1. Stability in terms of currencies

Although the recent interest in the possibility of a “hard” SDR has coincided with the debate surrounding proposals for a “hard” ECU, the idea of strengthening the SDR’s value or making it more stable in some sense goes back several years. Since the inception of the SDR, considerable importance has been attached by member countries of the Fund to its attribute of stability, both in its role as a reserve asset and as a unit of account. The interpretation of the concept of stability, however, has differed, depending on the context and the particular method of valuation being considered for valuing the SDR. The alternative approaches to the definition of stability were explored to some extent in the discussion of the valuation of the SDR in 1973-74 when the basis of valuation was shifted from gold to a basket of currencies. (See IMF, 1973). In a world of constantly changing relative prices of commodities and currencies, a choice had to be made regarding the medium in terms of which the value of the SDR was to be stabilized. Three possible alternatives for choosing the medium in terms of which stability could be defined are: in terms of commodities; of currencies in general; or of a subset of currencies that are regarded as stable. Although other approaches were discussed and debated, the adoption of the so-called standard basket for valuation of the SDR meant that the concept of stability adopted was one in terms of a specified group of currencies, i.e., those included in the SDR basket. This assured the holder of SDRs the same spread of risk as he would have achieved by diversifying his reserve holdings into the currencies in the basket in the proportions used to construct the basket. Thus, if a country’s import composition in terms of currency denomination approximated the currency composition of the basket, the effect would be to insulate the purchasing power of the SDR over imports from exchange rate changes.

The standard basket, however, does not possess the property of stability of value in terms of currencies in a different sense. Thus, an appreciation (depreciation) of any currency in the basket in terms of all other currencies raises (lowers) the value of the SDR in terms of those other currencies. If it were desired to give the SDR a more stable value in terms of its component currencies than that resulting from the standard basket approach, it would be necessary to introduce increases in the amount of each currency in the basket at periodic intervals. This suggested the alternative approach of an asymmetrical basket. Under this approach, the value of the SDR is set equal to that of a basket of currencies, but whenever any currency in the basket is devalued or depreciated, the number of units of that currency in the basket would be increased in proportion to the extent of the depreciation. The rationale behind this procedure was to prevent depreciations of currencies in the basket from reducing the transactions value of the SDR in terms of other currencies in the basket. However, under this approach, an appreciation of any currency in the basket would still raise the value of the SDR in terms of other currencies, as under the standard basket. Thus, the asymmetrical basket would put a floor under the value of the SDR in terms of its basket currencies, but would not prevent its appreciation as a result of a basket currency’s appreciation.

Under the third variant considered at the time, viz. the adjustable basket, the treatment of depreciations of currencies in the basket under the asymmetrical basket approach would be applied to appreciations of currencies as well, thereby preventing the exchange rate changes associated with all appreciations or depreciations of the currencies in the basket from influencing the value of the SDR in terms of other currencies. ^1/ Although the standard basket was preferred over the other approaches on account of its advantage, above all, of simplicity and understandability, the idea, inherent in the asymmetrical basket, that the value of the SDR could or should be protected from erosion or even be strengthened from time to time, resurfaced in various discussions of the valuation of the SDR. Indeed, in the early discussions of the Fund Executive Board on this question, the Japanese Executive Director had supported the view that the value of the SDR should be based “on the strongest currencies,” that is, those appreciating in the exchange markets. ^2/ However, this idea did not find sufficient support at the time.

The principal weakness of the asymmetrical basket was that in the context of a definition of stability in terms of currencies in the basket, it had no operational meaning in a world of floating exchange rates. The notion of an asymmetrical basket was not easy to implement in a world of predominantly floating currencies, because this method of valuation would require a determination of which currencies were appreciating or depreciating, with reference either to par values maintained for at least some currencies in the basket, or to a clearly defined base period from which the relative exchange rate changes would be measured. In a world of floating currencies, no such common benchmark exists. The ease of operation and adaptability of the standard basket to a floating exchange rate regime probably proved to be its most important attribute which resulted in its adoption in preference to the other alternatives considered. Indeed, with stability understood to mean stable value in terms of an average of its component currencies, the standard basket alone offered an unambiguous and consistent valuation of the SDR irrespective of the nature of the exchange rate regime. Among the alternative basket methods considered at the time, it was the only method that offered a clearly defined concept of stability in value over time in terms of “an average value of currencies” and the opportunity of adopting a formula for the SDR interest rate which was guaranteed to preserve the yield on the SDR constant relative to that on currencies in general.

2. Stability in terms of commodities

The principal concern about valuing the SDR in terms of the standard basket was that an asset that performed with the “average” of currencies would not be sufficiently attractive to establish itself in an inflation-prone world. If the SDR could not be made stronger through an asymmetrical basket, some other means, such as a commodity link, was considered by some as a promising route. The idea of establishing a stable value for the SDR in terms of commodities rather than in terms of a group of currencies had obvious attractions because reserves are held in part to enable countries to finance temporary deficits in their balance of payments, and therefore, the ability of reserves to finance a given volume of imports was regarded as critical. Despite this strong argument for stabilizing the SDR’s value in terms of its purchasing power over commodities, the extent to which a reserve asset like the SDR should maintain its value in terms of changes in price levels was not explicitly addressed at the time the SDR was negotiated. In the Articles of Agreement, the unit of value of the SDR was stated in terms of a weight of gold, which was equal to the par value of the U.S. dollar. ^3/

The idea of establishing the SDR’s value in terms of a group of commodities rather than a group of currencies found its most concrete expression in the proposal for an international commodity-reserve currency. The theoretical antecedent to this idea can be traced to proposals such as that by Benjamin Graham (see Hart, 1966) to define the U.S. dollar in terms of a fixed weight basket of 23 commodities and for the Federal Reserve to issue notes against warehouse receipts for the basket thus defined. He selected his commodities on the strength both of their economic importance and their storability, the list varying from coal to wood pulp. The proposal, made in the 1930s, was motivated by a desire to counter the depression, in the belief that support for commodity prices would help stabilize overall economic activity. By the same token, sale of the commodity basket was expected to limit inflationary pressures in a boom, both by supplying commodities out of stocks and by contracting the money supply. The idea of a “tabular” standard, advocated by Irving Fisher, is the logical extension of the commodity-reserve money to the entire basket of goods and services in the context of a national economy.

