II. Potential Contribution of the SDR to Economic Stability

Juanita Roushdy
Published Date:
March 1987
  • ShareShare
Show Summary Details

The need for greater stability and improved performance of the international monetary system has been recognized in the recent reports of the Deputies of the Group of Ten36 and the Deputies of the Group of Twenty-Four.37 The report of the Group of Ten noted that “while the process of liquidity creation has been made flexible and the scope for countries to manage their international reserve position has generally increased, the working of the system has not been entirely satisfactory.”38 The report of the Group of Twenty-Four was more forceful: “The record of creation and management of international liquidity in the past decade has been unsatisfactory. During this period, the supply of international liquidity was uneven and grossly inadequate, especially for developing countries.”39 While there is not yet agreement on any comprehensive international action that could improve the performance of the international monetary system, Fund members do aim, individually and collectively, at creating a more stable framework in which to pursue their objectives with respect to economic growth, external balance, and price stability. This quest for a more stable economic environment is appropriately pursued within the Fund, whose Articles of Agreement encompass the objective of promoting a stable international monetary system.40

The achievement of stability in the world economy depends on a number of factors, the most prominent of which is economic policies that combine to foster stable economic conditions, specifically sustainable economic growth, stable national price levels, and external equilibrium. Countries’ economic policies can be affected by various features of the international monetary system—the framework within which national policies are conducted. Two elements of the system are particularly important in this context, namely, the obligations of countries with respect to exchange rate policy and the process through which reserves are made available to countries. This paper deals chiefly with the second of these two elements, that is, with the international reserve system, its effects on countries’ policies, and the way in which SDRs, as part of the reserve system, could influence these policies.

Countries hold international reserves so that they can carry on with their chosen policies when encountering temporary disturbances that, in the absence of reserves, would force them to modify their policy stance. Insofar as countries’ policies promote stable economic conditions—say, sustainable noninflationary growth—an adequate stock of reserves could be judged to exert a stabilizing influence by allowing these policies to continue. Moreover, ample reserves can have a favorable influence on the choice of a country’s policies by providing a basis for the confidence needed to adopt a relatively free and open system of trade and payments, which can in turn have beneficial effects on economic growth and stability. If there could be assurance that all countries invariably adopted policies promoting economic stability provided their reserves were adequate for that purpose, the reserve system could be designed so as to supply reserves relatively freely to satisfy any existing demand.

Command over reserves, however, permits the continuation not only of stabilizing but also of destabilizing policies. Moreover, an assured supply of reserves, at low or no net cost, regardless of the character of the policies that are being pursued, may tempt countries to move away from stabilizing policies. Since any resources that may become available to a country to be held as reserves are equally usable for the settlement of external deficits, an open-ended supply of free reserves could invite policies that induce such deficits and thereby reap the advantages entailed in the free use of foreign resources.

To be compatible with economic stability, a reserve system must, therefore, fulfill a dual function: on the one hand, it must permit and encourage the continuous pursuit of stabilizing policies in the midst of domestic and external disturbances, but on the other hand, it must avoid any undue inducement of policies that might destabilize participating economies by fostering excessive external deficits. Various historical reserve systems, beginning with the gold standard, have met this dual requirement in different ways. Although these historical systems will not be reviewed in detail here, a discussion of their main features can provide a perspective for the present reserve system and for consideration of the role that the SDR can properly play in it.

The remainder of the paper is divided into three sections. The first explores the strengths and weaknesses of the present reserve system, with special emphasis on the contribution of the mechanisms through which reserves are generated to the stability of the world economy and on perceived deficiencies in the reserve system that are, or could be, detrimental to economic stability. The second considers the contribution that the SDR might be capable of making to world economic stability by improving the reserve system. The third provides a summary of the role that the SDR might be perceived to play in fostering economic stability.

The Reserve System and World Economic Stability

The availability of reserves and the cost of acquiring and holding them exert an influence on the economic policies of countries. A change in the reserve system can thus affect the performance and stability of the world economy. This section begins with a brief review of the relationship between the supply of reserves and the formation of economic policy. It continues with a description of the present multicurrency reserve system, the relationship between the supply of reserves and economic stability in that system, and alleged shortcomings in that relationship.

Reserve Supply and Choice of Policies

The primary purpose of holding reserves is to provide a temporary “shock absorber” to mitigate the effects of sudden changes in a country’s external payments position. To be sure, a country could, in principle, allow its external accounts to come into balance without ever using, or even holding, official reserves. To do so while avoiding undue exchange rate instability, a country would have to rely on two important elements: first, its policies would have to succeed in keeping the external accounts reasonably close to a path of equilibrium by responding quickly and flexibly to external or internal disturbances; and, second, private capital markets would have to be guided by stabilizing expectations with respect to the exchange rate and respond readily to any profit opportunities created, in the light of these expectations, by temporary payments imbalances. Some industrial countries, and a few developing countries as well, have at times allowed their exchange rates to be determined entirely by market forces, including responses to their own policies. No country, however, has seen fit to eliminate its official reserve holdings, and few countries have been willing to see these holdings permanently reduced.

The inclination of countries to keep their official reserves intact in an international monetary system that permits the floating of exchange rates is not difficult to understand. When faced with the need to react to disturbances of the payments position by either financing part or all of an external deficit or allowing instant adjustment to take place, perhaps through a depreciation of the exchange rate, the use of reserves for balance of payments financing is a welcome additional policy instrument. Since the authorities always have the option of not using reserves held by them, that is to say, of not employing this additional policy instrument, the expected outcome of policymaking could hardly be worse, and may be considerably better, when reserves are available than when they are not. An appropriately timed temporary use of reserves is likely to foster economic stability, especially in two circumstances: first, use of reserves to finance a temporary payments deficit may avoid costly adjustments that would have to be reversed when the effects of the underlying temporary disturbances have passed. Second, when sudden disturbances are believed to be permanent, a temporary use of reserves permits a more gradual, and thereby more balanced and less disruptive, adjustment.

