Increases in Quotas¹⁴

Increases in quotas in the Fund have a dual aspect. They increase the drawing facilities afforded by the Fund to its members—assuming no change in the Fund’s policies governing drawings—and they simultaneously increase the resources available to the Fund to meet the drawing requirements of members. They thus affect the liquidity both of Fund members and of the Fund itself.

The immediate effect on the external liquidity of members is complex, but on balance expansionary. Twenty-five per cent of any increase in quotas has normally to be paid in gold and the rest in domestic currency. The subscribing member’s owned reserves are, of course, reduced by the payment of gold. In return, however, the member’s potential drawing facilities increase by at least 125 per cent of the addition to its quota.¹⁵ Where the member, just before the increase in its quota, has no drawings outstanding beyond its former gold tranche, one fifth of the additional facilities—equivalent to 25 per cent of the addition to its quota—will consist of quasi-automatic drawing rights in the gold tranche, and the member will suffer virtually no decline in unconditional liquidity. Where, however, the member has previously drawn beyond the gold tranche, less than one fifth, if any, of the additional drawing facilities will be in the gold tranche and the member will suffer some decline in unconditional liquidity. Moreover, in either case, to the extent that the member finds the gold for its subscription, not out of its own holdings but by converting foreign exchange reserves, there will be a secondary loss of gold reserves in the reserve currency countries. On the other hand, conditional liquidity in the form of drawing facilities in the credit tranches will be increased by an amount lying between 100 and 125 per cent of the addition to member quotas—an increase which, for most members, is much more important than the deterioration in the composition, or decline in the amount, of their unconditional liquidity.

The foregoing discussion of the effects of quota increases on liquidity has proceeded on the assumptions that quasi-automatic drawing rights are available only to the extent of gold tranche positions, and that the acquisition of gold by the Fund in subscriptions is neither reduced under Article III, Section 4(a), nor offset by the use of gold owned by the Fund to acquire currencies or investments. The effects of removing these assumptions are discussed later in this paper.

So much for immediate effects. In the longer run it may be assumed that, with unchanged drawing policies, outstanding drawings in the credit tranches will come to represent as great, or almost as great, a proportion of the enhanced quotas as, in the absence of the quota increase, they would have represented of the old quotas.¹⁶ When this adjustment has been accomplished, gold tranche positions will have increased in much the same proportion as quotas, and the increase in such positions will exceed the gold subscriptions associated with the increase in quotas. Nevertheless, as long as quasi-automatic drawing rights are confined to the gold tranche, it is very doubtful whether quota increases, even in the longer run, make any significant net contribution to the expansion of unconditional liquidity.

If, provisionally, one ignores any effect which the transfer of gold to the Fund may have on the amount of foreign exchange that members are willing to hold in their reserves, the Fund’s net contribution to unconditional liquidity up to any point of time can be measured by the amount of members’ gold tranche positions less the amount of gold held by the Fund. At the end of 1963, the net contribution, thus measured, amounted to some $1600 million. Of this amount, $800 million was attributable to the fact that between 1956 and 1960 the Fund invested that amount of its gold in U.S. securities. The remaining $800 million, reflecting the Fund’s ordinary transactions, particularly the amount of outstanding drawings in the credit tranches,¹⁷ represented some 5½ per cent of member quotas. This figure has varied between 3 per cent and 8 per cent of quotas during the past decade. If, now, one takes account of the indirect effect of the net transfer of gold by members to the Fund in reducing members’ holdings of reserve currencies, and assumes that one fourth of the gold so transferred¹⁸ was obtained by the conversion of reserve currencies, it would appear that the consequential reduction in foreign exchange reserves, and hence in total owned reserves, may have amounted to some 4 per cent of total quotas. It will be seen that any net contribution of the Fund to unconditional liquidity as a result of ordinary transactions has thus far been insignificant.

As long as additions to Fund quotas keep pace roughly with the rising demand to use the Fund’s drawing facilities and as long as present drawing policies and repurchase arrangements remain unchanged, there is no reason why there should be any great change in the average proportion that outstanding drawings in the credit tranches bear to quotas; under these conditions, quota expansion must be regarded almost exclusively as a means of increasing the supply of conditional, as distinct from unconditional, liquidity.

