly no room for the implication of a gold clause into treaties entered into following the collapse of the Bretton Woods system of parities.56


Under these circumstances, the principle of nominalism must apply to liquidated obligations arising under public international law. It follows that the debtor State is bound (and entitled) to pay the nominal amount of the agreed currency, irrespective of its intrinsic value—in terms of purchasing power or some other currency—as at the date of payment. By contracting on the footing of a specific national currency, States incorporate into their treaties the monetary legislation of the country concerned and, to that extent, those treaties contain a renvoi to the lex monetae. The problem is always one of construction; when a treaty refers to a national monetary system, the contracting States are aware that such system is created by a municipal system of law, and that monetary obligations can only be defined by reference to that system. To that extent, their treaty necessarily adopts pro tanto that national monetary legislation. In a private contract which is subject to English law and which stipulates for the payment of a sum of foreign money, the system of law which regulates that money effectively becomes a part of the law applicable to the contract.57 It is for this reason that the definition of what the stipulated unit of account means is referred to as the lex monetae by a universally followed rule of the conflict of laws. It is for the same reason that monetary obligations under public international law are subject to the lex monetae in so far as the definition of the unit of account is concerned.


On this basis, the further question arises of whether the reference to the lex monetae envisages (a) the particular law of the currency as it exists at the time of conclusion of the treaty, or (b) such law of the currency as it may be amended from time to time. As has been shown,58 nominalism has the latter meaning in a private law context. Once again, the reasoning which underlies this rule must apply equally to interstate monetary obligations. If the parties have merely referred to a domestic currency without using some indexing or similar protective measure, then they must be taken to have contracted by reference to that currency as a measure of value from time to time. No ‘value maintenance’ or similar provision can be implied in such a case.