The seminar next turned to the question of the role that the SDR might play in the provision of unconditional liquidity to the Fund’s member countries if one looked beyond the traditional role defined in the Fund’s present Articles of Agreement. Three papers were presented, by Helmut Hesse, Ariel Buira, and Azizali Mohammed and Iqbal Zaidi (whose paper was co-authored by Muhammad Yaqub).

The seminar next turned to the question of the role that the SDR might play in the provision of unconditional liquidity to the Fund’s member countries if one looked beyond the traditional role defined in the Fund’s present Articles of Agreement. Three papers were presented, by Helmut Hesse, Ariel Buira, and Azizali Mohammed and Iqbal Zaidi (whose paper was co-authored by Muhammad Yaqub).

Is There Still a Rationale for the SDR as a Source of Unconditional Liquidity in the International Monetary System?

Helmut Hesse

In examining the question, “Is there still a rationale for the SDR as a source of unconditional liquidity in the international monetary system?” I should like to place the emphasis on the word “still.” There are good reasons for this: there have been fundamental changes in the international economic and monetary system since the mid-1960s when the creation of the SDR was being discussed. Besides the transition from the Bretton Woods system of fixed parities to one in which exchange rate arrangements can be freely chosen, further crucial changes have taken place. These include the enormous growth of the financial markets, the trend toward a multicurrency reserve standard, and the development of regional trading blocs. Although such blocs have been formed, all markets have become increasingly integrated globally—a process assisted by major advances in communication, an accompanying higher standard of information, and the ability of market participants to react more quickly to international developments. Against the backdrop of this completely changed environment, I believe that the question does indeed arise of whether there is still a need for the SDR.

The developments of the past three decades have provoked a radical change of mind on the part of many economists who had initially been strongly in favor of introducing or strengthening the SDR instrument. Nowadays there is a great deal of awareness that most of the reasons advanced for the introduction of the instrument are no longer valid under present circumstances. Since this instrument is in place, however, there is a desire not to forgo the option of employing it in future contingencies. This widespread attitude toward the SDR is aptly described in the well-known words of Professor Henry Wallich, who said, “Let’s put it on the shelf.” I too am inclined to endorse that pragmatic view. While working on this paper, I did, however, also wonder occasionally whether, from a purely economic viewpoint, one should not rather say, “Let’s get rid of it, it’s no longer needed.”

I suspect that many participants in this seminar take a different view. I should therefore like to take a broader look at the issues involved and at the debate thirty years ago when the SDR was introduced. In this, I shall confine myself to economic arguments; I shall deal only in passing with political aspects or legal issues.

This paper is structured as follows: I shall begin with a brief description of the reasons why the SDR system was created, after which I shall examine whether those reasons are still relevant and whether, in today’s international monetary system, there are new, additional arguments that might justify the use of the SDR instrument. After that, I shall deal with the question of whether, with regard to that instrument, the status quo should be maintained. In conclusion, I shall briefly mention the preconditions that would have to be fulfilled for the SDR to become the principal reserve asset of the international monetary system.

Reasons for Introducing the SDR Instrument and Its Envisaged Role

The scale of the global changes since 1969 becomes clear if one recalls those developments in the 1960s that led to the creation of an instrument for providing unconditional liquidity and why the role of the SDR was extended in the 1970s. I can deal with this point comparatively briefly, since it has already been discussed in other sessions of the seminar. The reasons for the introduction of the SDR may be summarized as follows:1

  • The main motive was the concern that a worldwide shortage of exchange reserves might occur and that deflationary trends in the world economy could result from this. The “traditional” sources of reserves (gold, U.S. balance of payments deficits) appeared to be inadequate, since less and less gold was finding its way into the exchange reserves, and the United States, the main reserve currency country, was striving for a rapid reduction of its balance of payments deficit. On the other hand, in view of the strong growth in world trade and in international movements of capital under conditions in which exchange rates were fundamentally fixed, a significant increase in reserve requirements was anticipated.

  • At the same time, the SDRs were supposed to supplement the limited physical gold reserves as a kind of artificial “monetary gold,” thus counteracting the further exchange of dollar reserves into gold and, ultimately, helping to defend the convertibility of the U.S. dollar into gold at the official price. As is well known, the ability to maintain that convertibility was increasingly being called into question at the end of the 1960s, as an ever-greater discrepancy emerged between the short-term liabilities of the United States to foreign monetary authorities, which were growing rapidly, and the continually decreasing gold holdings of the United States.

  • In the light of sharp price swings for gold and the U.S. dollar from the beginning of the 1970s, the objective was later added of replacing these previous cornerstones of the monetary system with the more stable SDR as the central point of reference, that is, as “numeraire,” of the system. Furthermore, as part of the efforts made to curb the role of the reserve currencies, the objective of making the SDR the principal reserve asset in the international monetary system was enshrined in the IMF Articles of Agreement in 1978.

  • Finally, it was also hoped that the SDR instrument would reduce the so-called asymmetries in reserve creation. The previous system had, above all, favored the main reserve currency country, the United States. It offered that country the opportunity of financing its balance of payments deficits by issuing its own currency, and thus of circumventing pressures to adjust, in addition to obtaining seigniorage profits. According to expectations at that time, these opportunities would be gradually reduced by the increasing use of the SDR (instead of the dollar). Simultaneously, the advantages associated with the creation of new reserves were supposed to accrue to all members of the Fund in relation to their quotas.

When the SDR instrument was created, it was laid down that SDRs must be allocated only “to meet the long-term global need, as and when it arises, to supplement existing reserve assets.” An allocation should be made “in such a manner as will promote the attainment of its [the Fund’s] purposes and will avoid economic stagnation and deflation as well as excess demand and inflation in the world.”2

The decision-making process that leads to an SDR allocation thus has two stages. First of all, the existence of a “long-term global need” must be firmly established. The words “as and when it arises” clearly point to the fact that the existence of such a need must not be taken for granted. Since an allocation can be made only if it is necessary “to supplement existing reserve assets,” it is clear, too, that it is not only the demand for reserves that needs to be taken into consideration but also the supply of monetary reserves. Thus, only the balance resulting from both variables can constitute a “global need.”3

If such a “long-term, global need” is established in a given situation, it must be met, that is, the size of the allocation must be determined in “such a manner” that it promotes the Fund’s objectives and has no inflationary effects. It is important to note that the absence of inflationary risks does not in itself justify an SDR allocation but is merely one aspect in determining the size of an allocation once the principal decision to allocate SDRs has been taken.

It follows unambiguously from the IMF Articles of Agreement that a long-term global need is the sole criterion that can justify an allocation. Article XVIII, which has the heading “Allocation and Cancellation of Special Drawing Rights,” mentions no other criterion. In this context, it is worth noting that the objective of making the SDR the principal reserve asset is not mentioned in Article XVIII, but in Article VIII, Section 7, and in Article XXII, where the members’ obligations to collaborate in matters of reserve policy and of the SDR in general are dealt with. To make this absolutely clear, the official commentary on the Second Amendment to the IMF Articles of Agreement states explicitly that the newly included objective of making the SDR the principal reserve asset of the international monetary system should not in any way change the existing principles for the allocation and cancellation of SDRs.4

Neither the IMF Articles of Agreement nor the 1968 report of the Executive Directors on the introduction of the SDR contain any specific statement of what is to be understood precisely by the term long-term global need. The report of the Group of Ten deputies does contain a number of indications, however.5 According to the report, “long-term” means that the provision of additional reserves must not be geared to short-term fluctuations in demand. “Global need” means the provision of adequate monetary reserves for the system as a whole, not the reserve requirement of individual countries or groups of countries. Moreover, the report points out that the SDR instrument should serve solely monetary purposes, and that its use must therefore not lead to a permanent transfer of real resources. Activating the SDR instrument for “short-term demand management” at the global level would likewise be unacceptable.

Even at that time, however, there were distinct reservations concerning the SDR instrument, inter alia, within the Group of Ten. The representative of one country in particular (which, interestingly, is now one of the most vehement advocates of SDR allocations) expressed his concern that the SDR instrument might be activated prematurely without real need and might thus weaken the willingness of deficit countries to adjust.6 Subsequent developments have shown that such fears were not unfounded.

Are the Reasons for Introducing the SDR Instrument Still Relevant Today, or Are There New Reasons?

Proposition 1:

The marked expansion of global liquidity, together with qualitative changes in the system, make the original objectives seem obsolete.

A large measure of consensus exists nowadays that the present international monetary system is capable of providing the reserves required worldwide even without recourse to an artificial reserve asset. The main reason for the creation of the SDR has thus become obsolete. This can also be demonstrated by figures: as Table 1 shows, since the introduction of the SDR in 1969, the growth of global reserves has been similar to that of global imports, which are regarded as a major indicator of the reserve demand. A global shortage of reserves cannot be deduced from this. Simultaneously, there have also been significant qualitative changes in the international monetary system, which have reduced the need for an additional artificial reserve asset as a liquidity reserve, a store of value, and a medium of transaction.

Table 1.

Trends in International Monetary Reserves and World Imports Since 19691

article image
Sources: International Monetary Fund, IFS Yearbook and Direction of Trade Statistics Yearbook; and Deutsche Bundesbank, Exchange Rate Statistics (Statistical Supplement to the Monthly Report 5).

End-of-year levels

The monetary reserves (including gold) denominated in SDRs in the IFS Yearbook have been translated into U.S. dollars for the entire period using the U.S. dollar values for the SDR given in the Exchange Rate Statistics.

The strong increase in the supply of reserves and the qualitative changes in the system are due mainly to the following developments.

  • The most important factor is probably the rapid growth of the international financial markets. Creditworthy countries are nowadays able to meet any appropriate (additional) reserve requirement by comparatively low-cost borrowing in those markets and thus can cushion temporary balance of payments disturbances without difficulties.7 The problem of asymmetries in reserve creation that was perceived formerly has been appreciably reduced by this development, since creditworthy countries now have the opportunity of financing balance of payments deficits, or a desired accumulation of reserves, by issuing debt instruments in the international financial markets. In that respect, what is relevant today is not so much the difference between reserve currency and nonreserve currency countries, but that between creditworthy and non-creditworthy countries. The pressure to adjust arising from the outflow of monetary reserves in the former par value system has thus been replaced to a certain extent by the pressure to maintain creditworthiness vis-à-vis the markets by pursuing sound economic policies.

    Of course, the growth of the financial markets has not brought only advantages, but has also caused new problems. I shall come back to this point in the next section; at this stage, I wish to confine myself to pointing out that, as a result of the development of the financial markets, some core weaknesses of the former gold dollar standard, which the SDR instrument was intended to overcome, no longer exist.

  • A second important development was that, contrary to original expectations, the balance of payments deficits of the United States have not declined significantly, but have, in fact, increased. They have thus remained an important source of reserves for the system.

  • Third, the change from the par value to the present open exchange rate system should also be highlighted. This development freed Fund members from the necessity of having to defend their parities with interventions. As a result, the room for maneuver in reserve policy has become distinctly greater.

  • A fourth central factor is the continuing trend toward a multipolar international monetary system, in which other currencies besides the U.S. dollar—such as the deutsche mark and the yen—have gained increasing importance as reserve currencies. Because of this trend, dependence on the main reserve currency, the U.S. dollar, which is occasionally subject to sharp exchange rate swings, has been significantly reduced and the diversification of monetary reserves made considerably easier. Interest on the part of official and private institutions in a greater use of the SDR has consequently declined. Moreover, the reserves available in the multicurrency reserve system are of a higher quality than the SDR, as they can be used directly for intervention purposes and do not first have to be exchanged for usable currencies.

  • For the sake of completeness it should be mentioned that it is also no longer necessary to strengthen confidence in the convertibility of the dollar into gold by recourse to an artificial reserve asset, as that convertibility was abolished as long ago as 1971.

In view of these fundamental changes in the international monetary system, it is hardly conceivable today that a global need to supplement existing monetary reserves will arise. For that reason, attempts are occasionally made to justify SDR allocations by interpreting the global need concept in a new way. From the sharp growth in reserves that has actually been observed in the past, this interpretation extrapolates a corresponding increase in the need for reserves in the future, a need that is supposedly to be met, at least in part, by SDR allocations. This approach, however, is at odds with the global need concept as interpreted in conformity with the Articles of Agreement and as called for in terms of anti-inflation policy. An additional global need for reserves cannot be deduced simply from the demand for reserves, but only from the difference between the global demand for and global supply of reserves. In other words, it has to be examined whether the “legitimate” global demand for reserves can be fully met by the conventional reserve sources. Such an examination will inevitably lead to the conclusion that nowadays every “legitimate” demand for reserves can be and is, in fact, satisfied by the highly liquid international financial markets.

Proposition 2:

The SDR is not an appropriate means of reducing the weaknesses of the present monetary system.

It is an undeniable fact that, in the present international monetary system, a number of shortcomings have emerged that have to be considered as serious weaknesses. These weaknesses include, principally:

  • First, the fact that the enormous growth of the financial markets has also produced an enormous “financial superstructure” on top of the real economy. Since the growth of the financial markets is subject to no kind of quantitative control, there exists an increasing risk of financial flows becoming divorced from trends in the real economy, and thus of fostering inflationary tendencies. In my view, this fact is a particular cause for concern. Furthermore, the increasing integration of the financial markets leads to a loss of monetary policy autonomy, thus making it more difficult to achieve price stability and/or exchange rate targets.

  • The financial markets’ size and mode of operation also encourage the emergence of volatile capital movements, which can result in various kinds of misalignments and imbalances. In countries with unexpected capital inflows, there may be an unwanted appreciation of the currency or a problematical expansion of the domestic money stock with matching inflationary risks, whereas countries with unexpected capital outflows are confronted with depreciations or deflationary trends domestically. The problem is frequently exacerbated by the fact that countries take up very short-term loans to obtain reserves, loans that are fraught with particular risks with regard to their renewal and interest rate changes. The comparatively large share of borrowed reserves in total global reserves, which has arisen in this way, thus places an additional strain on the system.

  • The fact that many countries have no or only very limited access to financial markets is also considered a problem. Although this is often the result of unsound policies that squander creditworthiness, there are also system-immanent factors, which may play a part in the loss of market access. Private lenders, for example, often tend to finance an unsound economic and monetary policy for too long, and then, as an overreaction, to end their financing abruptly. A financial crisis triggered by such behavior may subsequently also spread to other countries (“contagion effects”) that pursue sound policies.