If quotas were increased more rapidly than the demand for drawing facilities, the proportion of outstanding drawings to quotas would tend to decline, and the contribution of the Fund to unconditional liquidity would tend to become negative—gold and foreign exchange reserves would decline faster than gold tranche positions would expand. On the other hand, the expansion in the supply of, relative to the demand for, conditional drawing facilities would reduce members’ need for unconditional liquidity, and this is probably much more important than the above-mentioned effect on the supply of unconditional liquidity. To take a highly simplified example, suppose that with total quotas of 100, Fund gold holdings amount to 25, drawings outstanding in the credit tranches to 6, and gold tranche positions to 31, and that the transfer of 25 of gold to the Fund has resulted in a decline of 6 in members’ foreign exchange reserves. The Fund’s net contribution to conditional liquidity in the form of unused credit tranche positions would then be 94, and its contribution to unconditional liquidity would be 0. If, now, quotas were raised from 100 to 200, without any corresponding increase in the need for Fund drawing facilities, the Fund’s gold holdings and the associated negative impact on members’ foreign exchange reserves might both be doubled. Drawings outstanding in the credit tranches, however, would probably rise less than in proportion to quotas, say, by 3. Gold tranche positions would rise by only 28, against a loss of gold and foreign exchange reserves of 31, so that the Fund’s contribution to unconditional liquidity would decline by 3. However, the rise of 97 in its contribution to conditional liquidity (drawing facilities in the credit tranches) would surely reduce the need for unconditional liquidity by far more than 3.

Regarded as a method of increasing the liquidity provided by the Fund, increases in quotas have the disadvantage that they tend to occur—if past experience is any guide—in large amounts at infrequent intervals. A general increase in quotas has thus far occurred only once in the Fund’s history—in 1959—although, of course, increases in individual quotas have occurred at other times. It is true that these abrupt and infrequent expansions in conditional liquidity are only gradually utilized in drawings. Nevertheless, it is arguable that the abrupt increases in conditional—and reductions in unconditional—liquidity associated with increases in quotas should themselves be smoothed out over time.

One possible method would be for general increases in quotas to be considered annually; this, however, runs into the difficulty that countries are unwilling to seek authorization from their legislatures for such increases as often as once a year. A more promising variant would be for increases in quotas to be undertaken at less frequent intervals, say, once in five years, but to come into effect by annual installments.

Changes in Drawing Policies

The Fund possesses considerable power to vary the conditions on which consent is given to drawings or to the granting and use of stand-by arrangements. In particular, it can alter the degree of scrutiny of applications for, or the severity of the conditions imposed on, drawings over any given range of IMF positions; or alter the amount of drawings (in relation to quotas) to which any given policy applies.

For example, if it were felt that the quasi-automatic drawing rights now applicable to the gold tranche were still not quite automatic enough to induce all members to regard them as fully equivalent to their other reserves, it might be possible to arrange, by permanent stand-by arrangements or otherwise, that members could draw within that tranche without their applications even having to be considered by the Executive Board. Members would still be under obligation to draw only to meet a payments deficit, and would consult with the Managing Director regarding the currencies to be drawn.

Again, the proportion of a member’s quota that may be drawn upon quasi-automatically could be altered. For example, drawings in, say, the first 5 per cent of quota beyond the gold tranche, instead of being granted on the condition (now applicable to the whole of the first credit tranche) that the member “is making reasonable efforts to solve its problems” could be granted on the same conditions as apply to the gold tranche. If such a step were taken, there would be an immediate increase in the amount of quasi-unconditional liquidity made available to members through the Fund. Unless further steps were taken, however, this increase in unconditional liquidity would involve some decline in conditional liquidity, since the amount that could be drawn within the credit tranches, subject to policy conditions, would be protanto diminished. However, the substitution of unconditional for conditional liquidity would no doubt have an effect on countries’ policies similar in kind, though not in degree, to that resulting from an outright increase in reserves.

Again, members might be permitted, through the exercise of the waiver power, to draw up to a higher maximum limit than is now customary,¹⁹ subject, however, to suitable conditions. Here, obviously, the immediate effect would be to increase conditional liquidity. However, since the adoption of this policy would lead to increased drawings, it would also tend to enhance the quasi-unconditional gold tranche positions in the Fund of those countries whose currencies are drawn.