The extent to which countries will wish to avail themselves of the option to use reserves for financing payments deficits, and indeed the amount of reserves they will decide to hold, depends on the costs and benefits of holding reserves. If countries attempt to add to their reserves until the cost of an additional amount matches its benefits, they will be sensitive in their reserve policy, and therefore in their policymaking generally, to two developments: first, to changes in the conditions on which international financial resources can be obtained (e.g., interest rates, interest spreads above basic lending rates, and the cost of the use of other policy instruments to achieve external balance); and, second, to changes in factors affecting the potential benefits of reserve holdings (e.g., the volatility of a country’s payments position, the size of capital movements, obligations with respect to exchange market intervention, and the availability and cost of access to alternative financial resources).

It is through policy responses to changes in these costs and benefits that a reserve system can exert an influence on global economic stability. The mechanism of the international gold standard provides an illustrative example as well as a comparator for the present reserve system. Under the gold standard, a country that expanded its economy faster than its trading partners might initially be able to finance resulting current account deficits through an inflow of capital but would in the end come to lose some of its gold reserves. At that point it would have to abandon its expansionary policies; in fact, a more contractionary policy stance would be the automatic consequence of the reduction in its monetary base caused by the loss of gold. In this way, the policies of individual countries were subject to the stabilizing discipline exercised by the system.

The discipline of the gold standard thus provided a kind of automatic surveillance over the balance of payments adjustment of countries that opted to follow the unwritten rules of that system. This automatic surveillance was, in principle, symmetrical in that payments surpluses generated expansionary pressure just as deficits generated contractionary impulses. Although countries could opt out of the system—and some did, for shorter or longer intervals—the stabilizing effect of the gold standard system on the world economy was assured as long as a substantial number of countries, including major trading countries, continued to adhere to it.

While this stabilizing tendency of the gold standard system took effect directly when at least some of the leading countries maintained stable economic conditions in their own economies, there was also a more roundabout mechanism that imposed discipline on all countries together by reining in any simultaneous expansion that would leave individual payments balances unaffected. If all countries attempted, through expansionary monetary and fiscal policies, to achieve faster growth than the resources of the world economy could accommodate, so that there was a tendency toward world inflation, the cost of gold mining would also rise relative to the price of gold, which was typically fixed in terms of major currencies for extended periods; gold production would consequently decline at the very time when the demand for reserves increased in line with a higher nominal value of trade and other international transactions. In trying to attract a larger share of gold reserves, some countries would switch to less expansionary policies and, by succeeding in their objective, would put additional pressure on others to do likewise.

This mechanism, which worked in a similar manner to counter world stagnation and deflation, entailed an automatic adjustment of the supply of reserves to global reserve demands generated at stable conditions of the world economy. Any excess or shortfall of the world’s gold stock would tend to be corrected by the effects of inflation or deflation on the current output of gold.41

In summary, reserves can have a stabilizing effect by insulating countries’ policies from temporary disturbances and by permitting orderly adjustment, which may be time-consuming, to abrupt permanent changes affecting the balance of payments. Reserves held by a country also permit, however, the continuation of destabilizing policies for as long as the stock of reserves lasts. Appropriate arrangements with respect to the supply of reserves must be relied upon for limiting the effects of destabilizing policies on the countries pursuing these policies and on the world economy as a whole. Although the international gold standard is today only of historical interest, it provides an illustration of a mechanism that limits the adverse effects of destabilizing policies in individual countries and in the world economy as a whole.

Reserves and Stability in the Multicurrency Reserve System

There is little to add under this heading to the general description above of the potential stabilizing role of reserve holdings in individual countries. One feature of the present international monetary system to be noted in this context is not so much a characteristic of the multicurrency reserve system as an aspect of the entire economic and financial structure of the 1980s. This feature is the dominant size of financial magnitudes—stocks and flows—over related real magnitudes. Exchange market transactions in a day or a week amount to values comparable to those of trade flows over a quarter or a year. As a result, the task of stabilizing the external sector of an economy through the use of reserves is more difficult than it was earlier, when financial flows were smaller. For this reason, reserves must now be larger and international liquidity more ample relative to trade than, say, two decades ago, to accomplish the same tasks.

In the present international monetary system, many countries can increase their foreign exchange reserves quite freely at relatively low net cost, either by purchasing reserve currencies with national currency in the exchange market or by borrowing balances of reserve currencies in private capital markets. Other countries, however, find their ability to acquire reserves constrained by the absence of an international market for their national currencies and by limits on their access to credit markets. Apart from any SDR allocations or conditional credits that they may receive, these countries have to provide themselves with reserves through surpluses in those accounts of the balance of payments that can be influenced by their policies, in particular the current account.

For countries not subject to these constraints, the present reserve system works, up to a point, quite satisfactorily. Countries that take timely action to adjust their economies to changes in the external environment, as reflected in their balance of payments positions, can ordinarily be confident in their ability to acquire and hold the reserves they desire. Moreover, the supply of reserves—or more accurately, the supply of liquid funds that can be accumulated as reserves or used to finance payments deficits—responds flexibly to changes in the demand for such funds by creditworthy countries. There are some respects, however, in which the present reserve system is less adequate, particularly, in the determination of the global supply of reserves and in the timing and the manner in which adjustment pressure is brought to bear on countries in deficit (or surplus).

The global supply of reserves depends mainly on decentralized decision making of numerous private financial institutions in several countries, generally acting in the interest of their shareholders. These decisions often constitute responses to developments that potentially affect the profitability of these institutions. The time lags in these responses, together with a tendency toward exaggerated reactions to stimuli in a group of competing agents, provide a decision pattern characterized by “overshooting,” with periods of excessive expansion in the supply of reserves being followed by periods of inadequate expansion or even contraction.

The problem of instability in the mechanism generating reserves was noted in the report of the Deputies of the Group of Ten. The Deputies agreed on the existence of the problem and emphasized that, although the supply of international liquidity had been ample, if not excessive, during most of the 1970s and 1980s, instances of sharp contraction in the availability of international credit also occurred. In the view of some of the Deputies “… present arrangements for the provision of liquidity have not been optimal and …, while ultimately producing very powerful effects, they have not always been conducive to a gradual adjustment toward steady noninflationary growth of the world economy.”42

The vulnerability of an international monetary system relying on private capital markets for the generation of the bulk of official reserves lies not primarily in reserves being held but rather in reserves being spent. The danger is that some countries can finance growing external deficits through unchecked borrowing from these markets. This financing of payments deficits raises the reserves of the surplus countries, which may at times react by pursuing more expansionary policies themselves. As a result, all countries may come to expand faster than permitted by the growth of labor and capital resources, without any of them incurring unsustainable external deficits.