  • Finally, in today’s multicurrency reserve standard with flexible exchange rates between the major currencies, the problem of adequate protection against exchange rate risks increasingly arises. Although a diversification of reserve holdings is possible through the markets, this can cause new or additional volatility in the markets and unwanted exchange rate fluctuations.

In order to alleviate somewhat these shortcomings of the system, “steady periodic SDR allocations at a moderate rate” have been proposed for gradually increasing the share of reserves that are “owned” and the use of which can be controlled by the IMF.8 This idea is based on the assumption that allocated SDRs will replace borrowed reserves, not least because holding SDRs that have been allocated free of charge is cheaper than holding borrowed reserves. This strategy envisages that fluctuations in the global reserve requirement should be met by a corresponding increase or reduction in the allocation rate; that is, if there is an excess supply of reserves, the allocation rate would merely be lowered below the multiyear average, but there would be no cancellation of SDRs. In this way, the proponents of this idea hope to progress toward the objective of making the SDR the principal reserve asset.9

Such proposals are fraught with problems. The obvious deficiencies of the present international monetary system would not be reduced by SDR allocations but, if anything, increased. It is simply unrealistic to assume that SDR allocations would merely replace reserves from other sources; it is, instead, much more likely that deficit countries would take up additional loans in the international financial markets on the basis of the allocated new “owned reserves.” The expansion of international liquidity would thus be given additional momentum, and the necessary adjustment processes would be further delayed.

It would be counterproductive if the response to a sharp growth in international liquidity were to consist of additional, disproportionately high SDR allocations in order to increase the share of SDRs in international exchange reserves. This would not only encourage additional inflationary expectations but also undermine confidence in the IMF as the guarantor of the monetary system’s stability. Finally, this would impair the quality and acceptability of the SDR as a reserve asset in the longer run. As SDR holdings in the world economy would grow rapidly in such a scenario, it is likely that increasing doubts would arise as to the exchangeability of SDRs into usable currencies. As a result, countries in a strong reserve position would probably initially show restraint in agreeing to voluntary SDR transactions and subsequently perhaps limit their SDR acceptance obligation by opting out of further SDR allocations. Even the possibility of some participants leaving the SDR Department—membership in which, as is well known, is voluntary for all Fund members—could not be excluded under such circumstances.

What is also problematic is the attempt to justify SDR allocations by the objective of making the SDR the principal reserve asset of the monetary system, which was introduced into the IMF Articles of Agreement in 1978 by Article VIII, Section 7. This objective is subordinated, after all, to the primary objective of controlling international liquidity, which was introduced into the IMF Articles at the same time. Since then the financial system has developed in such a way, however, that the implementation of these two objectives has become increasingly unrealistic.

Apart from that, within the Fund’s membership there have from the outset been differing interpretations of the goal of making the SDR the principal reserve asset of the system, that is, qualitatively, on the one hand, and quantitatively, on the other. This objective has probably been achieved to a great extent in qualitative terms, since the SDR has largely assumed the formerly predominant role of gold within the IMF. It has become one of the most important media of transactions among Fund members and is comparable with other reserve assets in terms of its remuneration and development of its value.

With regard to the quantitative aspect of the future role of the SDR, it has to be borne in mind that the aforementioned objective remains subordinated to the global need criterion as the sole precondition for allocations. In this connection, as mentioned above, the commentary on the Proposed Second Amendment of 1976 clearly states that “the general obligation of participants to collaborate… has been broadened by adding the objective of making the special drawing right the principal reserve asset of the international monetary system…. The principles for the allocation and cancellation of special drawing rights remain unchanged…” (pp. 70–1.). SDRs cannot play a major role in the monetary system in quantitative terms as long as there is no global need to supplement existing reserves with SDRs.

To conclude this section, it should be noted that, under present conditions, the SDR instrument cannot be used to alleviate the shortcomings of the international monetary system. Other strategies have to be employed for that purpose, in particular the IMF’s surveillance of its member countries’ economic policies and close cooperation among its members.

Proposition 3:

The SDR is not a suitable instrument for financing adjustment, development, and growth programs.

As the original reasons for the creation of the SDR have lost much of their earlier significance, increasing calls have been made over the past few years to use the SDR instrument for purposes for which hitherto, for good reasons, only conditional funds were used. Such proposals include special allocations to “poor” countries, that is, developing countries and countries in transition with weak reserve positions, similar to the “link” schemes of the 1960s and 1970s. Here, the offer is occasionally made to tie such allocations to the reintroduction of the reconstitution requirement. This is intended to dispel concerns that the allocated SDRs will not be held, but quickly spent by the recipients. There are also proposals that the industrial countries should contribute allocated SDRs to a trust fund, which could provide additional financial assistance to Fund-supported adjustment programs.

One of the reasons given for these proposals is that the poor countries are experiencing increasing difficulties in adding to their reserves because, in view of the budgetary problems of many industrial countries, flows of bilateral aid have become more sparse and alternative ways of obtaining reserves—such as borrowing in the international financial markets or achieving adequate current account surpluses—are too costly for them or entail excessive burdens of adjustment. Moreover, the additional demand to be expected from the recipient countries could, it is said, contribute to overcoming the sluggishness of growth in industrial countries, whereas inflationary effects are not to be feared in view of the large overhang capacity in industrial countries and the limited volume of such SDR allocations. It is also claimed that a more uniform distribution of exchange reserves could ultimately be achieved in this way.

A fundamental objection to these arguments is that support payments to needy countries—which I regard as important and desirable—should be granted through the official development aid instruments designated for that purpose. Providing freely usable exchange reserves through SDR allocations for these ends would be inappropriate and contrary to the system (see below). Apart from that, under today’s conditions, if individual countries or groups of countries have insufficient reserves, this is frequently the result of unsound economic policies and not the outcome of an overly tight global supply of reserves. Appropriate changes in economic policies must therefore be the principal response to such a shortage of reserves, and not the provision of unconditional low-cost financial resources, which, as experience shows, is an incentive to postpone the inevitable processes of adjustment. If a temporary external financing requirement arises in the context of such a strategy, it is possible to draw on appropriately conditioned balance of payments assistance from the IMF and, possibly, on supplementary financial assistance from the international community; moreover, external capital inflows soon start up again if adjustment processes are progressing successfully, as the example of a number of Latin American and Asian countries has recently shown.

Thus, there exists no need for the provision of additional unconditional liquidity if there are sufficient adjustment efforts. Such a provision of unconditional liquidity would not be prudent either, since it would undermine existing pressures to adjust. That would be in the interests neither of the international community nor of the country itself, since the cost of adjustment, ultimately, is all the higher later on if inevitable processes of adjustment are delayed. To ensure a suitable relationship between adjustment and financing, it is important for the Fund’s financial assistance to remain appropriately conditioned and for the Fund to confine itself to the role of a catalyst in the funding of stabilization programs.

The reason that is sometimes given for calling for SDR allocations in favor of poor countries is not that those countries have a lack of reserves but that it is necessary to promote development, transformation, and growth. This is intended as a justification for SDR allocations even if the groups of countries concerned have no discernible shortage of reserves. The argument against this line of reasoning is that goals such as development, transformation, and lasting growth are longer-term tasks, the financing of which logically also requires longer-term capital, and not short-term liquidity. The Fund can provide neither conditional nor unconditional funds for long-term financing of this kind. Such financing requirements must instead be covered by funds that are, ultimately, obtained by reduced consumption. Primarily, private investment has a proper role to play here, as do loans by the development banks and donor country assistance.

By contrast, as a monetary institution, the IMF can only grant financial assistance for overcoming temporary balance of payments imbalances and for regaining macroeconomic stability. As a sustainable balance of payments position and macroeconomic stability are basic preconditions for growth and development, the Fund thereby makes a major indirect contribution to achieving those goals; it is not able, however, to promote them directly. In this context, I refer to Article I (ii) of the IMF Articles of Agreement, which states that the Fund should make only an indirect contribution to “high levels of employment and real income” and “the development of the productive resources of all members” by facilitating “the expansion and balanced growth of international trade.”

Apart from that, there is scarcely any evidence for links between the size of exchange reserves—which can be influenced by adjustment programs supported by the Fund—and longer-term growth processes. This is indicated, at any rate, by a current comparison between the industrial countries (large reserves with low growth) and the newly industrializing economies (initially low reserves with high growth).

Allow me to add some specific comments to the more general considerations outlined above. First, it is not clear whether the groups of developing and transitional countries actually do have an acute shortage of reserves as claimed. With regard to the developing countries, the important ratio of “exchange reserves to imports,” for example, was more favorable in the mid-1990s than at any time over the past two decades. Among the countries in transition, there are without doubt many countries with large financing requirements; some of them, however, have been able to overcome initial shortages of reserves by pursuing sound economic policies, thereby demonstrating how a reserve problem can be mastered by countries’ own efforts.

Even if there were a shortage of reserves in a number of countries, this would not in itself constitute a “global need … to supplement existing reserve assets” within the meaning of Article XVII of the IMF Articles of Agreement, as has already been explained in detail. This can be demonstrated by comparing it to the situation at the national level: individual regions’ special financing problems cannot be solved solely by monetary policy means but can be mastered only through budgetary assistance and/or private loans in conjunction with necessary structural adjustments. A national central bank would lose control of domestic liquidity if it took on regional problems. Much the same applies to the work of the IMF.

An additional factor is that SDR allocations without a real global need would stir up inflationary expectations that would make the world economic environment worse in the longer term and tend to make adjustment processes more difficult. At the risk of some exaggeration, it could be said that there is no adjustment without pressure to adjust and that without adjustment there is no sustained growth! Finally, the availability of new reserves might give rise to the temptation to put off necessary adjustment measures.

The proposed funding of development aid through SDR allocations would involve numerous problems. In view of the expected long-term demand for official development assistance, and bearing in mind the fact that most governments face severe budgetary problems, there is cause to fear that extensive SDR allocations might be made on a regular basis, with a corresponding destabilizing impact on the international monetary system. This would, moreover, result in a permanent transfer of real resources by means of the SDR instrument—something that was specifically ruled out when it was created and that also cannot form any part of a monetary system geared to stability. In the end, transferring real resources by monetary means would raise serious doubts concerning constitutional legitimacy—in countries with a democratic constitution, at least. A redistribution of resources without a simultaneous disclosure of who is going to be taxed would breach basic principles of budget law. But that is not all: because of the well-known risk that policymakers will attempt to use monetary instruments to avoid necessary adjustment measures, it is now a widely accepted principle that central banks should be independent of government and parliament so that the stability of the monetary system can be better maintained. Since the Fund is, by definition, not independent of the governments, but the guardian of international monetary stability, it ought to be a matter of course for it to reject all notions that amount to a transfer of real resources using monetary instruments.

Finally, a transfer of real resources would be likely to have a serious impact on the economic structure of the producing countries. If the SDR-financed demand encounters utilized capacities in the producing country, a mechanism sets in that pushes back real demand. That can either be inflation or appreciation. Demand is then also pushed back for products of sectors where comparative advantages and disadvantages interface in the international division of labor.

It is also questionable whether a more even distribution of reserves could be achieved by targeting special SDR allocations to certain groups of countries. This would be possible, at most, in an initial step, but hardly on a permanent basis. Rather, it would have to be expected that those countries—for reasons that are quite understandable—would soon spend a large part of their newly allocated SDRs. This would be even more likely if there were a prospect of further regular and large-scale SDR allocations. Ultimately, it is likely that the SDRs originating from special allocations would flow to the countries with large reserves, and the discrepancy that already exists in the distribution of reserves would become even greater as a result.

Reintroducing the reconstitution requirement (which existed up to 1981) would contain such a development only if very far-reaching—and not just marginal—reconstitution requirements were agreed upon, and if appropriate powers to enforce them were conferred on the Fund. That would probably be difficult to achieve politically, however. Furthermore, in the past the reconstitution requirement was regarded as a severe bias against the SDR and would probably be called into doubt once again as soon as a substantial number of countries had increasing difficulty meeting it.

At the conclusion of this section I should like to address a special problem that has been a subject of lively debate for some time. That is the question of whether those (new) members that have not participated in previous SDR allocations should receive a one-off “equity allocation” to enable them to take full part in the SDR system. This call for an equity allocation cannot be justified in economic terms, as the SDR instrument was created to meet a global need for reserves and not to satisfy the need of a selected group of countries only.

For political reasons, there is nonetheless an apparent widespread willingness—even by those opposed to general SDR allocations—to approve a one-off special allocation in favor of those new members. Although this is not convincing from a purely economic viewpoint, one can only hope that any damage such allocations might cause would be limited. What is of paramount importance here is to ensure that such an equity allocation takes place only on the basis of an amendment to the IMF Articles of Agreement that clearly stresses its exceptional one-off character.

In summary, it can be said that the original reasons for the creation of the SDR system have to all intents and purposes ceased to apply, that the sole allocation criterion of a “long-term global need … to supplement existing reserve assets” is, at least for the foreseeable future, unlikely to be met, and that in the present situation the necessary preconditions for further SDR allocations therefore do not exist. Attempts to assign other or additional functions to the SDR instrument would be contrary to the original intentions and would be economically questionable, as these are tantamount to a desire to use monetary means to solve problems of development financing and of the international distribution of income. Bearing this in mind, the question arises of whether there are any reasons at all for maintaining the status quo with regard to the SDR.

Should the Existing SDR System Be Abolished or Are There Reasons for Maintaining the Status Quo?

In examining this question, it may be assumed that the emergence of a global need for supplementing existing reserves is highly improbable for the foreseeable future, although the possibility of this happening cannot be ruled out. Such a global shortage of reserves could nevertheless be met quickly and effectively by drawing on the extensive, appropriatety conditioned credit facilities of the IMF or through the use of existing bilateral lines of credit between central banks. Recourse to the SDR instrument would not be necessary.

In the 1985 report of the deputies of the Group of Ten, the proposition was put forward that the SDR instrument could be maintained as a safety net against potential functional crises in the international financial markets—as we experienced recently in connection with Mexico.10 This would not, however, conform with the role originally intended for it. Besides, it is unlikely that the SDR instrument could be activated at short notice—as is required in the event of a crisis—since an allocation would have to be based on a protracted procedure of approval within the Fund and would need ratification by parliaments in a large number of member countries. In addition to that, employing the SDR instrument selectively in favor of those countries principally affected would not be possible either, because the Fund is only entitled to make general SDR allocations, that is, to work here with the “watering can,” so to speak. In the final analysis, employing SDRs in this way would have to be weighed against the risk of worldwide inflationary effects likely to be associated with a general allocation. Any crises will therefore probably have to be overcome primarily by targeted, coordinated assistance by the IMF and other creditors. Employing the SDR instrument for such purposes does not appear to be reasonable.