Clearly, the above-mentioned policies could be combined. For example, if gold tranche drawing rights were extended to the first 5 per cent of the member’s quota beyond the gold tranche, the normal maximum of drawing facilities—apart from the compensatory financing facility—could be extended into the fifth credit tranche, e.g., to the point where the Fund would be holding 205 per cent of the member’s quota. In this way, unconditional liquidity would be increased without any reduction in conditional liquidity.

The examples of changes in drawing policies given above have as their object an increase in the amount or automaticity of drawing facilities.²⁰ It is easy to see how they could be applied in reverse so as to bring about the opposite result.

Each extension of quasi-automatic drawing rights gives rise to a once-for-all increase in unconditional liquidity, followed by recurrent increases with each subsequent expansion of quotas. For example, the extension of quasi-automatic drawing rights to the first 5 per cent of quota beyond the gold tranche would yield an increase of $600-750 million in quasi-unconditional liquidity,²¹ and—if we assume a long-run constancy in the proportion of outstanding drawings to quotas—would mean that any subsequent general expansion in quotas would increase unconditional liquidity for this reason alone by some 4 or 5 per cent of the increase in quotas.

Now, suppose it were desired that the Fund should contribute to the expansion of conditional and unconditional liquidity, respectively, at specified annual rates. One way of achieving this result, in principle, would be to expand drawing facilities, both conditional and unconditional, through a liberalization of drawing policies, while keeping quotas constant. Such a course, however, would soon run the Fund into difficulties in finding the resources wherewith to implement the drawing facilities it offered. A more practical approach would be to increase quotas annually by an amount equal to the total increase in liquidity desired, while simultaneously extending quasi-automatic drawing facilities at the expense of conditional facilities. Since, as time went on, a given quota increase as such would generate a larger and larger increase in unconditional liquidity, the need for a progressive substitution of quasi-automatic for conditional drawing facilities would diminish and finally cease. Thus, suppose that the world “needed” a Fund contribution to unconditional liquidity amounting to $600 million per annum, and a contribution to conditional liquidity of equal amount.²² In order to add $1200 million to total liquidity, it would be necessary to expand quotas by some $1280 million.²³ On the assumption that an increase in quotas under present conditions adds nothing to unconditional liquidity, the $600 million of unconditional liquidity required would have to be obtained in the first year by extending quasi-automatic drawing facilities at the expense of conditional drawing facilities by an equivalent amount (i.e., less than 4 per cent of present quotas). In the second year, a further addition of $1280 million to quotas under the new drawing conditions would, of itself, add some $45 million to unconditional liquidity, thus reducing to $555 million the addition to unconditional liquidity for which a further substitution of quasi-automatic for conditional drawing facilities would be required. And so on. Once quasi-automatic drawing facilities had been extended approximately through two credit tranches, amounting to 50 per cent of quotas, all required further expansion in unconditional liquidity would be provided by quota increases.

Turning to the second of the objectives listed above (p. 180), the question arises whether drawing policies can be adjusted with sufficient speed and flexibility to make a significant contribution toward meeting cyclical or short-term variations in the need for international liquidity. Though changes no doubt occur fairly frequently in the detailed interpretation given to these policies, they have been formally altered on only a few occasions in the history of the Fund, and only once or twice in a restrictive sense. To some extent this is inevitable. An explicit change in drawing policies requires a laborious process of formulation, and after this process has been gone through, time is required before members can see what the change means in practice. Moreover, the adoption of more restrictive policies might be regarded by some members as a breach of the understanding on which they had previously requested increases in quotas. Nevertheless, some of the inflexibility could be remedied. For example, it is not necessary that the tranches with respect to which policies are formulated should always be multiples of 25 per cent of quota. Moreover, despite the difficulties mentioned above, a more systematic periodic reconsideration of drawing policies might be aimed at, with a view to their being tightened up at times when inflationary pressures prevail or when countries show an undue reluctance to adopt realistic rates of exchange, and relaxed at times when the tendency is toward deflation or when a spread of restrictions tends to curtail the volume of international trade and payments.

Changes in Gold Policies

The Fund acquires gold (1) as part (normally 25 per cent) of initial quota subscriptions or of increases in quota subscriptions; (2) in a certain proportion of repurchases under Article V, Section 7(b), and Schedule B; (3) in some repurchases outside Article V, Section 7(b); (4) in charges under Article V, Section 8(f); and (5) in sales of currency for gold under Article V, Section 6(a).