In the mid-1960s, the idea of a commodity-reserve currency was revived in an international context by Hart, Kaldor, and Tinbergen (1964). This proposal has been updated for the post-SDR world by Hart (1976). He suggested defining an international reserve asset such as the SDR as a basket of standardized and storable commodities, with the Fund purchasing the basket of commodities, aiming at an amount equal to, say, 25 percent of annual world trade in these commodities and issuing SDRs in exchange. The SDRs so issued would be the principal source of new international reserves. Hart offered an illustrative list of 31 commodities that could make up the basket. The SDR over time would maintain a stable value in the commodity basket through purchases or sales of the basket, within a margin of plus or minus 5 percent. The proposal thus involved the creation of an international currency anchored in a basket of economically significant commodities.

Although enjoying consistent support from some academic economists, the idea of a commodity reserve currency has had little official support. Its obvious attraction is that it would make the value of the SDR stable in terms of a group of commodities and thus insulate it from erosion in real value. The principal objection to it has been that in the absence of a stock of the commodities comprising the basket, the continuing convertibility of the money denominated in this commodity reserve unit into equivalent purchasing power could not be guaranteed. ^4/ The storage and turnover costs would also be significant and would be economically wasteful.

If the objective is to stabilize the SDR’s value in terms of goods and services in general, then the approach of defining the SDR as a basket of specific commodities would not be appropriate mainly because of a lack of sufficient correspondence between the prices of a selected group of commodities and the general price level in the world economy. For instance, as noted by Cooper (1989), a prominent supporter of an international commodity-reserve standard, the price index for the commodities included in the early commodity-reserve basket proposals rose by 177 percent during the period 1947-86, while that for finished manufactures rose by 291 percent, and the price of services in the consumer price index rose by 684 percent. Thus, adoption of the suggested commodity basket to define the SDR would not generally prevent the erosion of its value in terms of a broader inflation index. ^5/

3. Stability in terms of purchasing power

As noted above, at the time of the redefinition of the SDR’s value in 1973 after the link between the U.S. dollar and gold was broken, there was some discussion of the purchasing power of the SDR and its capital value. The possibility of giving the SDR “a stronger capital value” than would result from the alternative basket approaches by introducing small proportional increases in the amounts of the basket currencies at regular intervals was mentioned in the final report of the Committee of Twenty in the context of its wide-ranging exercise on the reform of the international monetary system. It was suggested that the effects of the resulting capital appreciation of the SDR against the basket currencies on the yield of the SDR could be offset by lowering its interest rate (see Section IV.2. below). The discussion in the Committee’s report, however, did not suggest any link between the rate at which the basket currencies lost purchasing power over goods and services and the magnitude and frequency of the increases in the currency content of the SDR. The idea of a “stronger capital value” of the SDR was thus a precursor of the “hard” SDR defined in terms of purchasing power over goods and services, but without an explicit correspondence between the “strength” of the SDR and the “strength” of the individual basket currencies.

The idea of a real SDR, viz. one with constant purchasing power over goods and services was discussed in the subsequent deliberations of the Executive Board on the valuation of the SDR in 1973-74. However, the predominant view, especially on the part of some of the Executive Directors representing large industrial countries was that indexation of the SDR in any form was not advisable. The general objection to indexation, in any form, strongly voiced at the time by the Federal Republic of Germany, was that it would weaken efforts directed against inflation by reducing some of its ill effects, albeit in an incomplete fashion (see de Vries, 1985).

In its role as a unit of account in international transactions, some consideration was however given to a “real” SDR, especially for expressing limits of liability or similar values intended to maintain a constant purchasing power over goods and services. In the past, such values had often been stated in terms of a gold unit such as the Poincare franc. The Secretariat of the United Nations Commission on International Trade Law (UNCITRAL), was interested toward the end of the 1970s in an internationally recognized unit of constant purchasing power and requested the Fund for a study of the SDR in this role. The Fund staff recommended the adoption of a unit of value established on the basis of the SDR basket of currencies in combination with an international price index composed of national price indices of the basket currencies, with weights corresponding to the weights of the respective currencies in the basket. The UNCITRAL subsequently adopted (in 1982) the SDR as a unit of account for expressing monetary values in international conventions. ^6/ The inflation-corrected SDR was not adopted but was endorsed as one alternative for expressing values whose purchasing power is intended to be protected.

Thus, ever since its establishment, there has been recurrent interest in the idea of a “hard” SDR although the precise form this would take has not been fully articulated by its advocates, and several misgivings--in particular, those centered on the advisability of any form of indexation--have generally prevented a full exploration of its merits. In contrast to this ambivalent attitude toward the SDR within the international monetary establishment, official and semiofficial international bodies have from time to time showed interest in some version of an SDR of constant value so as to be able to use it in international contracts of various types as an objective and universally recognized standard of constant purchasing power (see Effros, 1982).