A possible argument in favor of maintaining the SDR as a reserve asset is that it exercises important functions within the IMF, such as being a transaction medium and store of value, and, in that respect, has partly taken over the former role of gold. Nevertheless, it is probable that these functions could largely be assumed by the five currencies represented in the SDR basket, for example, or by one of those currencies alone. The subscription payments that are to be made in SDRs, for instance, could be made entirely in usable currencies. Incidentally, abolishing the SDR reserve asset would not necessarily prevent the IMF from continuing to use the SDR unit in its accounts, that is, translating all its currency holdings into SDR units defined by a currency basket. At the same time, members could be placed under an obligation, as before, to keep the SDR equivalent of the Fund’s holdings of their currencies at a constant level. This would mean that, if their currencies depreciated against the SDR unit of account, they would continue to have to pay additional amounts of currency into the Fund and vice versa.

It can be concluded from the above that, in economic terms, even maintaining the status quo with regard to the SDR does not appear to be particularly convincing. As a political realist, I have doubts, however, whether radical conclusions will be drawn from the cited economic arguments that call the SDR into question.

Conditions That Would Have to Be Fulfilled for a Further Development of the SDR System

To conclude my paper, I should like to outline briefly the conditions that would have to be fulfilled if the SDR instrument were to be developed in line with the role originally intended for it. As is well known, in the discussions on reform at the beginning of the 1970s, the proposal was made to make the SDR the principal reserve asset in a system with the core elements of fixed but adjustable exchange rates, symmetrical processes of adjustment with asset settlement, and the payment of foreign exchange reserves into a substitution or consolidation account, as well as an appropriate control of all sources and types of international liquidity.11 So as not to create any misunderstanding, I should like to emphasize straightaway that the setting up of a system of that kind and a concomitant rise in the status of the SDR cannot be expected in the foreseeable future. I nevertheless feel it is useful to indicate the most important conditions that would have to be met to justify a more intensive use of the SDR instrument.

To make the SDR the principal reserve asset in the area of official transactions in quantitative terms, the use of the former reserve currencies would have to be reduced accordingly and replaced by SDRs. A major step toward achieving that goal would be the establishment of a so-called substitution account,12 into which Fund members would have to transfer a large part of their official reserves in exchange for SDRs, which they could use only under the Fund’s control in case of a balance of payments need.

A major problem in introducing a substitution account of this kind, however, would be that of meeting possible account losses, which can arise from differing trends in the interest and exchange rates of the SDR and of the counterpart reserve currencies that have been transferred to the substitution account. From a global point of view, it would be desirable for the respective account losses to be offset by the countries whose currencies have depreciated against the SDR. This would subject these countries to a certain pressure to keep the SDR value of their currencies stable in the longer term by pursuing suitable economic policies.

From an economic point of view, the setting up of a substitution account would have several advantages: the diversification of exchange reserves would be made easier and would no longer burden foreign exchange markets; the greater part of the official exchange reserves would be subject to a certain degree of international control; and a step in the direction of symmetrical adjustment would be made, since the reserve currency countries, too, would be subjected to a certain pressure to adjust (through the loss compensation scheme). At the same time, however, there would be the danger of potential temporary crises in the financial markets if—owing to the reserve currencies’ decreasing importance—there were private speculative movements out of assets denominated in these currencies. Furthermore, from the national perspective, economic policy autonomy would be restricted.

It is mainly for the last-mentioned reason that the substitution account probably has no prospect of being realized. The two rounds of discussions on this subject in 1973–74 and 1978–80 showed that the majority of the Fund’s members do not wish to relinquish their unrestricted power of disposal over the greater part of their reserves. In addition, it is not likely that agreement on covering potential account losses could be achieved under present circumstances.

The establishment of a substitution account would, moreover, not be sufficient to control international liquidity, since it would embrace only the official reserves and thus no more than a comparatively small part of overall international liquidity. Therefore, the introduction of more far-reaching instruments would be required with the aim of influencing the large international financial markets. To this end, the IMF would probably have to be upgraded to a kind of “world central bank” and equipped with efficient monetary policy instruments, in particular with the power to pursue open market policies with SDR-denominated assets and issue SDR assets for that purpose, supplemented perhaps by the power to impose non-interest-bearing minimum reserve requirements.

A core problem of this approach would be that, because all markets are highly integrated, policy measures taken by the Fund in the international financial markets would probably have major repercussions on the national financial markets, and would thus make it much more difficult to pursue an autonomous national monetary policy. A liquidity policy that is perhaps quite reasonable in global terms might nevertheless be inappropriate for individual countries and have adverse effects on their economic development—by fostering inflationary trends, for example. The question would also have to be settled of how any losses arising from a global liquidity policy by the Fund ought to be met. Such losses could always arise if the interest income from the currencies accepted were lower than the interest to be paid on the SDR instruments that had been issued. The problem would probably not arise if the Fund were to rely principally on a minimum reserve instrument as part of its global liquidity policy. It is likely, however, that it would be very difficult to enforce such minimum reserve requirements worldwide, especially in offshore centers. From a political perspective, it should also be said that many countries would probably not be willing at the moment to transfer such far-reaching powers to the Fund, and thus surrender a large part of their monetary policy autonomy. Any thoughts of making the IMF a kind of world central bank can therefore, under the present circumstances, only be described as “utopian.”

Should the Fund ever achieve an effective control of international liquidity by means of the SDR instrument, it would become especially important to ensure that SDR allocations were not made for inappropriate reasons. Since, in such a world, the SDR would probably be one of the most important reserve assets in terms of quantity, any allocations would attain a substantial magnitude, and false decisions on allocations would therefore have a particularly severe impact on the international monetary system. With that in mind, it would probably be sensible to develop rules or to formulate intermediate targets (similar to those for monetary targeting at the national level) for international liquidity management. Formulating such intermediate targets is likely to be much more difficult at the international level than at the national level, however. For that reason, a global liquidity policy would be faced with enormous problems, and it would remain to be seen whether the Fund could overcome those difficulties.

It ought to have become sufficiently clear from the description of such a hypothetical system that establishing the preconditions outlined above will scarcely be possible in the foreseeable future. This simultaneously emphasizes the doubts mentioned above about the usefulness of continuing the present SDR system. Ultimately, those who wish to make the SDR the principal reserve asset of the system ought to avoid anything that might arouse additional doubts about the instrument’s constructive role.

The Potential of the SDR for Improving the International Monetary System

Ariel Buira 1

This paper focuses on the shortcomings of the current arrangements for the provision and distribution of international liquidity and how they affect the working of the adjustment process and the international trade and payments system. The case is made that despite the rapid growth of increasingly integrated international capital markets, the prevailing mechanisms for the creation and distribution of world liquidity result in a system that is economically inefficient as well as inequitable. To address the shortcomings of the current system, a proposal is developed based on an alternative arrangement, which combines the Fund’s ability to create liquidity through SDR allocations and its surveillance function, to distribute international liquidity according to estimated needs.

Shortcomings of the Current International Monetary System

Among the objectives of the International Monetary Fund as envisaged by its founders was “to assist in the establishment of a multilateral system of payments in respect of current transactions between members and in the elimination of foreign exchange restrictions which hamper the growth of world trade.” Moreover, the IMF would “give confidence to members by making the general resources of the Fund temporarily available to them under adequate safeguards, thus providing them with the opportunity to correct maladjustments in their balance of payments without resorting to measures destructive of national or international prosperity.”2

In 1969, 25 years after the Fund was established, the First Amendment of its Articles of Agreement expanded its responsibilities in matters concerning international liquidity through the creation of the SDR. The Amendment sought to deal with the international community’s dissatisfaction with a monetary order in which the main source of international liquidity was the U.S. balance of payments deficit, by enabling the Fund to increase existing reserve assets in a timely manner whenever the need arose. In this way, the Amendment laid the foundations for a more rational and effective mechanism to adjust the supply of international reserves to the requirements of the world economy.

However, 27 years later, little progress has been made, as only a minority of countries have adequate access to international liquidity. In this regard, countries may be divided into four groups.

  • Reserve currency countries: mainly the United States, Germany, and Japan, whose currencies are accepted as reserve assets and used as vehicles for international payments. These countries enjoy the privilege of not having to pay for their external deficits with goods and services.

  • Countries with ample market access: mainly industrial countries that face a very elastic supply of liquidity at market interest rates. These countries can increase their foreign exchange reserves at a relatively low cost, either by purchasing reserve currencies with their national currency in the exchange market or by borrowing balances of reserve currencies in private capital markets or from other central banks.

  • Countries with limited market access: mainly middle-income developing countries that face a steeply sloped supply curve of international liquidity. For this group of countries, holding borrowed reserves imposes a significant interest cost equal to the spread between the borrowing rate and the rate of return on reserve assets.

  • Countries with no market access: mainly low-income developing countries that have to run current account surpluses in order to accumulate reserves. For this group of countries, the cost of acquiring reserves is the forgone consumption of investment implied in the generation of a current account surplus.

For the first two groups of countries, the current system works well. The supply of liquid funds that they can hold as reserves or use to finance payments deficits responds flexibly to changes in the demand for such funds. These countries can follow their chosen economic policies and maintain free trade and payments systems, since they face few risks derived from eventual liquidity constraints. The other two groups of countries, which comprise the vast majority of Fund members, are less fortunate.

The current arrangements for the provision and distribution of international liquidity give rise to a number of important shortcomings that, if overcome, would lead to a better performance of the world economy. Among these, the following may be considered:

  • Seigniorage and asymmetry in access to liquidity.

  • Insufficient financing of adjustment programs.

  • A deflationary bias resulting from inadequate access to international liquidity.

  • Volatility of capital flows.

Seigniorage and Asymmetry in Access to International Liquidity

Reserve currency countries with payments deficits have the freedom, at times very considerable, of settling them without undertaking adjustment policies. The acceptance of their currencies by the rest of the world allows them to finance their payments deficits simply through the issuance of currency. That is to say, unlike other countries, they do not have to pay for their trade imbalances through the sale of goods and services or by running down their international reserves. An undesirable feature of this arrangement has been the absorption of a substantial portion of the world’s net savings by the United States rather than by other less developed countries.

Although it is not possible to quantify the benefits of seigniorage that accrue to these countries, the simple sum of the holdings of their currencies by other central banks produces staggering amounts. At the end of 1995, total world foreign exchange holdings amounted to $1,320 billion.3 Out of this total, 79.5 percent corresponded to the three main reserve currencies: 58.1 percent to the U.S. dollar, 14.1 percent to the deutsche mark, and 7.3 percent to the Japanese yen.4 While these reserve holdings receive interest, the many billions of dollars and of other reserve currencies in circulation either as a store of wealth or as a means of payment in the former Soviet Union, Eastern Europe, Latin America, Asia, and Africa constitute huge transfers from poorer to richer countries in the form of interest-free loans.

The United States has been the largest beneficiary of these transfers, but over the last decade, as the share of the deutsche mark and the yen in international reserves has increased, Germany and Japan have also benefited from having their currencies become a major reserve asset and a vehicle for international payments.5

Other industrial countries have easy or large access to credit from international capital markets and are thus able to sustain large fiscal or external imbalances for long periods. Most developing countries, on the other hand, benefit neither from the creation of international liquidity and the seigniorage that benefits reserve currency countries nor from easy access to international financial markets and are therefore subject to a stricter discipline. Indeed, to the well-known asymmetry resulting from the burden of adjustment falling on deficit countries—as surplus countries have little pressure to correct their surplus—an additional asymmetry is added: that resulting from an unequal access to international financing and from an inequitable distribution of the benefits of international liquidity creation.

Insufficient Financing of Adjustment Programs

To attain the goals established in its Articles of Agreement, the Fund would make its general resources temporarily available to its members under adequate safeguards, “thus providing them with the opportunity to correct maladjustments in their balance of payments without resorting to measures destructive of national or international prosperity.”6

In practice, however, an IMF-sponsored program does not necessarily lead to a successful adjustment. The evidence indicates that the availability and cost of financing have important consequences for the quality and effectiveness of countries’ programs for macroeconomic stabilization. A study of 266 IMF-funded programs during 1979–89 found that 52 percent of them failed in that they did not achieve their objectives and the countries were unable to disburse all resources available to them.7 The programs classified as successful by the authors received on average IMF support that financed 17 percent of their imports, whereas the programs that failed received only 6 percent. These results suggest that the probability of achieving a successful adjustment is enhanced when adequate financing is available. Conversely, when programs are provided with little financing they will be harsher and, thus, more prone to failure.

When financing is inadequate, the adjustment necessary to eliminate a balance of payments deficit may have to be extremely severe, with high costs in terms of national income and employment.8 Moreover, the conflict implied over the distributional effects of adjustment is exacerbated. Adjustment programs should take due account of the origins of an imbalance and provide the resources required to correct it over a reasonable period. A focus on aggregate demand management to correct external imbalances is adequate in many cases, but it is not necessarily appropriate when the disequilibria are of a structural nature. When this is so, the measures adopted to regulate aggregate demand only suppress or alleviate the symptoms of the imbalance but do not correct the underlying disequilibria. The recessionary adjustment associated with these programs implies high political and social costs. These costs, in turn, reduce the possibility of bringing the program to a successful conclusion.

Moreover, in order to strengthen the external payments position, adjustment programs generally imply a tightening of the fiscal stance even in countries in which public finances are already in balance. Often, to maintain fiscal balance in the face of a decline in activity requires that public investment in infrastructure, research and development, and other spending be reduced, postponed, or eliminated, and on many occasions health and education programs also suffer. This introduces additional costs because the correction of certain structural imbalances, requiring public expenditure, may be delayed, and the productive capacity of the economy over the medium and long terms impaired. However, because their effect is not always felt immediately, these costs are seldom considered.

The combination of lower economic activity and sharply higher interest rates that often accompanies the adjustment program has a disastrous impact on small and medium-sized businesses, causing loss of capital and employment with the consequent human distress. In addition, such a combination leads to an increase in nonperforming loans that may put financial institutions under strong pressure. The ensuing halt in the provision of new credits and in the rollover of outstanding loans adds to the disruptions in economic activity. Moreover, the fiscal cost of programs designed to support the financial sector to overcome these difficulties can be considerable and can extend over a prolonged period.