The Fund may use gold (1) to replenish its holdings of scarce currency under Article VII, Section 2(ii); (2) to acquire income-earning investments under implied powers; or (3) to repay loans contracted under Article VII, Section 2(i).

The Fund has little power to vary the rate at which it acquires gold, though (1) in connection with an increase in a quota, the Fund may reduce the proportion of the increase to be paid in gold if the reserves of the member are less than its increased quota; and (2) it may be possible to induce members to include more or less gold in their repurchases outside Article V, Section 7(6). Apart, however, from the possibilities, discussed below, that the Fund might borrow gold or that it might secure the repayment in gold of investments originally made in gold, the main scope for varying the Fund’s gold holdings probably lies in the discretion it possesses with respect to the use of gold.

If the Fund uses its gold to purchase currencies that it needs for drawings, or if it induces members to repurchase in currencies—necessarily the currencies of net creditor countries—rather than in gold, the result in both cases is likely to be the same, namely, a decline in the Fund’s holdings of gold (as compared to what they would otherwise have been) and an increase in its holdings of the currencies of net creditor countries. The effect on the liquidity of the members primarily concerned will be slight, the increase in their gold reserves being offset, or nearly so, by the decline in their gold tranche positions. However, the countries whose currencies have been bought will probably consider their ratio of gold to foreign exchange assets as having increased and therefore use a fraction of their gold acquisitions for the purpose of acquiring reserve currencies. In this case there will be a secondary expansion in the reserve positions of the reserve currency countries.

The use of gold by the Fund to purchase creditor currencies may have further effects on world liquidity via its effects on the Fund’s own liquidity. If the currencies acquired are urgently needed, their acquisition may enable the Fund to avoid a contraction in drawings or a drawing of unsuitable currencies that would otherwise have been necessary. Such purchases will therefore affect the amount and distribution of international liquidity in an expansionary sense. On the other hand, if the currencies acquired are not urgently needed, the Fund’s liquidity may on balance be adversely affected, since the currencies acquired may not be those that will be most needed later, and the gold which could have been used later to acquire the latter currencies will have been dispersed.

To reduce the gold proportion of subscriptions to the Fund would have an expansionary effect in some ways greater, in other ways less, than the use of gold to acquire creditor currencies. Some, at least, of the countries whose gold subscriptions are remitted are likely to have substantial amounts of drawings from the Fund outstanding. Such countries, had they paid subscriptions in gold, would have lost reserves and gained, not drawing facilities in the gold tranche, but merely conditional drawing facilities in the credit tranches. The remission of the gold subscription thus enhances their liquidity. Moreover, since they would probably have obtained the gold by selling foreign exchange, the remission avoids a drain on the reserves of the reserve currency countries also. On the other hand, the substitution of holdings of currencies for holdings of gold tends to reduce the liquidity of the Fund.

If the Fund uses gold to repay indebtedness previously contracted under borrowing arrangements, the effect on international liquidity will probably be less expansionary than the operations so far considered. If, as is probable (see the section, Borrowing Arrangements, p. 203, below), the Fund instruments of indebtedness have conferred on their holders unconditional liquidity similar to that conferred by a gold tranche position, the “automatic” effects of such repayment (other than those resulting indirectly from repercussions on the Fund’s own liquidity) will be similar to those of a use by the Fund of gold to acquire a creditor currency, i.e., mildly expansionary. The effects on Fund liquidity, however, will be as adverse as in the circumstances described in the preceding paragraph, and the reaction on the Fund’s drawing policies might be a contractionary one.

The expansion in international liquidity that can be achieved by reducing the Fund’s gold stocks is necessarily limited in amount and is of a once-for-all character. On the other hand, in the absence of acts (of replenishment or investment) specifically designed to return gold to countries’ reserves, the Fund might gradually accumulate a stock of gold in excess of what it requires in the interests of its own liquidity. Such an accumulation would exercise on the growth of unconditional liquidity in the world a negative influence which, even if not strong, might well be undesirable.