1. The “hard” ECU proposal

The U.K. authorities announced their proposal for a “hard” ECU in June 1990, representing an elaboration of their position on the European Monetary Union (EMU) earlier spelled out in a paper by the U.K. Treasury (1990). The U.K. proposal was motivated by the basic premise that the path to EMU should be gradual and evolutionary, that it should be consistent with market forces and should strengthen incentives for price stability.

Under the proposal, the ECU would be changed from a currency basket to a parallel currency with a par value and intervention limits against each of the EC currencies (it being assumed that all EC currencies would be participating in the ERM). A key feature of this new 13th currency in the EC would be that it would never be devalued against any other EC currency. Thus, upon any realignment, the value of the “hard” ECU would move with the strongest currency, instead of moving with the basket.

The proposal envisaged a new EC institution, the European Monetary Fund (EMF), to be set up to manage the “hard” ECU. The EMF would issue “hard” ECUs against surrender of holdings of national (EC) currencies. The proposal seems to envisage that the EMF would initially hold 100 percent of reserves (in the form of national currencies) against its “hard” ECU liabilities, and that it would set the interest rate on its deposits. Thus, initially the EMF would have the ability to change the composition of EC money supply between national currencies and the “hard” ECU, but not the total money supply.

The market was expected to exert an anti-inflationary discipline as firms and individuals switched from currencies with expected poor purchasing power to those with better prospects. To enhance the normal anti-inflationary discipline exerted by the market on those currencies which are expected to depreciate, the EMF could be empowered to require national central banks to repurchase their own currencies at the EMF for “hard” ECUs at the exchange rates prevailing at the time when the currency balances were originally deposited with the EMF. As the market for private “hard” ECUs grew and the EMF acquired credibility, the functions of the EMF could be expanded to resemble those of a national or a supranational central bank, including the ability to engage in open market operations aimed at influencing market interest rates on “hard” ECUs as well as on national currencies of EC members, coordination of members’ exchange market intervention against external currencies and possibly, extension of medium-term balance of payments lending to members. The EMF was also gradually expected to take over from the European Central Bank Governor’s Committee and the European Monetary Cooperation Fund most of the management of the exchange rate mechanism of the EMS and its financing facilities. Nevertheless, the key function of the EMF was envisaged as the monetary authority for the “hard” ECU with responsibility for ensuring that it was never devalued against any of the EC currencies.

The “hard” ECU would be free to vary within the small band of the exchange rate mechanism (ERM) of the EMS against other EMS currencies in line with market demands. It would be the responsibility of the EMF to ensure that the “hard” ECU is never devalued. If this no-devaluation guarantee is taken as fully credible, it follows that the purchasing power of the “hard” ECU would be dissipated by inflation less rapidly, or no more rapidly, than any other EMS currency. It would, however, not mean that the “hard” ECU would provide either stable or predictable purchasing power, as would a constant-value currency, indexed to a cost of living price index. Nevertheless, relative to the existing 12 EC currencies, it would have an advantage regarding its purchasing power over goods and services.

Critics of the proposal have noted that this advantage of the “hard” ECU may not be sufficient to overcome the costs, uncertainties and inertia associated with switching away from an existing domestic currency to the “hard” ECU as a means of payment. This skepticism is based on the historical experience that even in the face of inflation differentials and exchange rate movements that are many times larger than those prevailing, or likely to prevail, among EC countries, residents display a strong affinity to the domestic currency for purposes of means of payments and unit of account. This would not preclude the use of the “hard” ECU by some firms doing predominantly cross-border business or for purposes of denominating travellers’ checks. But the transactions demand for the “hard” ECU as a substitute for the domestic currencies is considered likely to evolve very slowly. In the EMU context, there would also be increased uncertainty due to the uncertain prospects of the “hard” ECU itself becoming the single EC currency and any shifts in its value and characteristics as a result of that process. These considerations lead critics to doubt whether removing existing impediments to the use of “hard” ECUs in the private sector and a no-devaluation pledge might still not be sufficient to stimulate a large shift toward them. A public sector intervention such as making the “hard” ECU a legal tender throughout the EC or a pledge by EC Governments to make it the single currency would help, but such “administered” approaches would be contrary to the competitive spirit of the U.K. authorities’ proposals.

Any reduced inflation or exchange rate risk of the “hard” ECU would presumably be compensated in the market by a lower nominal interest rate on “hard” ECU-denominated interest bearing assets. The growth of the “hard” ECU as a transactions currency itself might spur demand for interesting-bearing “hard” ECU assets over time. It is, nevertheless, possible that, like the present basket ECU, the “hard” ECU may become an important investment currency, despite the absence of a transactions demand. The hardness of the “hard” ECU could give it an edge in making it an even more attractive investment currency, irrespective of the growth in transactions demand.

2. The “hard basket ECU” proposals

The U.K. proposal for a “hard” ECU led to a debate and in turn generated alternative ideas. Although there was considerable skepticism about the idea of creating another currency alongside the existing ECU, the basic idea of hardness had widespread appeal and was reflected in a suggestion by Spain (1991) and Germany, for hardening the existing basket ECU. Under this proposal, the present basket ECU would be revised at any future realignment to ensure that it is never devalued vis-á-vis any of the basket currencies. Thus, the proposal tries to accommodate the concept of hardness within the existing basket structure of the ECU. Hardening in this context refers to the adoption of any measure that would make the value of the ECU less dependent on the weaker, or depreciating, currencies of the EMS. In the Spanish proposal for a “hard basket ECU,” “the amounts of the individual national currencies would be adjusted in the event of any realignment in the ERM, so as to preserve a fixed central exchange rate between the ECU and those currencies not experiencing devaluation.” Similar to the U.K. authorities’ proposal, the Spanish proposal envisaged the creation of a European monetary institution that will not interfere with the conduct of monetary policy by national central banks, but inter alia, “building on existing practices in the private ECU market, will foster its development by establishing a direct link between the official and the private ECU.”