In contrast to the gold standard, the multicurrency reserve system provides no global automatic constraint on the policies of all countries taken together in the situation just described. If all countries pursued unduly expansionary policies at the same time, resulting in world inflation but little change in countries’ external payments positions, the additional foreign exchange balances that countries would wish to acquire so as to maintain the real value of their reserves intact would be readily available to them. As a result, the present reserve system, in contrast to the gold standard system, cannot ensure world economic stability in the circumstances just described. This shortcoming was acknowledged by the Deputies of the Group of Ten in their report: “… the stability of the system depends primarily on all countries pursuing sound policies to achieve sustainable noninflationary growth.”43

A second aspect of the present reserve system that should be noted in the context of economic stability has to do with the initially delayed, but ultimately quite abrupt, application of pressure on a country to undertake balance of payments adjustment. Comparison with the gold standard system can again illustrate the point. Under that system, a country with an external deficit will lose reserves steadily until the deficit is eliminated or reversed. As reserves decline, the need to adjust becomes more obvious and more pressing all the time. In the present system, by contrast, creditworthy deficit countries can prevent the decline in reserves through external borrowing, albeit at the cost of interest payments accruing on the new loans. The mounting indebtedness and the higher interest charges may induce adjustment action, but they may fail to do so, especially if the funds needed for making interest payments can themselves be borrowed. In that case, the deficit imposes no immediate real burden on the country in question and may be allowed to continue until the indebtedness reaches a level at which credit-worthiness is suddenly lost and harsh adjustment measures must be taken.

From the viewpoint of the contribution of the reserve system to economic stability, the delay in the effective signaling of the need for adjustment measures is a systemic weakness. There are also some related weaknesses, which will be discussed below, that have to do with the difficulty of financial institutions in assessing country risks, which can lead to an unjustified loss of creditworthiness or prevent the timely restoration of creditworthiness once the conditions causing its loss have been reversed. But the phenomenon of loss of creditworthiness itself is not a weakness or flaw of the present reserve system. It constitutes a necessary signal in a system that otherwise lacks a clear indicator of the situation in question: it corresponds roughly to the point under the gold standard system at which a country is about to use the last of its gold holdings.

A third weakness of the current reserve system, with its reliance on private financial markets as the principal source of reserves, has to do with the effect of misinformation on the supply of reserves to individual member countries. Changes in the availability of reserves have often reflected shifts in market perceptions of creditworthiness that have been based in some periods on inappropriate considerations or incorrect perceptions. Critics of the current system have noted that the late 1970s and early 1980s witnessed a tendency toward overlending. At the other extreme, the availability of credit to a country can be curtailed for reasons that do not reflect exclusively the conduct of its policy. For example, it has been suggested that private markets often evaluate the creditworthiness of an individual country in terms of its regional location or stage of development, rather than on a case-by-case basis. Thus, when one country in a particular group experiences difficulties in servicing its external debt, private institutions will suspend new credits to all countries that might have similar problems. While this could be a rational response of individual financial institutions to uncertainty and a lack of complete information, the resulting liquidity shortage could harm even prudent debtor countries and thereby destabilize the international monetary system. Debtor countries can also be harmed by changes in the availability of private credit that arise from changed conditions in financial markets of the industrial countries rather than from changes in the creditworthiness of the debtors.

An additional criticism has focused on the long lags and other difficulties of restoring creditworthiness after countries adjust their policies. Despite the vigorous adjustment policies pursued by many developing countries, which have led to a sizable improvement in their current account positions since 1982, their creditworthiness has not always been restored. In fact, the expansion of bank lending to developing countries has continued to slow and has elicited proposals designed, in part, to reverse this decline.44

Critics of the current system also point out that private lending decisions, which determine the availability of borrowed reserves, are heavily influenced by nonmarket forces. It has been argued that decisions by commercial banks regarding the creditworthiness of a particular country are influenced not only by an evaluation of that country’s policies, but also by the attitudes of official agencies regulating banks and extending export credits. The imposition of special loan-loss provisions or the suspension of export credit insurance to any country or group of countries could more than offset the effects of the country’s adjustment efforts on the market’s perception of its creditworthiness. Thus, countries may not face an efficient and competitive financial market in attempting to satisfy their demand for reserves.

Perhaps the most serious deficiency of a system in which countries hold borrowed reserves is that their availability has been unreliable during periods of greatest need for reserves. At such times, as already noted, countries without access to international markets must often seek to acquire reserves through current account surpluses. The adjustments in policies and income needed to generate current account surpluses in excess of those required to service external debt positions can impose a high real cost on the acquisition of reserves. Uncertainties about the terms and conditions under which reserves would be available have led countries to try to increase their owned reserves even though this has often required a sharp contraction in imports.

It has been suggested that this additional adjustment undertaken to accumulate reserves may impose an unnecessary burden on the countries in question and on the world economy as a whole. Moreover, it has been noted that in attempting to improve their external payments position to acquire reserves, countries without market access have been confronted with movements in exchange rates, inflation, growth, and trade that have been affected as much by the domestic policies pursued in the major industrial countries as by those in their own economies. In addition, it has been argued that the policies of the major countries have not been internationally consistent and have led to levels of real interest rates and fluctuations in exchange rates that have affected the cost and availability of reserves without adequately promoting adjustment.

A further problem of the current reserve system arises because reserves are held in the form of financial assets denominated in a number of reserve currencies. This allows the value of reserves to fluctuate in the face of movements of exchange rates of these currencies. Decisions by countries to alter the currency composition of their reserves might in some circumstances destabilize exchange markets. This would most likely be the case where official actions are taken as evidence by private portfolio managers that official institutions have lost confidence in a currency.