Recessionary programs lead to a waste of resources that ultimately produces considerable economic and social costs for the country in question and for the global economy as a whole. Therefore, it is in the interest of the international community to minimize these costs. To help economies adjust without undergoing the recessionary features of traditional programs, in October 1985 the IMF adopted the strategy of “growth-oriented adjustment,” a concept introduced by former U.S. Secretary of the Treasury James Baker.9

Clearly, adjustment programs consistent with economic growth require a much greater amount of financing than that provided under traditional programs. Indeed, the funds available should be consistent with the attainment of a minimum acceptable rate of economic growth. This target rate could, for instance, be the rate of growth of population or that of the labor force. In many cases even a program constructed to sustain zero growth would be an improvement on a program that implies a sharp decline in GDP.

Programs designed to promote adjustment with growth would have to go beyond the usual emphasis on aggregate demand management. In particular, such adjustment programs would not merely aim at the immediate restoration of equilibrium in a country’s external accounts, since this can always be attained through deflation. As a minimum they would be designed to sustain levels of investment required to preserve the country’s physical and human capital.

Unfortunately, since the quota increases have lagged far behind the financing needs of IMF member countries, the size of the Fund has declined from 7 percent of world trade in 1950 to less than 2 percent at present, which has sharply reduced the Fund’s ability to support adjustment with growth and in practice this concept has rarely been applied. To avoid the recurring difficulties associated with quota increases one could think of a Fund based entirely on SDRs as proposed by Polak (1979).

Deflationary Bias Resulting from Inadequate Access to International Liquidity

Under the present arrangements for the creation and distribution of international liquidity, access to it is concentrated in industrial countries and a few developing countries. As Table 1 shows, at the end of the first half of 1995, 22 industrial countries and 18 developing countries with access to international capital markets controlled 88.4 percent of international reserves. The rest of the world (more than 130 countries) have historically held a very low share of total world reserves in addition to having little access to private international financing. As a result, these countries are forced to “acquire more adequate reserves only through expensive borrowing in the market—in the few cases where markets would be prepared to lend—or compression of domestic demand and imports that would be inimical to their adjustment, reform, and growth efforts” (Camdessus, 1994b).

Table 1.

Shares in Total Reserves by Country Group1

(In percent)

article image
Source: International Monetary Fund, international Financial Statistics.

Figures are end-year except for 1995 (end-June).

France, Germany, Japan, the United Kingdom, and the United States.

Developing economies with limited or no market access are not allowed to exploit their growth potential because of a lack of liquidity. Indeed, they are forced to rely almost exclusively on exports to satisfy their liquidity needs. Moreover, these countries do not have a diversified export base and thus are very vulnerable to swings in commodity prices. Their limited access to international liquidity forces them either to restrict aggregate demand or to constrain their imports to their limited supply of foreign exchange. But, of course, restricting imports implies restricting investment, which ultimately translates into low rates of economic growth. In this way, the lack of international liquidity introduces a deflationary bias into the performance of these countries and, through lower international trade, on that of the world economy. Clearly, this is inconsistent with the Fund’s Articles of Agreement.

Unfortunately, the current system does not ensure that the supply of liquidity and its distribution will respond to global needs. The IMF has not played the central role entrusted to it in the process of creating and distributing international liquidity. Despite acceptance of the objective of making the SDR the principal reserve asset, SDRs have declined as a proportion of world nongold reserves, from 5.8 percent at end-1975 to 2.3 percent at end-1995.

For a large number of countries insufficient liquidity requires the adoption of restrictive economic policies. Moreover, the effects are also felt by other countries and by the world economy as a whole. For instance, the liquidity shortage of the 1980s led to severe economic stagnation or decline in many of the Latin American economies. Some analysts have identified this as one of the factors behind the stagnation of demand growth for real U.S. exports between 1980 and late 1986, and thus, an important contributing factor to the deterioration of the U.S. external position during this period. On the other hand, the period of abundant liquidity during the early 1990s, which contributed to the resurgence of economic activity in many Latin American countries, made an important contribution to world export growth.

Volatility of Capital Flows

The process of financial innovation and the deregulation and liberalization of financial markets throughout the world have led to more integrated and globalized international capital markets. In addition, technological advances have increased the speed at which international financial transactions take place. As a result, the amount of capital traded in international markets has surged in recent years. For instance, in 1994 the value of international equity issues was almost 12 times larger than in 1990, while issues of bonds more than doubled in the same period.10” The volume of transactions in global foreign exchange markets has grown to over $1.2 trillion a day,11 which is equivalent to the value of world trade in goods and services in a whole quarter.

By providing expanded opportunities to both borrowers and lenders, this process is clearly positive for the world economy. However, financial markets have shown a tendency toward great swings. Thus, there are periods of overshooting and, therefore, an unwarranted expansion of liquidity, followed by periods of undershooting or inadequate expansion or even contraction of liquidity for a country or group of countries. This poses new and difficult challenges for the financial authorities and the international community.

Massive capital inflows can complicate the conduct of economic policy because, regardless of the exchange regime, the currency tends to appreciate in real terms. Thus, financial authorities face the dilemma of sterilizing the incoming capital at the cost of holding reserves that yield a lower return than domestic instruments and inducing high interest rates, or allowing an excessive expansion of the monetary base that could translate into increased expenditure and inflationary pressures. Moreover, when portfolio investments account for an important portion of the funds, the risk of a sudden reversion of capital flows that could force the country to endure a painful adjustment is considerable.

Capital flows to emerging economies are often volatile for reasons that bear little relation to country risk:

  • Exogenous and unanticipated changes in the financial conditions in the industrial countries can produce severe destabilizing effects in capital importing countries that are unrelated to their creditworthiness. Industrial countries generally formulate their economic policies based solely on domestic considerations, with little regard for the international repercussions of their actions. On several occasions, this has led to higher levels of real interest rates and exchange rate fluctuations that have increased the cost and diminished the availability of international financing to developing countries (Calvo, 1993).

  • Capital inflows and commercial bank lending have been markedly pro-cyclical. Capital comes most readily when economic and business prospects are good, whereas capital flows tend to decline when the economy slows or when problems or uncertainties of any kind appear. Thus, capital markets themselves tend to undermine the creditworthiness of countries. As the Bank for International Settlements (1983, p. 130) stated some time ago:

    To put this view into perspective, it may be useful to imagine what would happen in a national context if during a recession banks were suddenly to cut off the flow of new credits to the corporate sector and to begin closing off existing short-term credit lines. The inevitable result would be a financial collapse which would threaten to engulf even soundly managed firms, including banks…. Such a financial collapse… would therefore not permit any easy inferences with respect to the quality of the pattern of bank lending and of corporate investment before the outbreak of the crisis, whereas the conclusion could safely be drawn that something had gone seriously wrong with the macro-economic management of the economy.

  • Market behavior is often characterized by information asymmetries and contagion effects. Country risk analysis, which is far from perfect,12 is often dominated by “herding” behavior. Recent episodes of financial market turbulence illustrate that a country might lose its creditworthiness overnight, leaving authorities little time to react. In some cases, the sudden loss of creditworthiness may be unjustified. Countries may face abrupt changes in the cost and availability of liquidity when unexpected events, such as major macroeconomic or financial shocks, trigger sudden shifts in market sentiment and create a strong need for additional reserves. As the market liquidity dries up, the country may face excessive adjustment costs. Moreover, there is a tendency to evaluate a country’s creditworthiness as a function of its regional location or stage of development rather than on its own merits. This is particularly true for developing countries. When a country in a certain group experiences payments difficulties, markets often tend to suspend new credits to all countries in the same region or with similar characteristics.

  • In many cases, the process of restoring creditworthiness is unduly delayed. Investors hold back and wait for the recovery of the country in question, thus making it more protracted and uncertain. The official agencies regulating banks and extending official credit, as well as the rating agencies, also play an important role, but being cautious they may wait to see compliance with the program for a period of, say, a year or more before revising their appraisal.

The bandwagon effects, which abruptly reduce liquidity across the board, produce harmful effects for recipient countries and have a destabilizing impact on the international monetary system. In view of the considerable risk associated with the volatility of capital flows, recipient countries should not have to rely too heavily on these flows to satisfy their demand for reserves.

Summary of Shortcomings

Asymmetries in access to international liquidity aggravate the asymmetries in the adjustment process that cause deficit countries to bear an excessive share of its cost, while surplus countries have little pressure to correct their surplus and some of them enjoy the benefits of international seigniorage. Moreover, most adjustment programs appear to be underfinanced and consequently rely excessively on restrictive policies with adverse effects on output, employment, investment, and growth. This situation leads to a high proportion of program failures and is inconsistent with the objectives of the Fund as defined in Article I.

The current arrangements for the provision and distribution of international liquidity are inefficient and inequitable. The limited access of most developing countries to international reserves and financing has introduced a deflationary bias in the world economy. The volatility and uncertainties that countries face in satisfying their demand for reserves through the private markets impose heavy costs on them. Indeed, for developing nations, short-term capital inflows represent a mixed blessing as they pose difficult policy dilemmas that are aggravated by the risk of sudden outflows in response to factors beyond the control of their financial authorities.

The shortcomings of the current international monetary arrangements require increasing the availability of international liquidity and improving the way in which it is distributed to ameliorate the workings of the adjustment process and allow countries access to the liquidity needed to exploit their growth potential.

Proposed Alternative Arrangement

Improving Access to Reserves

The reserve needs of most countries are not met by present sources of supply. The existence of a stable relation between the demand for international reserves and the growth in world production and trade13 makes it possible to estimate the world economy’s liquidity requirements. However, at present no international organization evaluates the liquidity requirements of the different components of the global economy. Under its Articles of Agreement, the IMF should assume responsibility for this task as one of its central duties.

Despite the objectives endorsed by the international community since Bretton Woods and, in particular, by the First Amendment to the Articles of Agreement of the IMF, which took effect in 1969, and authorized the Fund—in Article XXI, Section 1—to allocate SDRs “to meet the need as and when it arises, for a supplement to existing reserve assets,” the growth of international liquidity is largely dependent on the monetary policies and payments disequilibria of the world’s major economies, which continue to be the primary source of international liquidity. Furthermore, ample access to secondary sources of international liquidity is confined largely to the industrial countries and to a few developing countries. As the Managing Director has emphasized on many occasions, most other countries have insufficient access to liquidity and must limit their imports and levels of economic activity to avert payments crises.

The failure by the IMF to use the SDR allocation mechanism in a regular fashion, as envisaged by its creators, may raise doubts about the meaning of this legal power. Since an 85 percent majority is required for an allocation, a few reserve currency countries have blocked all new initiatives for general allocations. Therefore, the deficiencies observed in creating and distributing international reserves can be attributed largely to the fact that the Fund has been unable to comply with its mandate. Hence, under current arrangements, it would be a pure coincidence if the supply of international liquidity were to meet the requirements of the various country groups, particularly those of the developing countries.

Just as the central bank of each country estimates the liquidity requirements of its economy every year, using certain assumptions about economic growth and price levels, so the IMF should conduct an annual exercise to assess the liquidity requirements of the world economy, the probable expansion of liquidity, and the likely access of countries or groups of countries to the various sources of international liquidity.14

The goal of liquidity creation by the Fund would be to help achieve a sustainable rate of growth of the international economy that is consistent with price stability. International reserves could be seen as the equivalent of the international monetary base. The latter is made up of reserve currencies and SDRs in central bank reserves and of reserve positions in the Fund. The primary operating variable is the level of reserves, including the SDRs that banks hold. Central banks’ demand for reserves is determined by precautionary liquidity requirements as well as by operational factors, notably the need for transaction balances in international payments, which are a function of growth and trade.

The Fund’s liquidity operations would comprise: (1) creation of international reserves; (2) financing in support of adjustment programs; and (3) exceptional lending. Reserve creation would be conducted through SDR allocations as outlined below and would constitute the Fund’s chief instrument to meet the liquidity needs of the international economy in compliance with the First Amendment. Financing in support of adjustment programs would strive to promote growth-oriented adjustment. In addition, the Fund would maintain an emergency lending facility: a type of Lombard loan, that is, swaps of domestic currency against SDRs or, preferably, advances against collateralized securities. These Lombard loans would be used by the central banks only to overcome unexpected liquidity shortages, would be accessible for amounts up to an individual country limit, and would have to be repaid in a space of up to six months. A country unable to repay would commit itself to undertake an adjustment program.

Proposal for Reserve Creation by the Fund

At present, SDRs are allocated in proportion to quotas as required under the Articles of Agreement. Thus, paradoxically, the main industrial countries, which have greater access to international liquidity, receive the bulk of the allocations as they command the largest quotas, whereas low-income countries, which have the greatest need to supplement their reserves, receive only a small proportion. On repeated occasions, there has been a call on industrial countries to redistribute their share of the SDRs to developing countries in order to improve the distribution of international liquidity among groups of countries. However, these proposals have not been implemented in spite of the support of France and Japan.15

A more efficient and more equitable arrangement would be to allocate SDRs following an agreed procedure in accordance with the reserve needs of individual countries, as determined by the Fund. For instance, the Fund could calculate the equilibrium long-run demand for reserves for each country. Reserve demand16 would be estimated according to the traditional explanatory variables, such as real output, the average propensity to import, and the variability of international transactions. Allocations would be made on a yearly basis to all eligible countries. The amount to be allocated to each country would be proportional to the calculated flow demand for reserves for the year. As under the current system, each country would decide if it accepted the amount allocated. Therefore, countries with market access could determine what part of their flow demand for reserves they desired to satisfy through SDR allocations and what part through other arrangements or through the market.

An alternative procedure for allocating SDRs in accordance with reserve needs could recognize that countries have different degrees of access to international capital markets and that, consequently, the cost of holding and acquiring reserves differs greatly among countries.17 A schedule for allocations could be agreed, in which the countries with market access would receive only a proportion of their flow demand, while those countries with limited or no market access would receive a higher proportion of their flow demand. Since market access and a country’s per capita GDP appear to be closely correlated, a schedule could be constructed by which the proportion of the reserve needs to be satisfied by SDR allocation is a function of per capita GDP. For instance, SDRs could be allocated according to a schedule as indicated in Table 2.

Table 2.

Illustrative Schedule for SDR Allocations

article image

As can be seen in Table 2, under this (alternative) scheme, industrial countries that can readily obtain liquidity in the international markets receive only a small proportion of their reserve needs through an SDR allocation. Developing countries receive a greater proportion of their reserve needs through an allocation to compensate for their lack of access to international capital markets. No country receives the full amount of its reserve needs. Presumably, the credit-enhancing effects of the allocation would facilitate a country’s access to market sources of liquidity.