The German authorities also proposed a similar “hard basket ECU,” but without the creation of an official monetary institution that would establish a formal link between the official and the private ECU. The existence of such a link or de facto narrow limits on the permitted deviation between the “hard basket ECU’s” market value and its theoretical value would be important to making the “hard basket ECU” viable for private financial use.

Another variation of hardening the existing ECU was the proposal by the European Commission, which called simply for a “freeze” in the currency composition of the ECU by ruling out future recomposition of the currency basket. This proposal thus envisaged a gradual hardening of the ECU. With the quantities of each EC currency in the ECU fixed, the weight of any appreciating currency in the ECU basket would rise, while that of the depreciating currency would decline.

It is important to note that all the proposals for hardening the existing basket ECU aimed at an ECU that does not lose value with respect to any of the basket currencies at the time of any realignment of parities. This does not rule out a softening of the ECU against one of its basket currencies at any moment in time. This underlines the fact that an application of the idea of an asymmetrical basket discussed in the last section is operationally feasible only in the case of currencies with parities or par values with reference to which any appreciation or depreciation of a currency can be measured. Any attempt to apply the “hard” ECU formula to intra-realignment periods leads to nonsensical results and instability in the ECU’s value. In recognition of this fact, all the proposals to harden the basket ECU have been restricted to realignments, which would provide a benchmark to define a depreciation.

At the Maastricht summit in December 1991, the EC member states agreed to a series of amendments to the Treaty of Rome. Apart from the major institutional innovations leading up to the prospective economic and monetary union, including the establishment of the European Monetary Institute, which would, inter alia, be responsible for the development of the ECU clearing system and facilitate the use of the private ECU, the treaty included a provision saying that “the currency composition of the ECU basket shall not be changed. From the start of the third stage of EMU, the value of the ECU shall be irrevocably fixed.” The European Commission’s proposal to “freeze” the ECU basket was thereby adopted.

1. A yardstick of price stability

The need for a universal unit of account of constant value for purposes of international legal conventions has been felt for some time. Since all currencies in the SDR basket continually lose value in terms of goods and services that they can purchase, the SDR, per se, does not guarantee the maintenance of value in terms of purchasing power. Various groups such as the Committee on International Monetary Law of the International Law Association have called for an international unit of constant value for facilitating international legal conventions.

A major attribute of a universal unit of account that influences the willingness of institutions and other economic agents to use it for denominating obligations is its ability to retain its value in terms of goods and services. This attribute tends to be less crucial for units of account for denominating financial instruments because changes in the purchasing power of such units generally tend to be reflected--and hence compensated for--in the interest rates attached to such instruments. Nevertheless, it seems evident that a universal unit of account of constant value in terms of a basket of all goods and services would be desirable.

The definition of the SDR’s value adopted since 1981 has succeeded to some degree in maintaining the SDR’s real value, because the rates of inflation of the five currencies in the SDR valuation basket have generally been lower than inflation rates in many other member countries of the Fund. However, this has not prevented a steady erosion in the SDR’s value in terms of goods and services (see Chart 1 and Table 1). Thus, the average value of the SDR in terms of goods and services eroded by 75 percent in the period of about 13 years from 1980 to mid-1993, reflecting the weighted average (with SDR basket weights) of consumer price inflation in the five countries whose currencies form the SDR valuation basket. Over a longer period of 19 years beginning in July 1974--when the SDR’s value was first defined as a basket of currencies--it has suffered from an average inflation in terms of the five component currencies of 197 percent. This steady erosion of the real value of the SDR could be a reason behind its limited adoption as a unit of account by the private sector, resulting in turn in a limited demand for SDR-denominated instruments. ^8/

Table 1.

Deflator for the “Hard” SDR

Note: The deflators are derived from consumer price indices in the respective countries. The deflator for the “hard” SDR is derived using the prevailing SDR basket weights.

Table 1.

Deflator for the “Hard” SDR

		United States	United Kingdom	France	Germany	Japan	Deflator for “hard” SDR
July	1974	100.0	100.0	100.0	100.0	100.0	100.00
July	1975	109.7	126.3	111.1	106.2	111.4	111.57
July	1976	115.6	142.4	121.6	110.2	121.9	119.66
July	1977	123.4	167.6	133.9	114.4	131.4	129.83
July	1978	132.9	180.6	146.4	117.5	137.4	138.51
July	1979	147.9	209.0	161.5	122.8	143.3	152.24
July	1980	167.4	244.2	183.1	129.2	154.0	170.45
July	1981	185.4	270.9	207.6	137.5	160.7	187.07
July	1982	197.4	294.6	232.4	144.9	163.9	200.26
July	1983	202.2	306.8	254.3	149.0	167.5	207.97
July	1984	210.6	320.6	273.2	152.4	171.8	216.92
July	1985	218.1	342.4	289.8	155.8	176.0	226.28
July	1986	221.6	350.7	295.7	155.4	176.1	229.53
July	1987	230.3	366.2	305.7	156.2	176.3	236.65
July	1988	239.8	383.8	314.1	158.0	177.2	244.52
July	1989	251.8	415.5	325.0	162.4	182.5	256.58
July	1990	263.9	456.1	334.9	166.3	186.7	269.53
July	1991	275.7	481.4	346.8	173.6	193.2	281.54
July	1992	284.4	499.4	354.3	179.3	196.3	290.02
July	1993	292.2	506.2	361.7	187.1	200.2	297.15

Note: The deflators are derived from consumer price indices in the respective countries. The deflator for the “hard” SDR is derived using the prevailing SDR basket weights.