Role of the SDR in the Reserve System

The preceding section has analyzed the potential effects on world economic stability of a number of shortcomings of the international reserve system. It is conceivable that economic stability could be enhanced by countering the effects of some of these shortcomings through arrangements involving the SDR. It is also likely, however, that some deficiencies of the reserve system cannot be remedied by the SDR. This section first seeks to identify those inadequacies that appear amenable to being remedied or mitigated through use of the SDR. It then discusses SDR arrangements and procedures, including allocation and cancellation of SDRs, that could provide the intended remedy or mitigation.

Possible Improvements Involving the SDR

In the mechanism through which the current reserve system provides liquidity to its members, imperfections with potential adverse effects on stability are found both in the determination of the scale on which liquidity is made available to the system as a whole and in the distribution of liquidity among countries. This is an important distinction in the present context for two reasons. First, it may be more difficult to reach consensus about distributional shortcomings, which are often seen as favoring one country group while harming another, than about system-wide (or scale) deficiencies, which may at times harm all countries. Second, the SDR mechanism set out in the Articles of Agreement, which relies on uniform allocation of SDRs in proportion to members’ quotas,45 is better fitted to counteract scale shortcomings than distributional ones.

In addition to these scale and distributional deficiencies in the provision of liquidity, the system also contains another source of potential instability, which has to do with the variability of the currency composition of existing reserve holdings. The contribution that the SDR could make to the alleviation of this instability problem was extensively discussed at the time active consideration was given to a substitution account in which reserves denominated in U.S. dollars, and perhaps other currencies as well, could be deposited in an account administered by the Fund in return for claims denominated in SDRs.

The possible contribution of the SDR to the alleviation of these three stability problems—for brevity referred to as scale, compositional, and distributional problems—will be briefly discussed in turn.

As was argued earlier, the major shortcoming of the current reserve system in determining the overall scale on which liquid funds are supplied is its inability to adjust that supply so as to exert a stabilizing influence on the world economy. Because it lacks an automatic “governor,” the supply of liquidity expands freely at times on terms and conditions that do not prevent, and indeed contribute to, excess demand and inflation in the world economy; and at other times international liquidity may become unduly tight, with many countries finding their access to international credit severely curtailed and therefore forced to reduce the growth of their economies and to restrict external trade. Since the SDR was created to supplement existing reserves when needed, it seems worth considering whether it could play a stabilizing role through the mechanism of allocation and cancellation provided in the Articles of Agreement.

The inability of the reserve-generating process, which is based on decentralized private decision making, to provide stable reserve growth at an appropriate rate also suggests the question of whether the SDR could help to bring that process under better international control. For this purpose, the SDR would have to be able to play the role of an international base money that is itself under the control of the international community and through which other international monetary aggregates can be influenced.

Since the process of supplying reserves, as well as international credit in general, has shown evidence of overshooting, it is conceivable that the reserve system may, in particular circumstances, fail to provide even the minimum amount of liquidity necessary for conducting international trade and financial transactions at an acceptable level. Considerations of stability could then require quick action to make up the deficiency. SDR allocation—a process already in place—could provide the necessary supplementation of private credit and thereby provide a safety net for the world economy. In the same vein, a sudden tendency toward overexpansion in private lending could be remedied by the cancellation of SDRs, provided the stock of SDRs was large enough to permit the intended reduction in reserves. Of course, the notion of a safety net, implying some sort of worldwide breakdown of existing processes, is a subject discussed with considerable reluctance. All the same, it would be wise to explore it as part of normal contingency planning.

As regards the compositional problem inherent in a multicurrency reserve system with floating exchange rates, the SDR as a basket of major currencies can provide some protection for reserve holders against fluctuations in the value of their holdings as a result of exchange rate changes. Since reserve holders can, without undue difficulty, diversify their reserve portfolios so as to make up their own currency baskets, the importance for them of a voluntary SDR substitution arrangement would be marginal and extend primarily to the saving of certain transaction costs. To the world economy as a whole, as well as to some of the reserve-currency countries, the potential benefit of a substitution account lies in permitting currency substitution, when it is sought by holders, without affecting market exchange rates.

In turning to distributional shortcomings of the current reserve system, there are two different types that must be distinguished at the outset. The first arises to the extent that some countries, in particular reserve-currency countries, have easier or more extensive access to, or face a lower net cost of, international credit than do other countries. Possibilities for remedying this shortcoming with the help of the SDR are limited. Evidently, the solution cannot be to provide for all countries the advantages of easy access to liquidity that may be enjoyed by a reserve country46 even if it were possible to do so, since this would jeopardize global stability by weakening discipline throughout the international economy. The only possible remedy—not a very practical one in present circumstances—would be the replacement of the reserve-currency system by an SDR standard with “asset settlement” instead of “liability settlement” for all countries, as elaborated in the reform discussions of the Committee of Twenty in the early 1970s.47

The second type of distributional shortcoming is found when some countries have less assured, or costlier, access to international private credit than do most other countries in similar circumstances.48 Here, too, the SDR in its present structure does not appear to be an adequate instrument for remedying the situation because of the need to aim the remedy at specific countries. It would be unusual to find all countries of a major group—for instance, all developing countries—in a similar situation with respect to the adequacy of access to credit markets. It would be more common for some countries to enjoy good access at reasonable cost while others, although deserving of the same status, have had their access abridged and yet others have rightly been judged to lack creditworthiness. It would seem difficult to employ the SDR mechanism for improving the position of the middle group without affecting the two other groups. Neither the allocation of SDRs nor a general link between SDR creation and development finance would seem to be particularly well adapted for directing the remedy to the intended recipients. Extension of credit through the Fund’s General Department, which is of course also denominated in SDRs, would appear to be a more flexible instrument to use for the purpose under discussion. If a country’s policies were, in fact, adequate and the absence of creditworthiness resulted from a collective error of private financial institutions, Fund assistance would not need to be subject to burdensome conditionality and might be less costly than credit from the capital markets.