At the time of the first allocation, countries may face an initial deviation between the actual amount of their reserves and the calculated equilibrium long-run level. A formula could be devised to close the gap gradually between actual and desired stock of reserves.18 The speed of adjustment could take into account the situation of world liquidity as a whole. However, for simplicity, this aspect can be left open for further consideration.

SDR allocations could be made at the beginning of each calendar year. SDRs would be allocated to all those countries that were classified as eligible during the previous year. Alternatively, SDRs could be allocated to a country on the day of its Article IV consultation if it met certain eligibility requirements.

To be eligible to receive the regular allocations, a country would have to meet a predetermined set of criteria relating to fiscal balance, external position, etc., which would indicate that it is following sound policies. A set of criteria of the Maastricht type could be devised as benchmarks for eligibility. Those countries that met the criteria would automatically qualify for the allocation. Countries that did not meet the criteria could exceptionally qualify subject to an evaluation by the Fund of their policies and of the progress made toward reaching the benchmark criteria.

Every year, in the context of the Article IV consultation, the Fund staff would make an appraisal of the country’s economic policies based on the above-mentioned indicators, as well as other qualitative analysis. Through this process, the Executive Board, on the recommendation of the staff, would determine whether the country is eligible to receive its SDR allocation or would lay out the necessary steps for it to become eligible.19 This exercise would certainly give an important boost to the surveillance function of the IMF. Countries with limited or no market access would have the greatest incentive to become eligible, since the availability of liquidity coupled with the cost saving involved in obtaining liquidity at the SDR rate would produce the largest benefits for them. For industrial countries, SDR allocations would be an interesting alternative to borrowing in international markets, particularly during episodes of financial market turbulence and for countries that have a high debt-to-GDP ratio.

As under the current system, countries would pay charges for the net use of the SDRs allocated to them, while countries receiving the SDRs in exchange for their currency in excess of their cumulative allocations would be paid interest. The rate of charge for using the allocated SDRs would be somewhat higher than the interest rate paid on the excess over the SDRs allocated. This spread between the user rate and the holding rate would provide the Fund with an extra source of income that could be applied to constitute an administered account to be used if a country went into arrears in the SDR account. Once reserves had been built up to a sufficient level, the additional income could be used to subsidize IMF lending to the poorest countries under its concessional facility.

This system would not give rise to a resource transfer from net holders to net users of SDRs. As is well known, SDRs are costless to create. A country that spends the SDRs on real resources would enjoy a higher level of real resource use than otherwise, but this would be financed by the interest payments that are being made on the net use of SDRs. Since the interest rate on the use of SDRs would be a market rate, in effect, the Fund would be extending credit to the net users of SDRs and would be paying net holders of SDRs interest for their willingness to exchange their currency for SDRs. Countries that simply need to boost their reserves could do so at no cost, since they would just hold the allocated SDRs.

It can be argued that insofar as the interest rate on the net use of SDRs is below a member’s cost of borrowing in international capital markets, a resource transfer is taking place. In effect, this is correct. The resource gain from SDR allocations is derived from the Fund’s capacity to create international liquidity and of ensuring the payment of interest by those members that use the SDRs.

In essence, the role of the Fund would be to provide countries with adequate liquidity by enhancing the “creditworthiness” of those that do not have ready access to international capital markets. Thus, the IMF, acting as a financial intermediary owned by all the world’s governments, would fulfill a valuable risk-reducing role. Through its surveillance of a country’s economic policies, through its analysis of the world economy, and through its preferred creditor status, the IMF is in a position to minimize effectively the risk of default of countries receiving SDRs.

By supplying liquidity through allocations to countries following appropriate policies, the IMF would, in addition to supplying resources, endorse the policies followed. In so doing, it would contribute to reducing risk in financial markets. This should help reduce the tendency of the international capital market to overreact and to underlend to developing countries. An important part of the membership would benefit from a significant narrowing of interest rate margins, particularly countries with limited or no market access.

Experience under the current system has shown that countries at different times have willingly held SDRs in their reserves and been satisfied with the risk-adjusted return on their SDR holdings. This would continue to be so, since the credit risk would be shared by the entire IMF membership. Every country would be assured the payment of interest on their excess holdings of SDRs by the institution. Even under the current system, the SDR Department is required to pay interest to each holder of SDRs, whether or not sufficient amounts of SDRs are received in payment of charges, which virtually eliminates risk.

What incentives would countries have to maintain their SDR account in good standing? Would this mechanism give rise to moral hazard problems? Members place a high value on their relationship with the Fund. Even in the face of extreme balance of payments difficulties, members have endeavored to meet their obligations to the Fund and to the SDR Department. Under the proposed system, the nonfulfillment of their obligations in the SDR account would mean that the country is in arrears with the whole international community and thus ineligible to receive allocations. Therefore, tremendous pressure would be applied to the country involved. In effect, a country in arrears would become isolated from the world’s trade and payments system, lose access to financial and trade credit, and lose any aid flows from development finance institutions and, more generally, from the international financial community. Given the effects of reneging on its obligations, domestic public opinion would also demand that policymakers take the necessary steps to become current in their payments and that the country become eligible for future allocations. Thus, the moral hazard risks involved in the proposal would be minimal.

What Would Be the Benefits Under the Proposal?

  • The direct benefits would accrue to member countries that have limited access to liquidity and a cost of acquiring reserves that is greater than the SDR interest rate. However, the main benefits are perhaps the indirect systemic benefits. The provision of adequate levels of international liquidity for all eligible countries would foster growth of trade and output, reduce the total credit risk in the global economic system, and provide real resource savings for countries collectively.

  • The SDRs allocated would provide reserves at a lower cost than they can be acquired for through borrowing or current account adjustment. Therefore, the average cost of obtaining reserves would decrease for the majority of Fund members. Moreover, the greater availability of international liquidity to countries following sound macroeconomic policies could lead per se to an improvement of a country’s credit standing and hence to a reduction in their marginal cost of borrowing.

  • If SDR allocations lead to a better balance of payments equilibrium and reduce the demand for foreign borrowing, they will reduce pressures on world savings and consequently on interest rates. The reduction in interest rates would stimulate investment and growth in the world economy. Major trading nations would benefit particularly.

  • The greater availability of reserves would increase the probability of success of adjustment programs, particularly as countries could avoid excessive balance of payments adjustment for purposes of accumulating reserves. Therefore, the adjustment process would work better as it would be less recessive.

  • The steady supply of SDRs would reduce the volatility of capital flows because inflows of capital to developing countries would not depend on the cyclical condition of the main industrial countries and would be less exposed to bandwagon and contagion effects. The increased stability in the flow of international trade and payments would allow capital-importing developing countries to pursue economic policies with a longer-term perspective.

Emergency Financing

Most countries that, despite fundamentally sound economic policies, experience short-term balance of payments or exchange market pressures do not have a readily available source of emergency financing.20 The rapid development of financial markets and their increasing importance, together with the high degree of volatility of capital flows in recent years, has called for renewed attention to the role that the IMF can play in assisting members facing unwarranted pressures in their financial markets. By providing a formal “safety net” to countries following sound economic policies, appropriate emergency financing would improve the functioning of the international monetary system. In particular, it would prevent countries from resorting to capital controls or exchange restrictions when confronting short-term financial market disturbances. Against this background, the IMF could establish an emergency financing facility. Emergency financing would be provided through a Lombard-type facility for short periods and in exceptional circumstances. The operation of this facility could be subject to the following principles.

Access and Credit Limits. In principle, access to the IMF’s Lombard facility is given to all member countries’ central banks, that is, banks desiring access are granted a credit limit that they must fully and permanently collateralize. In case of an unexpected liquidity shortage, banks may borrow in part or in whole up to the credit limit, with maturities ranging from a minimum of one month, up to three months, renewable once. The Fund, in turn, could reserve the right to reduce a country’s credit limits at any time upon appropriate notice.

Collateral Eligibility and Rules. The Fund could accept eligible foreign and domestic securities, discountable bills, and gold as collateral for emergency loans, namely, (1) securities issued by public entities; and (2) highly rated bonds issued by recognized institutions provided they are traded in organized markets with regular publication of quotations. The securities would be subject to lending limits that vary between 40 percent and 80 percent of their market value, with foreign securities generally given a higher valuation than domestic securities.

Interest Rate. Emergency loans would be a source of exceptional financing from which central banks could draw on their own initiative. Loans could not remain outstanding for more than a few months in any 12-month period. The loans could serve as a safety valve for banks to tide over unforeseen liquidity shortages, particularly those arising from sudden capital movements. Credit would be granted at the SDR Lombard rate, which would be some 2 or 3 percentage points above the SDR interest rate in order to discourage their use.


The current international monetary system has important shortcomings: liquidity creation is governed by the monetary policies and payments disequilibria of the world’s major economies and capital flows by the vagaries of the markets. To the well-known asymmetry resulting from the burden of adjustment falling on deficit countries, a further asymmetry is added, resulting from the inequitable distribution of the benefits of international liquidity creation and from an unequal access to international financing. As developing countries are forced to limit their imports and levels of economic activity in order to avert payment crises, these asymmetries introduce a deflationary bias in the world economy. Moreover, despite widespread recognition of the benefits of adjustment with growth, financing for adjustment programs is generally insufficient to make it feasible.

In spite of having the instruments at its disposal, the central institution of the international monetary system, the IMF, has not been allowed “to supplement existing reserve assets in such a manner as to promote the attainment of its purposes and to avoid economic stagnation and deflation as well as excess demand and inflation in the world.”21

Clearly, to improve the functioning of the international monetary system, alternative arrangements, like the ones outlined in this paper, are needed. Such arrangements would ameliorate the workings of the adjustment process, reduce the risks associated with the volatility of capital markets, provide real resource savings for countries, and foster growth of world trade and output.

An enhanced spirit of international cooperation is required to render the international monetary system more efficient and equitable. It calls for the recognition by all IMF members of the needs and concerns of the less privileged members that are striving to pursue market-oriented policies and of the benefits that would accrue to the world economy.

A Focused SDR Allocation Muhammad Yaqub, Azizali Mohammed, and Iqbal Zaidi

A notable feature of the international monetary system as it has evolved since the SDR system was established by the First Amendment of the Fund’s Articles in 1969 is the rapid expansion of international capital markets. This has significantly enlarged the capacity of national authorities to acquire reserves through external borrowing, thereby making international liquidity a much broader, albeit more amorphous, concept than international reserves. Instead of relying exclusively on officially owned reserve assets for financing balance of payments deficits, a nonreserve currency country can tap additional resources by borrowing on private capital markets. This structural change makes it more difficult to derive meaningful criteria for assessing the long-term global need for reserve supplementation.

Recent discussions on the subject have focused on issues of access to international capital markets and on whether larger availability of “owned” reserves relative to “borrowed” reserves would improve the performance of the international liquidity system. This paper adopts this approach and presents a case for allocation of SDRs to those countries that would benefit most from them: developing and transition countries that either lack access to, or have to pay a significant risk premium when seeking to tap international markets as sovereign borrowers. In the ensuing discussion they are designated as nonaccess or low-access countries.

The proposal in this paper for a focused SDR allocation is quite distinct from the link proposals that would use the SDR system for transferring resources to developing countries. Unlike the link proposals in which the central element for the transfer of resources was the subsidy on the SDR interest rate, the case for the focused SDR allocation is made on the basis of a competitive SDR interest rate and how the SDR system helps to tackle the problem of maldistribution of access to international liquidity but without calling for a transfer of resources.1 The focused SDR allocation proposal outlined here has some common elements with the proposal in Williamson (1994), which would have the IMF issue SDRs if a substantial block of countries were unable to borrow internationally at an interest rate close to the SDR interest rate. Williamson had suggested that all member countries would receive SDRs on the scale needed to satisfy the revealed reserve needs of the less creditworthy countries, but a focused SDR allocation would be confined to developing and transition economies with large reserve needs but with little or no access to market borrowing.

The first part of the paper argues the case for the constructive role that annual, relatively modest SDR allocations to these countries can play in strengthening the international liquidity system, by reducing the high cost of reserve supplementation faced by many countries and by promoting growth-oriented adjustment in countries that have no or low access. The second part provides illustrative orders of magnitude for annual allocations to such countries, which stay well within the bounds of the projected growth in the demand for international reserves, but might prove sufficient to help members maintain adequate reserve levels relative to their projected needs.

Maldistribution of Access to International Liquidity

Issues of “fairness” have featured prominently in historical debates on international liquidity. In the first stage, it was the “advantage” presumably enjoyed in the earlier postwar decades by the United States because of its reserve country status that enabled it to finance its balance of payments deficits by issuing foreign liabilities denominated in its own currency instead of having to draw down its foreign assets. This advantage has since been extended to a few other countries, notably Germany and Japan, as their currencies have been increasingly accepted—or permitted by their national authorities to be accepted—as reserve assets. The distinction of being a reserve currency has been further eroded by the growth of international capital markets that has enabled many more countries to borrow against their own IOUs, although for most such borrowers their lack of reserve currency status has meant that they could not denominate their obligations in their own currency but had to denominate them in one or another reserve currency. The liberalization of capital movements and the floating of major international currencies have allowed reserve currency countries to argue that they now enjoy no special advantage. Indeed, they might even be subject to a degree of competitive disadvantage because of the demand for their currencies to be held as reserve assets by other countries, and to higher volatility in the exchange markets for their currencies because of the actions of such holders in switching from one reserve currency to another. In any event, all such holdings by foreigners must be treated as voluntarily held at prevailing exchange and interest rates.

It has been argued by some observers that if countries accumulate reserves efficiently, the costs of acquiring reserves through different channels (for example, borrowing or current account adjustment) would be equated at the margin. As a result, there would not be a significant differential in the costs of acquiring reserves, whether those costs were to come from the external borrowing (if the country holds borrowed reserves) or in terms of the costs of domestic absorption forgone to generate reserves through current account improvements. This would suggest that the cost of external borrowing can be taken as an approximation of the cost of acquiring reserves in other ways. This would be true only if all countries were able to optimize at the margin, but many countries with no or low access to capital markets are not in a position to equate their external borrowing costs with current account adjustment costs because they face credit rationing in international capital markets. There are crucial differences between countries that have access to capital markets and those that do not, and, among the former, between sovereign borrowers that can acquire reserves in such markets at little differential cost, relative to the return on reserve assets, and those that pay a significant premium for such borrowings.