The accumulation and decumulation of gold by the Fund could, in principle, be used as a means of bringing about variations in world liquidity but, for the reasons explained, their effect is likely to be moderate and largely confined to reserve currency countries. In any event, gold operations constitute a reasonably flexible instrument only in an expansionary direction. The Fund has no established means of acquiring gold in substantial quantities, save as a part of a general increase in quotas. One way of strengthening this instrument for use in a contractionary direction might be for the Fund to borrow gold. The effect of this would not be the precise opposite of the use of gold to repay indebtedness, since the enhancement which would occur in the Fund’s own liquidity could be, as it were, sterilized; it need not be allowed to affect the Fund’s policies on drawings, etc. However, any contractionary effect of such borrowing on the external liquidity of member countries would, as has been shown, be a mild one. Another possible method for the Fund to acquire gold would be for it to obtain repayment in gold of short-term investments which it might possess.²⁴ The right to secure repayment in this form would be natural in the case of investments originally acquired through the payment of gold, and is indeed provided for in the case of the gold investments at present held by the Fund for income purposes. If the repayment were made in gold, the contractionary effect on world liquidity would be somewhat greater than if the repayment were made in currency.

Changes in Repurchase Provisions

The resources of the Fund are used to assist members in meeting temporary balance of payments deficits, including deficits arising out of seasonal, cyclical, or emergency situations. Under the Articles, members are obliged at the end of each year to repurchase on the basis of the development of their reserves. As a matter of Fund policy, members undertake to, or represent that they will, repay any drawings within three to five years of the date of drawing, and even earlier if the payments problem for which the drawings were made has been solved. The express repurchase provisions of the Articles show clearly that drawings from the Fund are intended to compensate or offset balance of payments fluctuations. The repurchase policies subsequently adopted imply a view that countries should aim at keeping their payments and receipts in balance over a moderate number of years. The Fund’s ability to induce drawing members to achieve such equilibrium by acceptable methods would be greatly reduced if there were not a time limit on repurchases. Under present policies, the Fund can say that it will renew drawings that fall due for repurchase only if suitable policies are adopted. Thus, the effective conditionality of the use of the Fund’s resources is linked to their revolving character.

By an act of policy, the Fund could alter its terms of repayment for drawings in ways that would allow drawings, or certain classes of them, to remain outstanding for longer periods. For example, the outside limit for drawings could be made four to six years, or five to seven years, instead of the present three to five years. The longer the period specified as the limit for repurchases outside Article V, Section 7(6), the greater would be the proportion of drawings which would have to be repaid under the provisions of Article V, Section 7. It could indeed be argued that the logical terminus to the process of liberalizing repurchases would be the abandonment of all stipulations as to repayment other than those arising under Article V, Section 7.

As has just been shown, such measures, particularly if carried to the point of abolishing all repayment undertakings, are open to the objection that they weaken the Fund’s ability to secure the adoption by drawing members of appropriate balance of payments policies. This objection would not, however, have great force with respect to drawings in the gold tranche, or in any other tranche to which quasi-automatic drawing rights may be extended, since in these tranches repayment provisions can easily be nullified by fresh drawings. The withdrawal of time limits on the use of the Fund’s quasi-automatic drawing facilities would therefore merely formalize the actual state of affairs, and would have the advantage of strengthening the resemblance of these drawing facilities to other reserves held by members. There may therefore be a case for rendering inapplicable any repurchase obligations, other than those arising under Article V, Section 7(6), where the effect of such repurchases would be to reconstitute quasi-automatic drawing facilities.

Where conditional drawing facilities are concerned, there are still further objections to the elimination of repayment undertakings or to extending their terms unduly. To make repayment contingent entirely on a recovery in reserves might weaken the incentive for countries to achieve a payments surplus and a consequential rise in reserves. Moreover, as the duration of the drawing extends beyond five years, it becomes more and more questionable how far it can be regarded as bona fide balance of payments financing. The concept of compensatory or balance of payments financing rests on the notion that countries should so act as to ensure that any deficits in the remaining (noncompensatory) items of the balance of payments are succeeded by equivalent surpluses in the foreseeable future. Where the balance of payments is concerned, it is difficult to forecast with any confidence a period more than five years ahead—though admittedly the possibilities of balance of payments planning vary greatly as between countries.