View Table

Table 1.

Deflator for the “Hard” SDR

		United States	United Kingdom	France	Germany	Japan	Deflator for “hard” SDR
July	1974	100.0	100.0	100.0	100.0	100.0	100.00
July	1975	109.7	126.3	111.1	106.2	111.4	111.57
July	1976	115.6	142.4	121.6	110.2	121.9	119.66
July	1977	123.4	167.6	133.9	114.4	131.4	129.83
July	1978	132.9	180.6	146.4	117.5	137.4	138.51
July	1979	147.9	209.0	161.5	122.8	143.3	152.24
July	1980	167.4	244.2	183.1	129.2	154.0	170.45
July	1981	185.4	270.9	207.6	137.5	160.7	187.07
July	1982	197.4	294.6	232.4	144.9	163.9	200.26
July	1983	202.2	306.8	254.3	149.0	167.5	207.97
July	1984	210.6	320.6	273.2	152.4	171.8	216.92
July	1985	218.1	342.4	289.8	155.8	176.0	226.28
July	1986	221.6	350.7	295.7	155.4	176.1	229.53
July	1987	230.3	366.2	305.7	156.2	176.3	236.65
July	1988	239.8	383.8	314.1	158.0	177.2	244.52
July	1989	251.8	415.5	325.0	162.4	182.5	256.58
July	1990	263.9	456.1	334.9	166.3	186.7	269.53
July	1991	275.7	481.4	346.8	173.6	193.2	281.54
July	1992	284.4	499.4	354.3	179.3	196.3	290.02
July	1993	292.2	506.2	361.7	187.1	200.2	297.15

Note: The deflators are derived from consumer price indices in the respective countries. The deflator for the “hard” SDR is derived using the prevailing SDR basket weights.

Deflator for the Hard’ SDR 1974 JUL = 100

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

Note: The SDR inflation rate is calculated by taking the weighted average of the consumer price inflation in the five basket currencies.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Deflator for the Hard’ SDR 1974 JUL = 100

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

Note: The SDR inflation rate is calculated by taking the weighted average of the consumer price inflation in the five basket currencies.

• Download Figure
• Download figure as PowerPoint slide

Deflator for the Hard’ SDR 1974 JUL = 100

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

Note: The SDR inflation rate is calculated by taking the weighted average of the consumer price inflation in the five basket currencies.

• Download Figure
• Download figure as PowerPoint slide

An SDR with a constant real value would offer the international community a supra-national unit of account of world standing and would spur its use for denominating obligations. Since it would be regularly calculated and published by the Fund, the hard SDR could form the basis for contracts among Fund members or among private contracting parties. The hard SDR would be expected to appreciate in relation to the conventional SDR and against currencies in general. This expectation would make it an attractive store of value. Its expected yield would be competitive with that on currency-denominated assets, although, as discussed below, its nominal interest rate would tend to be relatively low. The increased demand for private SDR could increase the demand for official SDRs as well. A unit of stable purchasing power in which contracts could be written is also likely to facilitate expansion of world trade and improve international resource allocation.

As noted above, the UN Commission on International Trade Law has adopted the SDR as a unit of account and has expressed its view that an international unit of constant purchasing power would be useful. A number of other international organizations use the SDR as their unit of account or as the basis for their unit of account. These include the African Development Bank, the Arab Monetary Fund, the Asian Clearing Union, the International Development Association, the International Fund for Agricultural Development, and the Islamic Development Bank. In addition, various international conventions use the SDR as a standard of value in which to express monetary magnitudes, notably conventions which express limits of liability for carriers in the international transport of goods and persons. In some cases, the SDR has been adopted in these conventions to replace the Poincare or Germinal franc, both of which are defined with reference to fixed quantities of gold. For instance, the unit of account for the Universal Postal Union and the International Telecommunications Union is the gold franc, which has been declared to have a specified relationship to the SDR. Therefore, in practice, international postal and telecommunication accounts are maintained in SDRs. In these and other contexts where a convention or international contract needs a unit of account of stable value, the “hard” SDR would be very useful.

The “hard” SDR would provide a clear yardstick for price stability in the world economy. Therefore, any country which desired the goal of price stability could, in principle, implement such a policy by pegging its currency to the “hard” SDR. The “hard” SDR could thus play a role as an unambiguous international monetary anchor for price stability.

2. Impact on Fund financial structure

As a reserve asset of constant purchasing power, the demand for “hard” SDRs would be expected to increase with the prospect of inflation. This would introduce a new element in the consideration of the desirability of allocating SDRs. It would seem that an allocation of such hard SDRs would be appropriate, despite--indeed on account of--inflationary tendencies in the world economy. Hard SDRs would, over time, tend to be substituted in reserve holdings for currencies of countries experiencing inflation. It is sometimes argued that the “hard” SDR would, through this process of substitution for inflating currencies, tend to exert a disciplining influence on the adjustment process in the world economy. The validity of this view is open to question since the mechanism through which the discipline is exerted is not very transparent. In any case, a “hard” SDR would at least maintain the proportion of SDRs in total reserves in the face of inflation, even if no new allocations were made.