High Differential Cost Countries

We look first at the case of countries that do have access but pay high net carrying costs for borrowing. This cost can be measured by the difference between the lending rate and the market yield on reserve assets. For syndicated lending, the differential cost of borrowing is represented by the spread over LIBOR2 or the U.S. prime rate. Although the interest rate fluctuates over the duration of the loan, the spread is usually fixed; it depends on the perceived risk associated with lending to a particular country as well as on the degree of liquidity in international capital markets. In the post-1982 “debt crisis” decade, syndicated lending to developing countries either ceased or remained confined to countries in East and Southeast Asia and a very few elsewhere that had somehow escaped the debt “contagion.” Even after its resumption in the early 1990s, syndicated lending has largely related to project finance for private enterprises or public sector entities borrowing for telecommunications and other infrastructure projects with identifiable returns.

An alternative measure is the differential cost in terms of fixed-interest sovereign bond obligations relative to the return on short-term U.S. Treasury paper. It is evident from available data that the net interest cost of holding reserves borrowed from private international markets is close to zero for most industrial countries and below 100 basis points for a very few others. The weighted cost of borrowing, as measured by the yield spread at launch for unenhanced bond issues by developing country sovereign borrowers, has varied between 200 and 300 basis points in the 1991–95 period, but with a spike to 450 points in the second quarter of 1994, presumably reflecting the aftermath of the Mexican crisis.3 Note that the yield spread measures the difference between the bond yield at issue for sovereign borrowers in developing and selected transition countries that have access to international capital markets and the prevailing yield for industrial country government bonds in the same currency and of comparable maturity.

When private markets are not willing to lend spontaneously to a country, a representative interest rate for the cost of external borrowing does not exist for that country. An alternative measure of the cost of holding reserves can be obtained by adjusting the U.S. Treasury bill rate, with the implicit yield evident in the international secondary market price for external debt.4 These implicit yield-to-maturity calculations for the cost of external borrowing indicate substantially higher costs for borrowed international reserves than those indicated by the interest rates on syndicated bank loans. Moreover, the implicit yield evident in the international secondary market price for external debt does not fully represent the cost of external borrowing for the developing country because it is derived under conditions of credit rationing.

Since the debt crisis of the early 1980s, commercial bank lending to countries with no or low access to capital markets has been concentrated in relatively secure, transactions-based businesses, such as trade and project financing, and banks have increasingly resorted to so-called structured finance (for example, collateralization) to shift the financing risk to borrowers. One technique to mitigate transfer and exchange rate risks has been to establish escrow accounts and to use future export receivables as collateral for commercial loans. However, these techniques have negative macroeconomic implications if pursued too vigorously or if risk-sharing attributes are skewed too heavily against the borrowing country. Furthermore, the earmarking of foreign exchange revenues for particular creditors reduces the country’s flexibility in responding to a balance of payments crisis. Finally, such contracts may have large upfront fees and they may set the price of receivables for countries with no or low access to capital markets at significantly below market levels (Collyns and others, 1993, p. 44). All this suggests that the interest rate on a commercial bank loan does not fully reflect the cost of holding reserves when developing countries borrow from international capital markets.

Beyond the substantial differentials, there is the problem that private capital markets have not been a reliable source of liquidity for most nonindustrial countries. Their access can be curtailed abruptly, often for reasons that have little to do with the policies pursued by individual countries but owing to “contagion” effects of developments elsewhere in the developing world or to “spillover” effects resulting from interest rate or other developments in the industrial countries.

Non-Access Countries

Even countries that have access at certain times are lumped together, in times of disturbance, with the much larger number of others that do not have access to portfolio capital, especially bond financing. Their reserve supplementation requirements can only be met by adjusting current account positions to expected capital inflows through reducing deficits or increasing surpluses. But such adjustments require reducing domestic absorption in order to expand net exports, typically achieved through compressing imports. Although it is difficult to quantify precisely the costs of forgone domestic absorption in order to generate any required current account change, several factors suggest why those costs should be even higher than the high differential costs that countries with access pay for their borrowed reserves.

The costs of import compression can be seen in the striking divergence between the average growth performance in the 1980s of the countries that encountered debt-servicing problems and others. These two groups had broadly similar growth records during 1968–80, which suggests that much of the sharp decline in output growth of the debt-distressed countries was related to the sharp curtailment of capital inflows. Moreover, since private capital flows had declined much more sharply than official capital flows, the average rate of economic growth of the market borrowers fell further between 1973–80 and 1981–87 than did the growth rate of official borrowers.

Many non-access or low-access countries are heavily dependent on imports of capital goods and intermediate products as inputs into production. External shocks that force countries to restrict imports of capital goods directly constrain investment. The strength of the link between import availability and output depends on the severity of the foreign exchange constraint, the relative openness of the economy, the degree of substitutability between imported inputs and domestic alternatives, and the extent to which a country’s policies permit shifts in relative prices to allocate available imports to uses with the highest opportunity costs.

Changes in the capacity to import will tend to have the largest impact on domestic output in the short run, when substitution possibilities are lowest. They will also have the largest impact on countries that do not possess a significant domestic capital goods industry and therefore have fewer possibilities of substitution between domestic and foreign goods in capital formation. When faced with severe foreign exchange constraints, the volume and speed of implementation of investment is slowed, at great cost to long-run economic performance. Marquez (1985) estimated the long-run elasticity of investment with respect to the purchasing power of exports (in terms of the imports of capital goods) at around 1.25 for the group of non-oil developing countries.5

Reducing Costs of Reserve Accumulation Through SDR Allocation

The costs of reserve accumulation for countries with no or low access range from those incurred through current account adjustment for countries lacking access, to the differential costs between borrowing rates and the rate of return on reserve assets for those with limited access. These costs have nevertheless to be incurred because of the benefits of holding reserves and the need to build them up in line with the growth of trade and other external financial transactions: the alternative of not doing so would be for the country to accept greater vulnerability to abrupt changes in the availability of external funding and having to undertake rapid import compression. However, these costs are saved if reserve supplementation takes place through the SDR system: the SDR system does not have a wedge between the interest cost of external borrowing and the return on reserve assets—the rate of charge levied against a country’s cumulative allocation is equal to the rate of interest paid on a country’s holding of SDRs.

It has been argued that the cost saving associated with SDR allocation is more apparent than real; it seemingly arises because the analysis understates a transfer of resources from countries with unimpeded market access to countries with no or low access (Corden, 1983). The transfer argument is based on the assumption that the wedge between lending rates and the return on reserve assets represents an entirely appropriate pricing of risk by markets and that the absence of the wedge in the SDR system therefore constitutes an implicit tax on creditors in the system. While recognizing the existence of the risk premium in private capital markets, it should be noted that many factors can give rise to a wedge between lending rates and returns on riskless assets, including market segmentation, cartel behavior by lenders or investors, and inefficiencies in financial market intermediation.6 In particular, in an international context, it is often very costly to obtain the necessary information to monitor closely a borrower’s investment activities, and this asymmetric information problem is made more serious by different national systems pertaining to accounting standards and disclosure requirements (Goldstein, Mathieson, and Lane, 1991, p. 24). The information problem increases the uncertainty that the financial market participants face and can lead lenders to demand a large risk premium from other than the primary borrowers. One of the conditions for the correct functioning of markets is perfect information about the goods and services exchanged. In situations where contracting parties have different information about some characteristics of the contract (for example, the quality of the product), there is “hidden information” and scope for adverse selection. On the other hand, when the possible outcomes of the contract depend on the actions of one of the contracting parties, which are not perfectly observable by the other party, there is “hidden action” and scope for moral hazard. Polak (1988, p. 180) has argued that “this is a case where the collectivity of members of the IMF, operating as a credit collective, can within certain limits provide capital more cheaply than the market could.” The resource gains of countries with relatively high differential costs of borrowed reserves are not at the expense of other participants.

The SDR system has been compared to the insurance principle of pooling risk and charging a premium based on the average risk of the pool. However, the adverse selection and moral hazard risks that insurance markets have to deal with are minimized in the SDR system. There is no adverse selection because almost all countries are members of the Fund; the moral hazard is minimized by the knowledge that a country that fails to meet its obligations would have to pay significant costs, not least of which would be exclusion from additional SDR allocations or even from access to conditional lending from the Fund.7

SDR Allocations and Conditional Lending

It has been argued that for countries in need of macroeconomic adjustment, an SDR allocation would be counterproductive because it would weaken their incentives to pursue appropriate policies. This is partly a question of the magnitude of allocations made on an annual basis and will be so argued in the following section. To overcome this argument against SDR allocations, several proposals have been made to link SDR allocations with adjustment programs supported by the use of Fund resources or to avoid excessive use of SDRs by reintroducing a reconstitution requirement (see, for example, Sengupta, 1987). In many respects, however, the issue of whether there are sufficient resources available for conditional lending is quite separate from the question of an SDR allocation. Questions regarding the adequacy of conditional lending should focus on such issues as whether the balance of payments deficits were being caused by permanent rather than by temporary shocks, whether certain policy measures that tended to promote efficiency in the economy had led in the first instance to a deterioration of the balance of payments and thus to a need for temporary financing, and the like. In contrast, questions on SDR allocations should focus on the costs of building up reserves through borrowing or import compression, and on how SDR allocations raise the proportion of owned reserves and thereby reduce the vulnerability to financial market disturbances of countries with no or low access to capital markets.

A more fundamental question is how inadequacy of “owned” reserves can undermine adjustment efforts. A country with adequate holdings of reserves and access to other sources of international liquidity can sustain a smoother adjustment process because it can finance its way out of temporary external shocks. However, where there is inadequate liquidity, even temporary shocks will force a country off its adjustment path. This will apply even to those with past access to markets if such shocks result in curtailing access at a time when it is most needed. Even when access is not impeded, the cost of such funding can be volatile for marginal borrowers, depending on conditions and policies of major industrial countries. SDR allocations, being “owned” reserves, can promote steadiness of adjustment policies in the face of market volatility.

Focusing SDR Allocations on Countries That Would Benefit

The case for SDR allocations on an unconditional basis has been argued in terms of their benefit for countries without access or with limited access to international capital markets. The basic flaw in the current arrangements is that they mandate allocations for a majority of members that no longer have a need for them, being able to borrow reserves at little net cost. It is clearly suboptimal to have to create SDRs in a multiple of the amounts actually required for those that would really benefit from them. It furnishes a false argument about the inflationary effects of SDR allocation, an argument less credible in current circumstances than when first advanced. It sets up opposition in some industrial countries about the undermining of the appropriations authority of parliamentary bodies.

The purpose of the illustrative exercise is to indicate that the magnitude of such allocations would indeed be rather modest. One could start with countries judged as eligible for concessional funding by the international financial institutions. Most countries eligible only for IDA assistance or eligible for the enhanced structural adjustment facility (ESAF) would be included in any list of countries with no or low access, even if most of them have access to trade credits or export credits, guaranteed by official insuring agencies in the exporting country. Nor should the presence of an occasional corporate borrower or a marginal investor such as an emerging market mutual fund be considered as making them market borrowers. It is the typical lack of access to fixedinterest bond markets or even adjustable-rate syndicated markets as sovereign borrowers that defines their status as countries with no or low access.

However, this status should not define eligibility for SDR allocations under the scheme proposed here. The first criterion of exclusion could be based on ratings assigned by major rating agencies, such as Moody’s and Standard and Poor’s Investor Services. Countries receiving an investment grade from one or the other rating agency would be considered as having the capacity to borrow net reserves at low cost: on this basis, two of the largest quota holders in the developing world—China and India—would be excluded. Countries not rated by either agency would qualify for inclusion as would those given a “noninvestment grade.”

The second criterion for exclusion could be reserve levels relative to imports in a preceding (say) three-year period. The threshold for exclusion might be a level of reserves exceeding (say) 50 percent of imports in a recent period, on the reasoning that countries already holding reserves equivalent to six months worth of imports did not need additional support through SDR allocations. On the basis of this criterion as well, China and India would be excluded. In addition, it may be noted that when the structural adjustment facility (SAF) and ESAF were set up, China and India had agreed not to make use of resources from these facilities.

Assuming that SDR allocations could continue to use Fund quotas as the key for distribution, one has a group with aggregate quotas of SDR 23.2 billion on the basis of applying the two exclusion criteria (Table 1). This represents an outer limit, however, as the scheme allows countries to opt out if they do not wish to be included. This would apply to such countries as Mexico that might consider their current loss of investment grade to be a temporary aberration or to some of the oil producers in the Gulf and in the Caribbean that were not rated because of lack of interest on their part. Some other countries might decide not to participate in a focused SDR allocation if they apprehend that inclusion in the list of those eligible might be seen as negative from a market-access point of view. Finally, countries could be given the right of opting in during any year that they felt the need to exercise their eligibility.

Table 1.

Fund Quotas of Countries Eligible for Focused SDR Allocations

(In billions of SDRs)

article image

The second issue is the quantum of allocation to be made on an annual basis. Here there are two constraints to be considered: the first is the risk that the size of allocations undermines the Fund’s conditionality and the second is whether creditors of the SDR system might begin to consider their holdings to involve an implicit tax. It has been argued earlier that from the perspective of countries with no or low access, an unconditional SDR allocation has the potentiality of strengthening the adjustment effort of countries.8 However, some have argued otherwise and it is essential that the magnitude of allocations be set at a level so as not to be seen as interfering with the normal conditionality of Fund programs. Assume that annual allocations would not exceed (say) 20 percent of quota. Comparing this percentage with access limits for programs under the Fund’s ESAF and extended Fund facility currently in place, it would be hard to believe that countries would put the availability of such larger sums at risk; this is especially true for a number of countries whose access to bilateral assistance and debt relief—and even to other multilateral funding—is contingent upon their meeting performance criteria or benchmarks under the Fund’s conditional facilities.