Use of the SDR to Improve Economic Stability

The general suitability of the SDR for alleviating instability in the current reserve system was reviewed earlier in this section. It remains to examine specific arrangements that could conceivably be helpful in the quest for economic stability. It is convenient to organize this discussion, not by reference to the problems identified in the preceding section, but in accordance with techniques through which the SDR can be employed for the purposes envisaged. Most attention will be devoted to techniques for which explicit provision is made in the Articles of Agreement, namely, allocation and cancellation of SDRs in specified circumstances. Some possible uses of the SDR that cannot be effected through allocation or cancellation in proportion to members’ Fund quotas will be touched on briefly, making reference to possible arrangements that are not specifically provided in the Articles.

SDR Allocation and Cancellation

This mechanism could in principle be used for several purposes related to the stability of the world economy. SDR allocation could supplement the creation of other reserve assets when their supply is judged to be insufficient. Similarly, cancellation of SDRs could reduce the rate of growth in the overall supply of reserves if it were judged to be excessive.49 The allocation/cancellation mechanism could also provide safety-net facilities. Moreover, as will be argued below, steady SDR allocation over a longer period could in effect serve a substitution function by gradually replacing borrowed currency reserves with owned SDRs and thus improve the overall quality of reserves. Finally, a substantial replacement of currency reserves through SDRs, which could make the SDR the main reserve asset in the international monetary system, could eventually create the basis necessary for attempting better international control of the process of reserve generation.

Use of the SDR through the allocation/cancellation technique for any of these purposes is subject to a number of requirements under the Articles as well as certain other considerations. According to Article XVIII, Section 1(a), decisions on allocation and cancellation of SDRs must be guided by the “long-term global need … to supplement existing reserve assets. …” There has been considerable discussion of this provision in the Fund over recent years. Generally, the requirement of a “global” need is taken to preclude allocation of SDRs to all Fund members for the purpose of meeting the reserve needs of a few of them. It is debatable how large a number of members must have a need for additional reserves to justify the judgment that this reserve need should be considered to be “global” under the provisions of Article XVIII. The presumption is that the need is global if the adverse consequences of not satisfying it would be global in the sense of affecting most or all countries or impairing the functioning of the international financial system.

The specification that the need for reserve supplementation must be of a “long-term” character is more difficult to relate to the various possible purposes of SDR allocation and cancellation mentioned above. This requirement is clearly intended to preclude SDR allocations that would satisfy purely short-term needs—say, needs related to short-term business fluctuations. The Articles do not provide explicit guidance with respect to the length of the time span over which a need for reserve supplementation must exist for it to be considered a “long-term need”; however, two benchmarks may be used: the first describes a need extending over such a long span as to be perennial in character; the second envisages a need existing over a substantial number of years that, if left unsatisfied, might have consequences that are irreversible or at any rate incapable of reversal in the short or medium term.

The first of these benchmarks has to do with the quality, rather than the quantity, of reserves. A variety of conditions, identified in Section I, make the international monetary system vulnerable to adverse effects of sudden changes in international credit availability. This vulnerability could be reduced through greater reliance on owned, rather than borrowed, reserves and through a moderation of the exposure of the value of countries’ reserves to changes in the exchange rates of reserve currencies and, more generally, to the economic policies of these countries. With a given long-term demand for reserves, a shift from borrowed to owned reserves could be achieved by steady periodic allocations at a moderate rate.50 The expectation of a steady rate of SDR allocation—generally a less costly source of reserves than funds borrowed in the capital markets—could lead to compensating adjustments in the long-run rate of growth of foreign exchange reserves so that total long-run reserve creation may be unaffected by the allocation of SDRs. The effect of allocation would thus be an improvement in the quality of reserves and thereby in the stability of the world economy. The provisions of the Articles clearly support SDR allocation satisfying a long-term global need of this character.

The second benchmark is suggested by the facts that the Articles provide not only for allocation but also for cancellation of SDRs and that basic periods of allocation or cancellation of a normal duration of five years are specified in the Articles. These features enable the international community to vary the rate at which SDRs are added to or removed from members’ reserve holdings in accordance with periodic decisions, provided that the pattern of these allocations and cancellations satisfies a long-term global need for reserve supplementation. Such changes in the rates of allocation or cancellation are not, or not fully, anticipated by national authorities and markets, and their effects are therefore quite different from those of continuing allocation at a steady rate. This is so because unexpected allocations or cancellations would alter economic and financial plans of governments and individuals made before these changes occurred. Such alterations in the rate of allocation could be used to offset unwanted developments in the volume of international reserves. For instance, changes in the secular rate of SDR allocation may be warranted to offset medium-term or longer-term variations in the supply of international liquidity generated in private financial markets. Consideration has also been given to the possibility of using SDR allocation as a “safety net” to cope with an international financial emergency of limited, though uncertain, duration. The Deputies of the Group of Ten in their report expressed the view that one of the contingencies under which SDRs could be made available as a safety net would be “the possibility of private markets being unable to respond adequately to a legitimate long-term global need for international liquidity.”51

In circumstances such as those just described, changes in SDR allocation might be made for a temporary period and reversed when the circumstances that induced them have ceased. These considerations apply both to action intended to offset the effects of inadequate or excessive growth of private cross-border credit and to the provision of a substantial injection or withdrawal of liquidity in the course of the activation of a safety net. Indeed, when SDRs are used as a safety net, it might be desirable to decide on allocations and (complete or partial) cancellations at the same time as part of the allocation/cancellation program for a particular basic period. If this were done, the long-term character of the need for reserve supplementation would be derived from the adverse long-term consequences of the conditions to be remedied by SDR allocation/cancellation rather than from the length of time for which allocated SDRs would be intended to be outstanding.


Ever since major exchange rates started to float, the idea has been discussed that the SDR could make a greater contribution to the stability of the reserve system if it were used to help insulate the values of reserve portfolios from the effects of exchange rate variability and to facilitate adjustments in the currency composition of official portfolios without generating exchange market effects. This could be accomplished through the substitution of SDRs or SDR-denominated instruments for foreign exchange reserves. Proposals to this effect were discussed most intensively in the periods 1973–74 and 1979–80, although the proposals for substitution in the period 1973–74 encompassed a broader set of changes in the structure of the international monetary system than the proposals in the period 1979–80.