Any extension of the time allowed for the repayment of drawings along the lines discussed above would increase the degree of utilization of quotas (i.e., the average ratio of outstanding drawings to quota), and would thus increase the proportion of unconditional liquidity to conditional liquidity (as measured by the credit tranche positions remaining unused) provided by the Fund. It is unlikely, however, that any mere lengthening of repayment terms while retaining the repurchase provisions of Article V, Section 7(b), would make a significant contribution to the growth of unconditional liquidity or reserves. If this effect were desired, it would probably be necessary for the Fund to transcend the sphere of balance of payments financing altogether and to provide, in addition to such financing, longer-term lending of noncompensatory kinds. It would probably be a wrong approach to the problem to seek to introduce a new type of long-term drawing facility that would be exempt from the repurchase provisions of Article V, Section 7(6), but would still be related to quotas and still be activated on the initiative of the drawing country. Since the primary purpose of the Fund’s entry into the longer-term lending field would be less to accommodate the immediate borrowers than to enable the Fund to make a net contribution to the stock of reserves, it would be more appropriate to consider such longer-term lending under the head of investment by the Fund, which is the subject of the next section.

Investment by the Fund

By “investments” are here meant essentially any securities (other than the Fund’s own instruments of indebtedness) which the Fund might buy or sell but which would not count as holdings of a member’s currency for the purpose of calculating drawing entitlements, repurchase obligations, etc., and would therefore leave unaffected such rights and obligations. Under implied powers, the Fund has used gold to the extent of $800 million to acquire investments, in the form of interest-bearing U.S. dollar securities redeemable on demand with a gold guarantee, for the purpose of maintaining the income of the organization and providing a certain reserve.

In what follows, we consider the consequences of extending the concept of investment by the Fund so that it may be undertaken for purposes relating not only to the Fund’s income but also to international liquidity, and so that it may be financed not only out of the Fund’s gold holdings but also out of its currency holdings (whether the currency of the country in which the investment is made or another) or by borrowing. For simplicity, it is assumed that the investments take the form of securities denominated in the currency of the country in which the investment is made (the “country of investment”), but subject, like currency holdings under the Articles, to a gold-value guarantee; and that they are bought and sold from or through the government of the country of investment.

Since this is a type of operation not yet undertaken, it can be given any characteristics desired. However, in order that the power of investment should be useful, we shall assume it to possess the following characteristics, the first of which follows from the definition given above. First, investments would not enter into the calculation of the Fund’s holdings of currencies, or of the IMF position of the countries of investment, and would not as such affect their drawing rights (though the method of acquisition of the investments might do so). Second, the formal initiative as to the buying or selling of investments in any particular country would lie with the Fund, though the consent of the country of investment would be necessary both for purchases and for sales before maturity. Third, investments would be made on long term as well as on short term.

By virtue of the first characteristic listed above, investment would permit of an increase in the amount of external liquidity, unconditional and conditional, available to Fund members at any given level of quotas. This effect would be most clearly seen if the Fund purchased an investment in a member country by using its ordinary holdings of that member’s currency. The decline in these holdings would improve the IMF position of the country of investment, and thus increase its drawing facilities, without impairing the IMF position of any other member. The initial position of the country of investment would determine whether the drawing facilities acquired were quasiunconditional gold tranche drawing rights or conditional facilities in the credit tranches.

If the Fund used its holdings of country B’s currency to invest in country A, the effect would be the same as if it had used its holdings of A’s currency for this purpose while at the same time A had made a drawing of B’s currency. If, as might often be true for this kind of investment, A had initially a debtor IMF position and B a creditor position, the effect of the drawing element in the transaction would be to transform what would otherwise have been an expansion of conditional liquidity into an expansion of unconditional liquidity.

Again, if the investment in A were paid for out of the Fund’s holdings of gold, the effect would be similar to a combination of an investment financed from the Fund’s holdings of A’s currency and a use of the Fund’s gold to purchase A’s currency. Generally speaking, the investment of gold would have a more expansionary effect on unconditional liquidity than any other kind of investment.

The second characteristic suggested for the power of investment, namely, that the Fund should have more initiative and freedom of action than with respect to drawings, is clearly one the degree of which is debatable. It would, however, seem desirable that member countries should have no right to receive an amount of investment, or even a share in total investments, in any way related to quota, though doubtless the solicitation of investment could not in practice be excluded. Though for investments, as for drawings, the consent both of the Fund and of the country of investment (cf. the drawing country) would doubtless be necessary, nevertheless the fact that the initiative lay with the Fund should give the latter much more effective power in influencing the amount, timing, and distribution of investments than it has with respect to drawings.