The concern voiced by some in the earlier debates on the hardening of the SDR regarding the advisability of indexing the SDR would need to be addressed. The concern stemmed from the belief that indexation of a reserve asset would imply accommodation of inflation and would remove the discipline on governments. The proposed role of the “hard” SDR as a unit of account of a constant real value, however, does not seem to accommodate inflation. It rather recognizes the fact that inflation, in the real world, erodes the purchasing power of the five basket currencies and offers a standard of value which is free from such erosion.

The “hard” SDR’s attractiveness as an asset whose value is protected from inflation would probably lead to a greater demand for it on the part of Fund members, although its lower interest rate would tend to offset some of its attraction. If a greater demand for it were to materialize, this may necessitate a reassessment of the policy and procedures regarding transactions in SDRs and the optimal size of SDR holdings by the Fund and members.

An important consequence of a “hard” SDR would be for the size of the Fund as measured by quotas. Since quotas are denominated in SDRs, the shift to a “hard” SDR would imply that the real value of the size of the Fund would be maintained even when inflation erodes the values of the SDR basket currencies. As illustrated in Chart 2 and Table 2, if the SDR was protected against erosion by inflation, the size of Fund quotas would have kept pace with the expansion of world trade and output caused by inflation. If quotas are denominated in “hard” SDRs, the general increases in quotas every five years would reflect only the increase necessitated by real increases in world trade and output. These general increases would therefore be smaller than those needed to keep pace with real changes in the world economy as well as the effect of inflation. To a large extent, however, the process of increasing quotas to reflect world inflation would become automatic when quotas are denominated in “hard” SDRs. The difficulties associated with the protracted process of consent to general increases in quotas would therefore be considerably alleviated by the adoption of a “hard” SDR for denominating Fund quotas. The need for general increases to allow for real growth in the world economy would, however, remain.

The Value of Fund Quotas Fixed in ‘HARD’ SDRs Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

The Value of Fund Quotas Fixed in ‘HARD’ SDRs Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

The Value of Fund Quotas Fixed in ‘HARD’ SDRs Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

Table 2.

Fund Quotas - Actual and in “Hard” SDR

(In billions of SDRs)

Table 2.

Fund Quotas - Actual and in “Hard” SDR

(In billions of SDRs)

Year	Actual quotas	Hard SDR-denominated quotas
1974	28.7	30.0
1975	28.7	32.9
1976	39.7	49.1
1977	39.7	52.6
1978	39.7	56.4
1979	39.7	62.9
1980	39.7	70.0
1981	61.0	117.6
1982	61.0	123.4
1983	61.0	129.0
1984	61.0	134.3
1985	61.0	139.4
1987	91.1	214.9
1988	91.1	223.4
1989	91.1	233.8
1990	91.1	247.1
1991	91.1	255.6
1992	91.2	262.2
March 1993	91.2	268.9

View Table

Table 2.

Fund Quotas - Actual and in “Hard” SDR

(In billions of SDRs)

Year	Actual quotas	Hard SDR-denominated quotas
1974	28.7	30.0
1975	28.7	32.9
1976	39.7	49.1
1977	39.7	52.6
1978	39.7	56.4
1979	39.7	62.9
1980	39.7	70.0
1981	61.0	117.6
1982	61.0	123.4
1983	61.0	129.0
1984	61.0	134.3
1985	61.0	139.4
1987	91.1	214.9
1988	91.1	223.4
1989	91.1	233.8
1990	91.1	247.1
1991	91.1	255.6
1992	91.2	262.2
March 1993	91.2	268.9

This would be potentially an important gain. The erosion of Fund quotas by inflation has been a recurring issue. For example, Polak (1984), in discussing the declining role of the Fund vis-á-vis the commercial banks in financing for developing countries, noted that “inflations, and, quota increases that did not take adequate account of inflation had, by the mid-1970s, reduced at least by half the quotas as a percentage of imports both for the membership as a whole and for its main customers, the non-oil LDCs. As a result, the amount of financial assistance that a country could expect from the Fund under its prevailing policies became too small, in relation to the size of its payments problem, to make recourse to the Fund an acceptable political choice.” This inadequacy of Fund quotas, led the Fund to take major remedial steps including an expansion of access limits, a shift of emphasis from one-year to three-year arrangements and lengthening of the maximum period of repayment under the Extended Fund Facility. The adoption of a “hard” SDR would make a major contribution to alleviating the recurring problem of inadequacy of quotas and therefore of the financial assistance the Fund can extend, in relation to the magnitude of its members’ payment imbalances.

Similarly, the stock of allocated SDRs, if it were to be denominated in “hard” SDRs, would rise with the increase in the world price level (as measured by the GDP deflator). The illustrative calculation in Chart 3 shows the current stock of SDRs if each of the six allocations so far had been denominated in “hard” SDRs. Thus, without any additional allocations, the outstanding stock of “hard” SDRs would have amounted to the equivalent of some 38 billion of current SDRs, compared with the actual size of 21 billion SDRs.

Another important consequence of a “hard” SDR for the Fund’s operations would be the impact on the real debt servicing by Fund members on outstanding Fund credit. Since Fund credit would be denominated in “hard” SDRs, and the repayment as well would be in “hard” SDRs, the impact of inflation on debt service would be eliminated. As is well known, inflation results in early repayment of the principal when debt service is measured in real terms. Shifting of Fund credit to a “hard” SDR basis would therefore eliminate this implicit early repayment of principal, thereby (implicitly) lengthening the maturity of Fund credit. A similar effect would apply, pari passu, to member’s credit positions vis-á-vis the Fund.