During the earlier period, substitution was viewed as a means of reducing the strains on the international monetary system that might result if one or more of the reserve-currency countries were to be faced with demands for the conversion of existing balances into other currencies. Unless countries were offered the possibility of shifting a part of their reserves into SDR-denominated assets without going through the exchange markets, it was feared that official portfolio adjustments could result in extensive exchange rate changes. The perceived adverse effects of substitution in discouraging adjustment in reserve-currency countries were to be reduced by requiring that the currency balances surrendered to the substitution account be eliminated gradually by asset settlement on the part of the reserve-currency countries. In the discussions of 1979–80, in contrast, the proposed voluntary substitution account provided only for adjustments in the composition of reserve portfolios without the requirement of asset settlement. While voluntary substitution was viewed as having the potential to dampen exchange market pressures and to enhance the role of the SDR,52 it was recognized that the contribution to stability would be limited unless accompanied by appropriate macroeconomic policies in the reserve-currency countries.53

Those opposed to the substitution proposals argued that the establishment of such accounts could increase rather than reduce in stability. One concern was that expectations regarding the strength of individual reserve currencies could be lowered and thereby induce private speculative movements out of assets denominated in these currencies.

The substitution account proposals were not adopted, principally because agreement could not be reached on how to make provision for any discrepancy in the values of the account’s SDR liabilities and its reserve-currency assets that could arise because of differences in interest rates not compensated for by exchange rate movements. This difficulty may suggest that there is merit in the type of gradual substitution through steady SDR allocation discussed earlier.

Conditional SDRs

Distributional shortcomings of the international reserve system cannot easily be remedied through measures such as SDR allocation, which have a fixed distributional incidence. To improve the distribution of reserves, resources must be provided selectively to specific countries. It has been argued that the SDR could nevertheless contribute to a better distribution of reserves if allocations were made available for the extension of Fund credit to members. Under this proposal, participants in the SDR Department would place all or part of their allocations at the disposal of the Fund. These SDRs would then be provided to members that seek Fund assistance for implementing adjustment programs and rebuilding their reserves.

The proponents consider that such a program could strengthen the role of the SDR and meet the criticisms that SDR allocations might weaken adjustment efforts and grant unconditional resources to countries that do not need them. The proposal has also met with strong opposition, however. The Deputies of the Group of Twenty-Four noted that “only an unconditional SDR allocation could provide the required reserve strength. …”54 Moreover, it was suggested that these arrangements could blur the distinction between the SDR and conditional Fund credit and that in any case. Fund resources for conditional lending were sufficient for the foreseeable future.

SDR Link

Another proposal with distributional implications involving the SDR system has been discussed since the 1960s. Its aim is to link SDR creation to the provision of development finance by changing, directly or indirectly, the distribution of allocated SDRs in favor of developing countries. The proponents of such a link have argued that it would contribute to the liquidity and stability of the international monetary system by strengthening the positions of developing countries. The Deputies of the Group of Twenty-Four noted in their report that “… a link between allocation of SDRs and development would not only meet the unfulfilled absorptive capacity of developing countries but also reduce the pressures on the industrial countries to accommodate an improvement in the current account balances of developing countries.”55

Those that have not favored linking SDR allocations with development assistance have generally argued that such a link has the potential to destabilize the international monetary system by leading to excess creation of international liquidity. Because of the large and urgent need for development assistance, liquidity creation should, in the view of these observers, be kept apart from the provision of international liquidity lest a link induce a larger allocation of SDRs than would be consistent with the growth of the long-term global need for reserves. Moreover, since a link would in practice have to be arranged in accordance with a general formula for the distribution of SDRs, it would be difficult to direct SDRs to countries in accordance with their differentiated needs. Balance of payments support can be more precisely aimed at intended recipients through the use of “unconditional SDRs” or indeed of the Fund’s general resources.

Control of International Liquidity

The contribution of the international reserve system to the stability of the world economy could be greatly enhanced if it were possible to exercise a measure of international control over the processes through which reserves are generated and distributed. In particular, it would be useful if the supplies and availabilities that determine the state of international liquidity could be augmented whenever liquidity was less than adequate and reduced whenever liquidity was excessive. As discussed earlier, direct control over the liquidity-creating process is not feasible under the institutional arrangements of the present multicurrency reserve system. While movements in the adequacy of international liquidity could, in principle and to a limited extent, be offset by a judicious policy of SDR allocation and cancellation, by far the larger part of international liquidity is controlled by market processes and is apt to overwhelm the relatively small portion that is under control by the international community. In these circumstances, international surveillance over the policies of all members presents the only practical mechanism for control over international liquidity.

These considerations led the Deputies of both the Group of Ten and the Group of Twenty-Four to stress the importance of Fund surveillance as a means of fostering “policies consistent with a more stable evolution of international liquidity.”56 It was noted in both reports that Article VIII, Section 7 of the Articles of Agreement calls on members to collaborate with the Fund and other members in pursuit of the objective of “better international surveillance of international liquidity.”57 In particular, it was acknowledged that surveillance of both exchange rate policies and international liquidity should be closely integrated.58


Various ways in which the SDR can and does contribute to the stability of the world economy have been reviewed. Some of the functions of the SDR system adding to economic stability have been exercised since the SDR was first created more than a decade and a half ago. Others could be exercised under present provisions of the Articles of Agreement. Further contributions of the SDR to stability are conceivable but may be more difficult to achieve under present provisions and interpretations. Some of these uses of the SDR have nevertheless been addressed in this general discussion of the topic.

Present Contribution of the SDR to Stability

Although Fund members are not required to hold reserves in the present system, which permits free floating of exchange rates, reserves extend a country’s policy options and can thereby add to the stability of its economy. Allocated SDRs, which form part of members’ reserve holdings, can therefore contribute to economic stability in the world.

Moreover, SDR holdings are considered by holders as “owned reserves,” at a time when many countries hold in their reserve portfolios funds derived from short-term borrowing in international capital markets, which are subject to periodic refinancing. The greater reliability of SDRs in providing assured command over resources can contribute to stability in the policies of some holders of SDRs and to the stability of the international reserve system as a whole.