There remains the question of whether creditors of the SDR system would be prepared to absorb annual allocations in the range of SDR 4.0–4.5 billion a year (assuming 20 percent of quota as the outer limit for annual allocation) without regard to acceptance limits currently established by the Articles. Three considerations have to be kept in mind. First, only when members receiving allocations actually use them would creditors in the SDR system come into the picture. Given the reserve supplementation function that is primary to the allocation process, one should expect most recipients to continue to hold their SDRs voluntarily; this would be especially true of countries that have obtained Fund assistance under its conditional facilities and must cover the exchange risk of repayment in SDRs from their reserve currency assets. Second, the two exclusion criteria and the opting out procedures earlier outlined would reduce the list of countries over time as their reserves build up, their access to private capital markets opens up, and their credit ratings improve; this would be particularly applicable to major transition countries like Russia and some of the countries in Central and Eastern Europe. Finally, there does appear to be a growing willingness on the part of a certain number of countries to acquire and hold SDRs in excess of their allocations; this is indicated by the virtual redundancy of the designation mechanism after 1987.9 The need to be concerned about acceptance constraints could be further reduced if the usability of SDRs were enhanced by enabling the major international commercial banks to qualify as “other holders” of SDRs, thereby making a market beyond the official circuits to which the SDR mechanism is presently confined.

Concluding Remarks

In making the case for an unconditional allocation of SDRs, this paper has focused on the point that if there are no SDR allocations, the countries with no or low access will have to incur substantial costs to acquire the additional reserves they need. Reserves that have been obtained by borrowing on international capital markets impose a significant net interest cost for countries below investment grade. Reserves that have been obtained through the compression of domestic demand and net imports impose a significant cost in terms of forgone consumption or investment. These costs exceed the true economic opportunity cost to the world of creating additional reserves through an SDR allocation. SDR allocations reduce the cost of holding reserves for countries with no or low access because SDRs have no net carrying cost if they are held as a part of reserves (that is, SDR charges are offset by SDR interest payments). Circumstances that would force some countries to use their SDR allocations temporarily would indicate a judgment that the net interest costs of such use are lower than the costs incurred by negative alternatives such as accumulating arrears, compressing imports, or taking other adjustment measures.

This paper has emphasized that SDR allocations improve the distribution of international liquidity and promote growth-oriented adjustment programs. The ability of a country to persevere in implementing an adjustment program in the face of shocks that adversely affect its balance of payments position depends on the terms and conditions under which it can obtain international liquidity. An analysis of the trends in net borrowing from private creditors by countries with no or low access shows that the volume as well as the price of private foreign capital inflows have varied considerably over time. However, when there is inadequate liquidity, even temporary external shocks can force the country to modify its adjustment policies. This choice may not be desirable when there is a strong adjustment program in place, because a sharp change in policies can needlessly depress investment spending and thereby limit growth.

The proposal put forward for a focused unconditional SDR allocation suggests that SDRs should be allocated to countries that have relatively large reserve needs but have little or no access to international capital markets. Since the proposal focuses on a group of countries that have a small share of Fund quotas, it helps to reduce any potential inflationary impact of SDR allocations. It is recognized that the Fund has no authority to make selective allocations, but, given the impasse concerning SDR allocations, it is time to look for ways beyond the current Articles of Agreement. It is hoped that the present proposal will be seen as a midway point between the divergent positions adopted by the great majority of the Fund membership on the one hand, and a few major creditor countries of the Fund on the other, and thus help to move the SDR discussion forward.

General Discussion

Horst Siebert, referring to the interconnection between international reserves, monetary policy, and exchange rate policy, raised the question of seigniorage, which, he suggested, was the reward for price stability. He asked Ariel Buira if newly industrializing countries could solve the seigniorage problem by setting up a currency board arrangement. The currency board would establish a currency’s reputation, and if the supporting monetary policy remained credible, the newly stable currency could draw seigniorage away from the other major currencies.

Ariel Buira responded that stability alone would not be sufficient to earn seigniorage from international usage of a currency. A large economy was also required, because such seigniorage could only be earned by a currency that was internationally accepted as a means of payment. However stable Guatemala may be, for example—and it was extremely stable for forty to fifty years—the quetzal was never quoted in international financial markets.

John Williamson noted that a currency board enabled the authorities to recapture some of the seigniorage derived from domestic money creation, but it was not a relevant mechanism for capturing the seigniorage that comes from international reserve creation. A currency board arrangement did not establish an independent currency that countries would want to hold as international reserves.

Helmut Hesse distinguished between two wholly separate problems that had been cited in discussing the roles of the IMF and the SDR in the international monetary system, and argued that these two problems should not be mixed together. If a country that lacked creditworthiness and, hence, access to private financial markets, experienced a liquidity crisis, it should receive help from the IMF—but that help should be in the form of conditional lending. However, if the issue was one of resource redistribution, then in his view the IMF had no role to play. Hesse said he favored the development of the IMF in the direction of the central bank of a financially integrated world—and central banks had nothing to do with redistribution.

Peter Kenen asked Iqbal Zaidi and his coauthors how their proposal differed from the “link” proposals of the 1960s and 1970s, under which SDRs would be allocated to low-income countries with weak reserve positions, and the industrial countries would in effect earn reserves through sales of goods and services to the developing world.

Kenen also said he had found some internal inconsistencies or tensions between the two parts of Buira’s proposal. An emergency financing facility operating as a safety net implied a certain short-term elasticity to the SDR on a fairly large scale. If, for example, the action taken by the IMF and others in response to the Mexican peso crisis had been effected through an emergency facility, it would have generated a large increase in the stock of SDRs, at least in the very short run. Yet if, at the same time, the IMF moved toward being a fully SDR-based institution, as the second part of Buira’s proposal suggested, then some member countries might need reassurance that the IMF’s ability to create credit and to provide balance of payments financing would not be unlimited in the aggregate.

Iqbal Zaidi said the link proposals of the 1960s and 1970s had been intended to “kill two birds with one stone.” On the one hand, SDR allocations to all member countries would solve the liquidity and deflation problems embodied in the Triffin dilemma. On the other hand, since the SDR interest rate was below market interest rates at that time, the allocation would act as an aid mechanism, providing a real transfer of resources to reserves-starved developing countries whose severe budgetary constraints effectively stifled growth and adjustment. Today, however, the SDR interest rate was no longer concessional, and selective allocations would not embody an aid mechanism or involve a transfer of real resources to poor countries since creditors to the SDR system receive market interest rates. Rather, selective allocations would be a means for these countries to acquire reserves without incurring the costs of import compression, and for the IMF to distribute liquidity more efficiently than the market.

Buira, responding to Kenen’s question, said the emergency financing mechanism he had proposed provided for repayment in the short term. But even a large loan that could not be repaid in the very short term—such as the $40 billion extended to Mexico—would not have a significant effect on the total stock of international reserves, which amounted to about $1.4 trillion. Moreover, it was unlikely that the major industrial countries would avail themselves of this facility, since the interest rates on the SDRs would be higher than the cost of borrowing in the market.

Jacques Polak mentioned another important distinction between selective allocations and the link proposal. Under the link proposal, an estimate would be made of the total amount of reserves that ought to be created, and this total amount would be fed into the system through the developing countries. Under the selective allocation proposal, only the reserve needs of the small number of eligible countries would be estimated and allocated directly to them.

Polak said both the paper by Buira and that by Yaqub, Mohammed, and Zaidi espoused the principle that since there was no longer a point in across-the-board allocations to all IMF members, allocations should go only to the countries that would benefit from them. He noted with approval that the Yaqub-Mohammed-Zaidi paper further emphasized that allocations should add up to a small proportion of quotas. This principle, he said, reconciled the conflict inherent in the IMF’s being in the business of dispensing both conditional and unconditional credit—a conflict that could only be resolved by making the availability of unconditional credit much smaller than conditional credit in terms of quota.

Polak concluded by asking Zaidi and his coauthors which countries they had in mind to receive the selective allocations and how long eligibility would last.

Azizali Mohammed said that countries that could borrow in capital markets or could borrow government bond paper at a very low cost would not be eligible to receive an SDR allocation under their proposal. Based on these exclusion criteria, two categories of countries would be eligible: many, but not all, of the countries eligible for ESAF (the IMF’s concessional financing facility) or IDA (International Development Association, the World Bank’s concessional lending facility), would receive allocations; and all of the transition economies.

In response to Polak’s question about how long a country could continue to receive allocations, Mohammed referred to two provisions that would limit eligibility. First, under an “opting out” provision, a country with a temporary liquidity problem that prevented it from borrowing at low cost from the market could nevertheless decline to accept an SDR allocation because of the possibility that the market might treat as negative the country’s willingness to accept an allocation.

Second, the scheme would not give rise to an ever-increasing injection of SDRs into the system, because a country that was able to accumulate reserves equivalent to 50 percent of annual imports over a three-year period would no longer be eligible for additional allocations. Thus, after an initial period of allocations large enough to build up and sustain a country’s reserves, the size of allocations would gradually come to equal only the amount consistent with the rate of growth of imports. As countries became able to accumulate reserves, improve their credit ratings, and obtain investment-grade ratings in capital markets that permitted them to borrow at low net cost, they would lose their eligibility for SDR allocations.

Michael Mussa said it was not clear from either the Buira proposal or the Yaqub-Mohammed-Zaidi proposal who would be issuing the SDRs and against whom the holders of those SDRs would have a claim. One option (suggested by Polak) would be for the IMF to issue SDRs, and for recipients of SDRs to hold claims against the IMF, including all of its assets. The IMF’s General Resources Account would thus assume the liability, thereby providing a guarantee of creditworthiness on the SDRs that had been issued. Another option would be to rely on the designation mechanism under the present system, which permitted a member holding SDRs to use its SDRs to obtain an equivalent amount of currency from another designated member. Pursuit of this option, however, presented a problem, since the allocations would go only to countries in need of reserves and would thus be heavily unbalanced. Because none of the creditor countries would be in the system, there would be no official agency to pick up the credit side of the transaction.

Buira noted that in his proposal the IMF assumed the liability and acted as a broker. In the event of market failure—for example, when a country suffered from an unjustified loss of creditworthiness—the IMF could enhance creditworthiness by attaching conditions to SDR allocations, such as the requirement that countries meet certain policy criteria to qualify. The moral hazard risk was minimal because a country that defaulted on its obligations would be in arrears to the entire international financial community and would face virtual “civil death.”

Mohammed indicated that he and his coauthors were also inclined to have the IMF assume the liability, as in the systems suggested by Polak.


  • Baker, James, Address to the Fortieth Annual Meeting of the Boards of Governors of the IMF and the World Bank, Seoul, Korea, October 1985, in Summary Proceedings (Washington: International Monetary Fund, 1985, pp. 508).

    • Search Google Scholar
    • Export Citation
  • Bank for International Settlements, Annual Report (Basle: BIS, 1983).

  • Buira, Ariel, “International Liquidity and the Needs of the World Economy,” in The International Monetary and Financial System: Developing-Country Perspectives, ed. by G.K. Helleiner (New York: St. Martin’s Press, 1996), pp. 5584.

    • Search Google Scholar
    • Export Citation
  • Buira, Ariel, Reflections on the International Monetary System, Essays in International Finance, No. 195 (Princeton, New Jersey: Princeton University, Department of Economics, International Finance Section, January 1995).

    • Search Google Scholar
    • Export Citation
  • Calvo, Guillermo A., Leonardo Leiderman, and Carmen M. Reinhart, “Capital Inflows and Real Exchange Rate Appreciation in Latin America: The Role of External Factors,” Staff Papers, International Monetary Fund, Vol. 40 (March 1993), pp. 10851.

    • Search Google Scholar
    • Export Citation
  • Calvo, Guillermo A., Leonardo Leiderman, and Carmen M. Reinhart, “The Management of Capital Flows: Domestic Policy and International Cooperation,” in The International Monetary and Financial System: Developing-Country Perspectives, ed.by G.K. Helleiner (New York: St. Martin’s Press, 1996), pp. 85112.

    • Search Google Scholar
    • Export Citation
  • Calvo, Guillermo A., Leonardo Leiderman, and Carmen M. Reinhart, and Enrique Mendoza, “Reflections on Mexico’s Balance of Payments Crisis: A Chronicle of a Death Foretold,” Paper presented at a Seminar on Speculative Attacks in the Era of the Global Economy: Theory, Evidence, and Policy Implications, University of Maryland at College Park, December 1995.

    • Search Google Scholar
    • Export Citation
  • Camdessus, Michel (1994a), Remarks to the Group of Ten on the SDR Allocation, April 24, 1994.

  • Camdessus, Michel (1994b), Address to the Institute for International Economics, Washington, D.C., June 7, 1994.

  • Camdessus, Michel “The IMF and the Challenges of Globalization: The Fund’s Evolving Approach to its Constant Mission: The Case of Mexico,” Address to the Zurich Economics Society, Zurich, Switzerland, November 14, 1995.

    • Search Google Scholar
    • Export Citation
  • de Larosière, Jacques, Opening Address to the Fortieth Annual Meeting of the Board of Governors of the IMF and the World Bank, Seoul, Korea, October 1985, in Summary Proceedings (Washington: International Monetary Fund, 1985), pp. 1933.

    • Search Google Scholar
    • Export Citation
  • Edwards, Sebastian, “El Fondo Monetario Internacional y los Países en Desarrollo: Una Evaluación Crítica,” El Trimestre Económico, Vol. 62 (July-September 1990), pp. 61163.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, World Economic Outlook, various issues.

  • International Monetary Fund, International Financial Statistics, various issues.

  • International Monetary Fund, International Financial Statistics, Communiqué of the Interim Committee of the Board of Governors of the International Monetary Fund, Press Release No. 86/10 (Washington: IMF, April 1986).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, International Financial Statistics, Communiqué of the Interim Committee of the Board of Governors of the International Monetary Fund, Press Release No, 87/28 (Washington: IMF, September 1987).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, International Financial Statistics, (1995a), Improving the International Monetary System: Constraints and Possibilities, IMF Occasional Paper No. 116 (Washington: International Monetary Fund, December 1994).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, International Financial Statistics, (1995b), Private Market Financing for Developing Countries, World Economic and Financial Surveys (Washington: International Monetary Fund, November 1995).

    • Search Google Scholar
    • Export Citation
  • Khan, Mohsin S., “The Macroeconomic Effects of Fund Supported Adjustment Programs,” Staff Papers, International Monetary Fund, Vol. 33 (June 1990), pp. 195231.

    • Search Google Scholar
    • Export Citation
  • Killick, Tony, “Continuity and Change in IMF Programme Design, 1982–1992,” Overseas Development Institute Working Paper No. 69 (London: Overseas Development Institute, November 1992).

    • Search Google Scholar
    • Export Citation
  • Killick, Tony, and Moazzam Malik, “Country Experience with IMF Programmes in the 1980s,” World Economy (September 1992), pp. 599632.