SDR Allocations in ‘HARD’ SDRs Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

SDR Allocations in ‘HARD’ SDRs Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

SDR Allocations in ‘HARD’ SDRs Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

The interest rate on the “hard” SDR would be a “real” interest rate, set by deflating the basket market interest rate by the rate of change in the SDR deflator. This would not present any difficulties, since the low interest rate on the “hard” SDR would be interpreted as a real or inflation-adjusted interest rate.

1. The asymmetrical basket

The implementation of an asymmetrical basket is feasible in a world of fixed but adjustable exchange rates, because in such a world there is a clear benchmark, i.e., the current parity or central rate, against which a devaluation or depreciation can be defined. The Spanish authorities’ proposal for a “hard basket ECU” is, in fact, an application of the asymmetrical basket idea to the basket ECU with devaluations defined in terms of downward realignments of EMS parities. The downward adjustments of the exchange rates of the basket currencies are offset by raising the number of units of the depreciating currency in the basket in proportion to the size of the depreciation.

In the current context of freely floating currencies of the major countries, the notion of stability of the SDR’s value is not easy to define if it is restricted to mean stability in terms of currencies. If the objective, following the notion of the asymmetrical basket, is to prevent depreciation of currencies in the basket from reducing the value of the SDR in terms of other currencies, it would be necessary to set benchmark levels of exchange rates of the basket currencies with reference to which a depreciation or an appreciation would be defined. Such benchmark levels would necessarily be arbitrary. In any case there would be little economic rationale to such arbitrary benchmarks and deviations of market exchange rates from them.

Despite these fundamental difficulties as to the meaning of a hard SDR defined in terms of an asymmetrical basket, it could, in principle, be constructed at a technical level if it were to be decided to pick a date--a necessarily arbitrary one--from which any appreciation or depreciation would be measured. However, the frequency with which the basket is changed would need to be decided, e.g., say monthly. A daily change would imply an unstable “hard” SDR, because in a floating exchange rate world, a currency which appreciates vis-á-vis the benchmark date may be depreciating the day after. Such instability in the composition of the SDR would be a major drawback of the asymmetrical basket. In any case, it would be difficult to attach any economic significance to a hard SDR based on such arbitrary benchmarks and deviations of market exchange rates from these levels.

2. SDR of constant purchasing power

The natural interpretation of a “hard” SDR in a world of generalized floating of major currencies therefore would center around a definition which would aim at maintenance of the SDR’s purchasing power over goods and services. At a technical level, there can be a number of ways that would maintain the SDR’s purchasing power. For instance, the “hard” SDR could be defined in terms of a basket of commodities, together with an agreement on the commodity price quotations to be used in calculating its value in terms of currencies. The major difficulty with such an approach, however, (as noted in Section II.2) is that the markets for individual commodities are in the short as well as the medium term, subject to several special influences, including speculation, which result in their prices deviating often sharply from the general level of prices of goods and services. The price of a collection of primary commodities would tend to be more stable than that of any single commodity, but still less stable than a general price index of a comprehensive nature, such as a consumer price index or GDP deflator.

The method of valuation which is the most suitable in the present context in order to establish a “hard” SDR is therefore one where the amounts of currencies in the SDR basket are adjusted periodically (say monthly) to reflect the loss of purchasing power of each currency, as measured by its rate of inflation. Here, the measure of inflations used for illustrative purposes is the consumer price index. The illustrative calculations in Table 3 show the amounts of currencies in the SDR basket each month, if the “hard” SDR had been established in July 1974 when the basket method of valuation was first adopted. The “hard” SDR’s currency basket is shown in Table 2 and Charts 9-13 and exchange rates of the “hard” SDR for the five basket currencies are shown in Charts 4-8. The detailed monthly calculations of the hard currency basket and exchange rates are given in the Appendix table.

It should be noted that under the method implemented for illustration, there is a discrete change at the end of each month in the number of currency units comprising the SDR basket. If such a discrete jump were to be avoided, a smoothing procedure involving a moving average of rates of monthly inflation might need to be used. In such a case, it would be possible to pre-announce the daily changes in the currency basket amounts at the start of each month, based on a moving average of rates of inflation in the preceding months.

U.S. Dollar Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

U.S. Dollar Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

U.S. Dollar Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

JAPANESE YEN Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

JAPANESE YEN Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

JAPANESE YEN Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

Deutsche Mark Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Deutsche Mark Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

Deutsche Mark Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

Pound Sterling Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Pound Sterling Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

Pound Sterling Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

French Francs Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

French Francs Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

French Francs Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

U.S. Dollar Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

U.S. Dollar Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

U.S. Dollar Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

Japanese Yen Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Japanese Yen Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

Japanese Yen Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

Deutsche Mark Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Deutsche Mark Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

Deutsche Mark Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

Pound Sterling Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Pound Sterling Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

Pound Sterling Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

French Francs Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

French Francs Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

French Francs Jul 1974 - Jul 1993

Citation: IMF Working Papers 1994, 025; 10.5089/9781451920796.001.A001

• Download Figure
• Download figure as PowerPoint slide

Table 3.

Amounts of Currency in the “Hard” SDR

(In units of respective national currencies)

Source: Staff calculations. For more details, see Appendix tables.

Table 3.