As a basket of the major currencies entering into world trade and financial transactions and held as currency reserves by Fund members, the SDR provides a convenient unit of denomination that fluctuates less in value against each component currency than that currency does against some of the other component currencies. The SDR denomination therefore enables members, national and international institutions, and private parties to keep accounts and make contracts and other arrangements across borders in a more stable unit than would otherwise be available. One of the most important uses of the SDR as a stable international unit of account is the denomination of the Fund’s own accounts, in particular its claims on and liabilities to its members. As members hold claims on the Fund as part of their reserves, the SDR denomination of these claims adds a further element of stability to the reserve system.

SDR allocation at a steady rate consistent with world economic stability could over time build a stock of SDR holdings forming a substantial part of the total reserves of Fund members and nonmember countries that are authorized holders of SDRs. With this rate of allocation being adjusted, as it must be in conformity with the Articles of Agreement, to the collective willingness of Fund members to absorb SDRs in their reserve holdings under conditions of stable developments in the world economy, the share of SDRs in total reserves would gradually increase. This change in the composition of reserves would over time improve the quality of reserves and the stability of the reserve system. A substantial rise in the share of the SDR in members’ reserve holdings, which might eventually be achieved, could present opportunities for better international control of international liquidity and thus for making more active use of the reserve system for international stabilization policy conducted by the international community.

These functions describe the role that the SDR plays, or could play, in enhancing economic stability in accordance with the present Articles of Agreement and the practices that have evolved under their provisions. There are other ways, which are discussed below, in which the SDR could, in principle, add to world economic stability by countering weaknesses in the present reserve system. Consideration of such new or extended functions would have to include addressing the question of their feasibility under the present Articles and interpretations, as well as their effects on existing functions of the SDR.

Possible Extension in the Stability Role of the SDR

The principal extension of the contribution of the SDR to economic stability considered in this paper concerns the possible allocation or cancellation of SDRs in two circumstances: first, to counteract any tendency of international liquidity to be inadequate or excessive over a short-term or medium-term period, and second, to provide quick remedial action in case of a sudden development calling for the deployment of a safety net.

In both of these circumstances, the initial action may need to be reversed sooner or later, in particular, if the developments to be counteracted are pronounced deviations from a longer-term trend. Reversal must also be presumed to be the normal expectation in order to allow the initial action, say, a substantial allocation, to be taken without undue hesitation. A more active use of cancellation of SDRs could, of course, reduce the assurance with which the SDR has come to be regarded as part of countries’ owned reserves in an era during which cancellation has, in practice, never been invoked.

In addition to the use of the allocation/cancellation mechanism for adjusting the scale on which international liquidity is made available to all member countries, allocated SDRs can also be used to alter the distribution among countries of the supply of liquidity. This could be accomplished by amending the provisions on allocation and cancellation so as to allow deviations from the proportional pattern prescribed in the present Articles or by agreements under which some members donate allocated SDRs for the benefit of other members. Under an “SDR link,” where the benefits flow to a specified group of members regardless of the circumstances of individual recipient countries, it is more difficult to achieve a stabilizing effect than under a system of “conditional SDRs,” where the benefits can be directed at countries most in need of temporary balance of payments support or reserve supplementation.

Finally, related decisions on allocation and cancellation could hardly be taken independently—at different times—as long as they require an 85 percent majority of the total voting power. In these circumstances, a decision to allocate SDRs and subsequently cancel them, completely or in part, would have to be taken at the initiation of the sequence, even though the cancellation could later in effect be postponed, if circumstances suggested it, through a further decision to allocate and cancel.

At present, the small share of SDRs in total reserves precludes the use of the SDR for international control of international liquidity. If a larger SDR share in the future were to move such control within reach of the international community, thought would have to be given to increasing the flexibility with which the volume of outstanding SDRs can be influenced.


    Group of TenStatement by Ministers of the Group of Ten and Annex Prepared by Their Deputies (London: Her Majesty’s Stationery Office1964).

    • Search Google Scholar
    • Export Citation

    Group of TenReport to the Deputies of the Group of Ten R. Ossola Chairman (Rome: Study Group on the Creation of Reserve AssetsMay311965).

    • Search Google Scholar
    • Export Citation

    International Monetary FundAnnual Report of the Executive Directors for the Fiscal Year Ended April 30 1964 (Washington1964).

    International Monetary FundInternational Financial News Survey Supplement (Washington) Vol. 17(1965) pp. 20916.

    International Monetary FundSummary Proceedings of the Twenty-Second Annual Meeting of the Board of Governors (Washington1967).

    International Monetary Fund(1968a) Annual Report of the Executive Directors for the Fiscal Year Ended April 30 1968 (Washington1968).

    • Search Google Scholar
    • Export Citation

    International Monetary Fund(1968b) International Financial News Survey (Washington) Vol. 20(1968).

    International Monetary Fund(1968c) Summary Proceedings of the Twenty-Third Annual Meeting of the Board of Governors (Washington1968).

    • Search Google Scholar
    • Export Citation

    International Monetary FundSummary Proceedings of the Twenty-Fourth Annual Meeting of the Board of Governors (Washington1969).

    International Monetary FundAnnual Report of the Executive Directors for the Fiscal Year Ended April 30 1971 (Washington1971).

    International Monetary FundAnnual Report of the Executive Directors for the Fiscal Year Ended April 301979 (Washington1979).

    International Monetary Fund(1985a) Annual Report of the Executive Directors for the Fiscal Year Ended April 30 1985 (Washington1985).

    • Search Google Scholar
    • Export Citation

    International Monetary Fund(1985b) Summary Proceedings of the Thirtieth Annual Meeting of the Board of Governors (Washington1985).

    International Monetary Fund(1985c) IMF Survey Supplement on the Group of 10 Deputies’ Report (Washington) Vol. 14 (July1985) pp. 116.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund(1985d) IMF Survey Supplement on the Group of 24 Deputies’ Report (Washington) Vol. 14 (September1985) pp. 116.

    • Search Google Scholar
    • Export Citation

    FritzMachlupPlans for Return of the International Monetary System International Finance Section Department of Economics Princeton University (Princeton: Princeton University Press, 2d ed.1964).

    • Search Google Scholar
    • Export Citation

    FritzMachlupand Malkiel B. eds. International Monetary Arrangements: The Problem of Choice. Report on the Deliberations of an International Study Group of 32 Economists (the so-called Bellagio Group) International Finance Section Department of Economics Princeton University (Princeton: Princeton University Press1964).