    • Search Google Scholar
    • Export Citation
  • Killick, Tony, and Moazzam Malik, and Marcus Manuel, “What Can We Know About the Effects of IMF Programmes?” World Economy, Vol. 15 (September 1992), pp. 57597.

    • Search Google Scholar
    • Export Citation
  • Landell-Mills, Joslin, “The Demand for International Reserves and Their Opportunity Cost,” Staff Papers, International Monetary Fund, Vol. 36 (September 1989), pp. 70832.

    • Search Google Scholar
    • Export Citation
  • Polak, Jacques J., “Thoughts on an International Monetary Fund Based Fully on the SDR,” IMF Pamphlet Series, No. 28 (Washington: International Monetary Fund, 1979).

    • Search Google Scholar
    • Export Citation


  • Buira, Ariel, Reflections on the International Monetary System, Essays in International Finance, No. 195 (Princeton, New Jersey: Princeton University, Department of Economics, International Finance Section, January 1995).

    • Search Google Scholar
    • Export Citation
  • Coats, Warren L., Jr., Reinhard W. Furstenberg, and Peter Isard, The SDR System and the Issues of Resource Transfers, Essays in International Finance, No. 180 (Princeton, New Jersey: Princeton University, Department of Economics, International Finance Section, 1990).

    • Search Google Scholar
    • Export Citation
  • Collyns, Charles, and others, Private Market Financing for Developing Countries, World Economic and Financial Surveys (Washington: International Monetary Fund, 1993).

    • Search Google Scholar
    • Export Citation
  • Cooper, Richard N., “International Liquidity and Balance of Payments Adjustment,” in International Reserves: Needs and Availability, by International Monetary Fund (Washington: International Monetary Fund, 1970), pp. 12545.

    • Search Google Scholar
    • Export Citation
  • Corden, W. M., “Is There an Important Role for an International Reserve Asset Such as the SDR?” in International Money and Credit: The Policy Roles, ed. by George M. von Furstenberg (Washington: International Monetary Fund, 1983)

    • Search Google Scholar
    • Export Citation
  • Goldstein, Morris, Donald J. Mathieson, and Timothy Lane, “Determinants and Systemic Consequences of International Capital Flows,” in Determinants and Systemic Consequences of International Capital Flows, IMF Occasional Paper No. 77 (Washington: International Monetary Fund, 1991), pp. 145.

    • Search Google Scholar
    • Export Citation
  • Leven, Ronald, and David L. Roberts, “Latin America’s Prospects for Recovery,” Federal Reserve Bank of New York, Quarterly Review, Vol. 8 (Autumn 1983), pp. 613.

    • Search Google Scholar
    • Export Citation
  • Marquez, Jaime, “Foreign Exchange Constraints and Growth Possibilities in the LDCs,” journal of Development Economics, Vol. 19 (September-October 1985), pp. 3957.

    • Search Google Scholar
    • Export Citation
  • Márquez, Javier, “Reserves, Liquidity, and the Developing Countries,” in International Reserves: Needs and Availability, by International Monetary Fund (Washington: International Monetary Fund, 1970), pp. 97111.

    • Search Google Scholar
    • Export Citation
  • Polak, Jacques J., “The Impasse Concerning the Role of the Special Drawing Right,” in The Quest for National and Global Stability, ed.by Wietze Eizenga, E. Frans Limburg, and Jacques J. Polak (Dordrecht: Kluwer, 1988), pp. 17589.

    • Search Google Scholar
    • Export Citation
  • Rhomberg, Rudolf R., “Failings of the SDR: Lessons from Three Decades,” in International Financial Policy: Essays in Honor of Jacques J. Polak, ed.by Jacob A. Frenkel and Morris Goldstein (Washington: International Monetary Fund, 1991), pp. 15069.

    • Search Google Scholar
    • Export Citation
  • Rojas-Suárez, Liliana, “Risk and Capital Flight in Developing Countries,” in Determinants and Systemic Consequences of International Capital Flows, IMF Occasional Paper No. 77 (Washington: International Monetary Fund, 1991), pp. 8392.

    • Search Google Scholar
    • Export Citation
  • Sengupta, Arjun, “The Allocation of Special Drawing Rights Linked to the Reserve Needs of Countries,” in Compendium of Selected Studies on International Monetary and Financial Issues for the Developing Countries, by UNCTAD (Geneva: United Nations, 1987), pp. 31129.

    • Search Google Scholar
    • Export Citation
  • Triffin, Robert, “The IMS (International Monetary System … or Scandal?) and the EMS (European Monetary System),” Banca Nazionale del Lavoro, Quarterly Review, No. 162 (September 1987), pp. 23961.

    • Search Google Scholar
    • Export Citation
  • Williamson, John, “The Rise and Fall of the Concept of International Liquidity,” in The International Monetary System, ed. by Peter B. Kenen, Francesco Papadla, and Fabrizio Saccomanni (Cambridge: Cambridge University Press, 1994), pp. 5364.

    • Search Google Scholar
    • Export Citation

Joseph Gold, Special Drawing Rights, IMF Pamphlet Series, No. 13 (Washington: International Monetary Fund, 1969), pp. 1–10.


Article XVIII, Section 1(a), of the IMF Articles of Agreement.


International Monetary Fund, Annual Report, 1969, p. 29.


Proposed Second Amendment to the Articles of Agreement (Washington: International Monetary Fund, 1976), pp. 70–1.


Group of Ten, Communiqué of Ministers and Governors and Report of Deputies, July 7, 1966, item 30, p. 7.


Ibid., item 46, p. 10.


This view seems to be shared by the IMF; see The Role of the SDR in the International Monetary System, IMF Occasional Paper No. 51 (Washington: International Monetary Fund, March 1987), pp. 7–8.


Ibid., p. 21.


Article VIII, Section 7; Article XXII.


Group of Ten, “The Functioning of the International Monetary System,” Report prepared by the Deputies of the Group of Ten, June 1, 1985, item 78, p. 39.


Committee of Twenty, international Monetary Reform: Documents of the Committee of Twenty (Washington: International Monetary Fund, 1974), p. 7–8.


Stefano Micossi and Fabrizio Saccomanni, “The Substitution Account: The Problem, the Techniques and the Politics,” Banca Nazionale del Lavoro, Quarterly Review, No. 137 (June 1981), pp. 171–89.


The views expressed in this paper are strictly personal.


Article I of the IMF’s Articles of Agreement.


This represents approximately 25 percent of total world imports for the year.


In comparison, note that in 1995 these countries carried out only about 30 percent of world trade.


The shares of the deutsche mark and the Japanese yen in total foreign exchange holdings increased sharply between 1975 and 1995. The deutsche mark’s share rose from 8.6 percent in 1975 to 12.6 percent in 1985 and to 14.1 percent in 1995. The yen’s share rose from 1.7 percent in 1975 to 6.9 percent in 1985 and to 7.3 percent in 1995.


See Article I of the Articles of Agreement of the IMF.


See Killick and Malik (1992) and Killick, Malik, and Manuel (1992).


Khan (1990) evaluated econometrically the results of IMF programs implemented in the period between 1973–88. He concluded that these were successful in reducing balance of payments disequilibria, and tended to reduce the inflation rate. However, he also concluded that the programs reduced the rate of growth of output. In light of these results, the hypothesis that IMF-sponsored programs are essentially deflationary cannot be rejected. Edwards (1990) arrived at similar results. Although macroeconomic crises are harmful to growth, a more recent study of eight country experiences with adjustment programs in 1974–94, prepared by the IMF’s Policy Development and Review Department (1995a), found that economic growth increased over the medium term in only three of the eight countries analyzed.


In his statement at the IMF’s Fortieth Annual Meetings in October 1985, Secretary Baker declared: “We need to build upon the current approach to strengthen its ability to foster growth.”

Moreover, in his opening address at the same meetings, Managing Director J. de Larosière stated: “Adequate financing is a second key requirement if adjustment is to be combined with growth. Commercial banks, official lenders, and multilateral institutions all have an important role to play in this connection. The debtor countries must grow out of debt How is this to be achieved? Fundamentally, it involves encouraging productive investment”

This new approach to adjustment programs has subsequently been confirmed. For instance, the Interim Committee of the Board of Governors of the IMF stated in its communiqué of April 10, 1986, that: “The Committee reaffirmed the central role of the Fund in providing its members assistance and finance to develop growth-oriented adjustment programs and in serving as a financial catalyst.”

One year later, the Communiqué of September 28, 1987, stated: “Members reaffirmed the central role that the Fund had to continue to play in helping indebted countries develop appropriate growth-oriented adjustment strategies and in mobilizing finance …. The forms and terms of … financing in turn need to reflect the economic situation and prospects of individual countries, with a view to supporting both normalization of their payments position and a return to more satisfactory rates of economic growth.”

All italics are mine.


Bond issues increased from $226.6 billion to $523.4 billion. Equity issues amounted to $8.2 billion in 1990 and $95.5 billion in 1994 (International Monetary Fund, 1995b).


See Camdessus (1995).


See Calvo and Mendoza (1995).


This result has been attained and confirmed by different authors. See estimates of demand for international reserves in IMF (1994).


Of course, this would require an amendment of Article XVIII, Section 2, of the Articles of Agreement.


The Mitterrand proposal: President Mitterrand of France, in a speech before the United Nations in September 1988, proposed, inter alia, a scheme that would provide for the use of newly allocated SDRs to industrial countries to guarantee interest payments on certain debt issued by developing countries. The Hashimoto proposal: in the early 1990s, the Japanese authorities developed a proposal for SDR allocation with postallocation redistribution, initially outlined by Governor Hashimoto.


In estimating the reserve needs of individual countries, Fund staff should take into account the different role that reserves play in different countries such as their role as a shock absorber and as a basis for confidence building in international markets. Nevertheless, since some of these qualitative factors are reflected in the traditional explanatory variables, the manner in which they could be incorporated could be an issue for further study.


Empirical work has shown that the reserve holdings of countries that also borrow on international capital markets are significantly affected by the cost of holding these assets. When a country’s reserve holdings are assessed in terms of the interest rates it pays on international borrowing, two conclusions emerge: international reserves can be costly to hold and they are vulnerable to changes in terms of international financial markets. When international interest rates rise, spreads increase, and, according to the results of this analysis, countries economize on reserves. When, in particular, the range of spreads expands, so that the less creditworthy countries face higher borrowing costs, these countries, which are shown by the analysis to be more responsive to international borrowing costs than others, will adjust reserve holdings more quickly. The result is that those economies with the greatest need for reserves economize more than others on the reserves when international financial markets are tight. See Landell-Mills (1989).


This would follow the lines of Article XVIII, Section 1(b).


This process can be considered a sort of credit rating exercise being conducted by the IMF.


Members of the European Union and a few other countries with a network of swaps are an exception. The existing emergency financing facility is essentially a fast-track procedure for considering requests for a stand-by arrangement or an arrangement under the extended Fund facility.


See Article 1 and Article XVIII, Section 1(a).


Coats, Furstenberg, and Isard (1990) demonstrate that if the SDR interest rate is competitive, then the net use of SDRs does not involve permanent resource transfers. They point out that, for many countries, SDR allocation provides resources at terms more favorable than the costs of borrowing or earning reserves, but as long as the SDR interest rate is competitive and holding SDRs is not risky, the resource gain or savings for a country from its own individual allocation of SDRs does not impose losses on other countries. Their analysis shows that the resource gains from allocations do not represent transfers from other countries, but instead represent either a welfare gain for the international economy as a whole or a resource saving offset by losses arising from changes in the terms under which the country can obtain funds outside the SDR system.


The LIBOR itself can be regarded as consisting of a risk-free interest rate, which could be represented by the U.S. Treasury bill rate plus a margin that reflects both the cost of bank intermediation and the risk premium that banks have to pay in funding their own on-lending.


Private Market Financing tor Developing Countries, World Economic and Financial Surveys (Washington: International Monetary Fund, November 1995), Table A8. Being weighted averages, these numbers are pulled lower by large sovereign borrowers like China, which enjoy low spreads—below 100 basis points.


Rojas-Suárez (1991) has calculated the implicit yield to maturity for the external debt of several developing countries by applying the present value formula on the observed secondary market price to the country’s external debt.


Estimates of the elasticity of output with respect to imports are complicated by the fact that a change in output will itself cause a change in demand for imports. Consequently, the direction of causation involved in any direct estimation of the relationship between changes in output and changes in imports may be ambiguous. Nevertheless, Leven and Roberts (1983) estimate the relationship between the growth of GDP in selected developing countries and the change in import volumes in current and previous years; their results suggest that a 1 percentage point rise in real imports leads to an increase in GDP of about 0.5 percent, of which about one half takes place in the same year and one half over the next five years.


An aspect of market imperfection in the financial intermediation process is the asymmetry between the efficiency with which information on the economies and policies of industrial and emerging market countries is conveyed to investors and lenders. Over the longer run, private markets should be expected to correct for such information gaps, but the transitional costs can be high and can persist, especially owing to sluggishness of financial market response, even after markets obtain fuller information.


In the current system, the Fund is required to pay interest to holders of SDRs regardless of whether interest is received on SDR allocations.


Polak (1988, p. 179) has noted that “when the Fund designs medium-term programs for low-income countries, the question always presents itself of the choice that has to be made between higher imports (which usually would mean higher growth) versus the build-up of reserves. To the extent that a country’s need for reserves can be met by allocation, this painful conflict between growth and reserve adequacy can be reconciled.”


In his paper for this seminar (chap. 7) Polak makes the interesting point that reduced concern about acceptance limits “was implicit in the proposal made by the countries of the Group of Seven for a special allocation of SDRs to members that had not, or not fully, participated in earlier allocations; with the exception of Switzerland, these were all potential users of SDRs, and 25 years ago any allocation that failed to generate matching increases in acceptance limits of prospective creditor countries would have been considered unthinkable.”

  • de Vries, Margaret Garritsen, The International Monetary Fund 1966–71: The System Under Stress (Washington: International Monetary Fund, 1976).

    • Search Google Scholar
    • Export Citation
  • de Vries, Margaret Garritsen, The International Monetary Fund 1972–78: Cooperation on Trial (Washington: International Monetary Fund, 1985).

    • Search Google Scholar
    • Export Citation
  • de Vries, Margaret Garritsen, The IMF in a Changing World 1945–85 (Washington: International Monetary Fund, 1986).

  • Horsefield, Keith, The International Monetary Fund 1945–65: Twenty Years of International Monetary Cooperation (Washington: International Monetary Fund, 1969).

    • Search Google Scholar
    • Export Citation