Amounts of Currency in the “Hard” SDR

(In units of respective national currencies)

	U.S. dollar	Japanese yen	Deutsche mark	Pound sterling	French franc
December 1974	0.5400	34.0	0.460	0.071	0.74
end-1975	0.6062	38.75	0.494	0.094	0.85
end-1976	0.6352	42.84	0.512	0.1078	0.93
end-1977	0.6779	44.93	0.530	0.1208	1.02
end-1978	0.7386	46.64	0.543	0.1308	1.11
end-1979	0.8359	49.27	0.572	0.1529	1.24
end-1980	0.9383	52.6	0.604	0.1756	1.41
end-1981	1.0214	55.0	0.644	0.1964	1.60
end-1982	1.0606	56.1	0.673	0.2068	1.76
end-1983	1.1007	57.02	0.691	0.2178	1.92
end-1984	1.1441	58.52	0.705	0.2277	2.05
end-1985	1.1871	59.34	0.717	0.2403	2.14
end-1986	1.2004	59.15	0.709	0.2493	2.19
end-1987	1.2529	59.60	0.716	0.2585	2.26
end-1988	1.308	60.18	0.729	0.2760	2.33
end-1989	1.369	61.73	0.751	0.2970	2.41
end-1990	1.4516	64.07	0.772	0.3242	2.49
end-1991	1.4964	65.7	0.804	0.3387	2.56
end-1992	1.5394	66.5	0.834	0.3473	2.61
July 1993	1.5662	67.3	0.858	0.3510	2.65

Source: Staff calculations. For more details, see Appendix tables.

View Table

Table 3.

Amounts of Currency in the “Hard” SDR

(In units of respective national currencies)

	U.S. dollar	Japanese yen	Deutsche mark	Pound sterling	French franc
December 1974	0.5400	34.0	0.460	0.071	0.74
end-1975	0.6062	38.75	0.494	0.094	0.85
end-1976	0.6352	42.84	0.512	0.1078	0.93
end-1977	0.6779	44.93	0.530	0.1208	1.02
end-1978	0.7386	46.64	0.543	0.1308	1.11
end-1979	0.8359	49.27	0.572	0.1529	1.24
end-1980	0.9383	52.6	0.604	0.1756	1.41
end-1981	1.0214	55.0	0.644	0.1964	1.60
end-1982	1.0606	56.1	0.673	0.2068	1.76
end-1983	1.1007	57.02	0.691	0.2178	1.92
end-1984	1.1441	58.52	0.705	0.2277	2.05
end-1985	1.1871	59.34	0.717	0.2403	2.14
end-1986	1.2004	59.15	0.709	0.2493	2.19
end-1987	1.2529	59.60	0.716	0.2585	2.26
end-1988	1.308	60.18	0.729	0.2760	2.33
end-1989	1.369	61.73	0.751	0.2970	2.41
end-1990	1.4516	64.07	0.772	0.3242	2.49
end-1991	1.4964	65.7	0.804	0.3387	2.56
end-1992	1.5394	66.5	0.834	0.3473	2.61
July 1993	1.5662	67.3	0.858	0.3510	2.65

Source: Staff calculations. For more details, see Appendix tables.

1. Frequency of adjustment

The first issue would therefore be the frequency with which the amounts of currencies in the valuation basket would be adjusted in order to correct for erosion (or accretion) in its real value. The competing considerations here would be the extent to which the SDR’s real value is eroded over different duration periods and the desirability of avoiding large, abrupt changes in the SDR’s value when the adjustments take place. The practical constraint on the frequency with which the adjustments in the currency amounts can take place is the availability of data on price inflation for the countries whose currencies comprise the basket. A period of a month, therefore, would seem to be the minimum feasible interval for adjustments. If the rates of inflation of the currencies in the basket are very low, say below 4 percent, quarterly or semiannual intervals could be considered. An advantage of making the adjustment interval as short as possible would be that it would reduce the size of the change to be made each time.

Private market participants would try to anticipate the changes in order to build them into the yields on SDR-denominated financial instruments. To the extent that monthly changes in price indices are more difficult to forecast than quarterly or annual changes, such adjustments would be difficult to anticipate for private markets. If the rules for adjusting the basket are made clear and well-known, the market can forecast on the basis of the available information.

2. Choice of the price index

The objective of preserving the real value of the SDR in terms of a comprehensive basket of goods and services in the world economy would point to the GDP deflators of the countries whose currencies form the SDR basket as appropriate measures of inflation. However, the long lags with which these are published would argue against their selection. If it were to be decided to use GDP deflators, the highest frequency of adjustment would be restricted to a quarter rather than a month.

3. Consistency with the Fund’s Articles

An important technical and legal issue concerns the requirement under the present rules that a new basket have the same value as the old one on the last day the old one is still in operation. Under the adjustment made each month to establish the value of the “hard” SDR, there would obviously be a significant departure from the present method because the value of the baskets would actually change--in fact that is the point of the exercise--on the days when the new baskets come into effect. The question of whether the present rules could be interpreted in such a way that the “hard SDR basket” adjustment to be done each month (or other selected interval) would be compatible with them or an amendment of the rules would be necessary is an open question needing further investigation.

4. Implications for the interest rate on SDRs

The establishment of a “hard” SDR would imply a concurrent change in the method of calculating the interest rate on the SDR. This should not present any additional technical difficulties because with similar inflation adjustment of weights of the SDR interest-rate basket, the interest rate on the “hard” SDR can be calculated on a daily basis. This interest rate, being inflation-adjusted, would be considerably lower than the one on the nominal SDR which is a weighted average of market interest rates prevailing in the five countries whose currencies comprise the SDR basket.