    • Search Google Scholar
    • Export Citation

    TriffinRobertTomorrow’s Convertibility: Aims and Means of International Monetary Policy” Quarterly Review Banca Nazionale del Lavoro No. 49 (June1959) pp. 131200.

    • Search Google Scholar
    • Export Citation

    TriffinRobertGold and the Dollar Crisis: The Future of Convertibility (New Haven: Yale University Press1961).


Communique of the Interim Committee of the Board of Governors of the International Monetary Fund, in International Monetary Fund, Summary Proceedings of the Fortieth Annual Meeting of the Board of Governors (Washington, 1985), para. 8.


Ibid., para. 10.


More precisely, it was limited by a deficit in the U.S. official settlements account, which is the counterpart to a net increase in U.S. liabilities to foreign official holders plus any net reduction of reserve assets held by the United States.


The U.S. current account surplus was also shrinking over most of this period.


Total reserves of industrial countries amounted to SDR 60 billion at the end of 1968.


This general argument was first formulated in the early 1960s. See Triffin (1961).


Speech by the Managing Director at the Institut d’Etudes Bancaires et Financieres, Paris, June 11, 1965, reproduced in International Monetary Fund (1965), p. 215.


Statement by the Governor of the Fund for the United States, Henry H. Fowler, International Monetary Fund (1968c), p. 53.


For a discussion of these proposals, see Machlup (1964).


See Group of Ten (1964), para. 2.


The Fund study was published as Chapters 3 and 4 in International Monetary Fund (1964).


International Monetary Fund (1968a), p. 16.


Article XVIII, Section 1(a).


See International Monetary Fund (1969), pp. 278–79.


International Monetary Fund (1971), pp. 18–19.


Article XXI, Section 2 of the Articles of Agreement as amended on May 31, 1968.


Communiqué, issued by the governors of the central banks of Belgium, the Federal Republic of Germany, Italy, the Netherlands, Switzerland, the United Kingdom, and the United States at the conclusion of their meeting in Washington, D.C. on March 16 and 17, 1968, reproduced in International Monetary Fund (1968b), p. 89.


Statement by the Governor of the Fund for the United States, Henry H. Fowler, International Monetary Fund (1967), p. 82.


See, for example, Group of Ten (1965), paras. 137 and 138.


The popular designation of the SDR in its formative period as “paper gold” also lost currency after the suspension of gold convertibility of the U.S. dollar.


International Monetary Fund (1985c), p. 8, para. 63.


International Monetary Fund (1985d), p. 2, para. 17.


International Monetary Fund (1985a), p. 62.


International Monetary Fund (1985d), para. 93.


Ibid., para. 94.


International Monetary Fund (1985c), para. 74.


International Monetary Fund (1985d), para. 23.


Allocation of Special Drawing Rights for the Third Basic Period: Proposal by the Managing Director of the International Monetary Fund, International Monetary Fund (1979), pp. 124–25.


International Monetary Fund (1985c), para. 78.


Articles of Agreement of the International Monetary Fund, Article VIII, Section 7.


International Monetary Fund (1985c), para. 61.




International Monetary Fund (1985d), para. 20.


“The Functioning of the International Monetary System,” reproduced in IMF Survey, Supplement on the Group of 10 Deputies’ Report (July 1985). (Hereinafter cited as Report of the Deputies of the Group of Ten.)


“The Functioning and Improvement of the International Monetary System,” reproduced in IMF Survey, Supplement on the Group of 24 Deputies’ Report (September 1985). (Hereinafter cited as the Report of the Deputies of the Group of Twenty-Four.)


Report of the Deputies of the Group of Ten, p. 8, para. 63.


Report of the Deputies of the Group of Twenty-Four, p. 3, para. 17.


See Article I and Article IV, Section 1.


In practice, the stabilizing effect of this automatic adjustment of the global supply of reserves was marred by the length of time it required, as well as by the uneven pace of gold discoveries and technical improvements in gold mining.


Report of the Deputies of the Group of Ten, pp. 9–10, para. 63.


Report of the Deputies of the Group of Ten, p. 10.


At the 1985 Annual Meetings of the Fund and the World Bank in Seoul, Korea, the Governor of the Fund and the Bank for the United States, Secretary of the Treasury James A. Baker III, outlined a program that included “increased lending by the private banks in support of comprehensive adjustment programs.”


The Articles also provide for uniform cancellation of SDRs in proportion of members’ net cumulative allocations.


There are also disadvantages of being a reserve country.


International Monetary Fund, International Monetary Reform: Documents of the Committee of Twenty (Washington, 1974), in particular, “Outline of Reform,” p. 14, para. 20; and Annex 5, pp. 37–40.


Examples have been given in Part One, Section I and include countries that may have lost creditworthiness for reasons other than their own inadequate policies or countries that may have failed to regain creditworthiness despite the pursuit of adequate adjustment policies for a sufficient length of time.


Cancellation is limited to the existing stock of SDRs, and the achievement of a reduction in the rate of growth of reserve supply is thus dependent on the existence of a sufficient initial stock of SDRs.


Report of the Deputies of the Group of Twenty-Four, p. 11, para. 104.


Report of the Deputies of the Group of Ten, p. 10, para. 72.


In these discussions, it was argued that substitution provided a means, other than SDR allocations, for “making the special drawing right the principal reserve asset in the international monetary system” (a requirement set forth in Article XXII of the Articles of Agreement).


A detailed discussion of the substitution account proposals can be found in “The Evolving Role of the SDR in the International Monetary System,” in International Money and Credit: The Policy Roles, ed. by George M. von Furstenberg (International Monetary Fund, Washington, 1983), pp. 475–535.


Report of the Deputies of the Group of Twenty-Four, p. 11, para. 105.


Ibid., para. 106.


Report of the Deputies of the Group of Ten, p. 9, para. 66.


Report of the Deputies of the Group of Twenty-Four, p. 10, para. 96; Report of the Deputies of the Group of Ten, p. 9, para. 66.


Report of the Deputies of the Group of Ten, p. 9, para. 67.

    Other Resources Citing This Publication