The third characteristic proposed for investment, that it extend to assets of long as well as short maturity, is, of course, of fundamental importance for the Fund’s ability to expand the total amount of credit—and hence the amount of unconditional liquidity created in the form of gold tranche drawing rights or instruments of the Fund’s indebtedness, relative to the amount of short-term to medium-term conditional drawing facilities which it provides. Investment by the Fund is thus a third way of bridging the possible gap between what the Fund ought to contribute to the expansion of unconditional liquidity or reserves and what, if anything, it contributes as the counterpart to the use of its conditional drawing facilities. The other two possible ways of bridging this gap, as already explained, are an extension of quasi-automatic drawing rights beyond the gold tranche and a lengthening of the term of drawings.

Investment differs from an extension of quasi-automatic drawing rights beyond the gold tranche in five main respects: (1) it can be distributed more selectively than an extension of automatic drawing rights, which would normally be applied to all members in proportion to their quotas; (2) insofar as the investment is made with a currency other than that of the country of investment, it resembles an extension of automatic drawing rights plus an actual drawing; (3) investment will create more drawing facilities free from repurchase obligations than will an equal extension of automatic drawing rights beyond the gold tranche; (4) investment will constitute a bigger drain on the Fund’s resources than will an equal extension of automatic drawing rights;²⁵ (5) whereas an extension of automatic drawing rights beyond the gold tranche automatically increases the effectiveness of increases in quotas in expanding unconditional liquidity, investment does not. To achieve a similar result, investment would have to be increased pari passu with the increases in quotas.

Investment of the type outlined differs from the introduction of longer-term drawing rights mainly with respect to the first two characteristics of investment discussed above. Owing to these characteristics, Fund investment has the advantage over the provision of long-term drawing facilities that it could more easily be kept distinct from the Fund’s everyday task of providing short-to-medium-term drawing facilities, and is therefore more likely to leave that valuable function unimpaired.

In one respect, investment might act more powerfully on a country’s policies than either of the other two methods of expanding credit from the Fund, viz., if it is regarded by the country of investment as entering into its balance of payments as a positive item “above the line.” An increase in liquidity that results from a payments surplus (or a decline in liquidity that results from a payments deficit) is likely to be more effective than one arising independently of the balance of payments.

Insofar as investment is employed for the purpose of enabling the Fund to contribute to a long-term expansion in the trend of unconditional liquidity, it would seem appropriate that it should take the form of a purchase of securities amortized over a long period of time—e.g., 20 years. While the Fund would, of course, not be precluded from holding long-term investments in the more highly developed countries, it would seem desirable that as high a proportion as possible should be held in those countries where the need for capital is, humanly speaking, greatest—namely, the less developed countries. Since the latter would be unlikely to use the proceeds of such investments to build up their own reserves, but would spend them on the products of the industrial countries, the effect on the reserves of the industrial countries would probably be almost as great as if the investments had been made in these countries in the first place. Moreover, insofar as the object of expanding liquidity is to encourage the industrialized countries to adopt more expansionary financial policies, the underlying purpose would be better achieved by investing in the less developed than in the more highly developed countries, since in the former case increased liquidity would accrue to the industrialized countries in the form of export receipts, with a stimulating effect on incomes, and in the latter case it would accrue only in the form of greater ease in the money and securities markets. Investment by the Fund in less developed countries should, of course, not exceed amounts which the countries in question could economically absorb at the relatively low rates of interest which the Fund would be able to charge. Such investments would not be part of the Fund’s own liquid reserves; they would be a means to the creation of liquidity in the hands of national monetary authorities, with beneficial side effects on the provision of resources for economic development.

Any long-term investment by the Fund in the less developed countries should presumably be carried out as far as possible through existing intermediaries, such as the IBRD and the IDA, though neither the terms and conditions of the investments nor even the investment criteria need be precisely those at present applied by either of these institutions. While the technique here discussed has some resemblance to the “Stamp Plan,”²⁶ there is no suggestion that the investments should be in any way tied to imports from such countries as are willing to accept additional claims on the Fund, nor is it suggested that those claims should necessarily take the form of transferable certificates rather than, for example, gold tranche positions.

The technique of investment (and disinvestment) has certain advantages over that of changes in the liberality or severity of drawing policies as a means of bringing about such temporary variations in the amount of liquidity as may be useful in counteracting cyclical fluctuations in the industrial world. Temporary changes in the amount of unconditional liquidity, such as may be called for by alterations in the condition of the foreign exchange markets, are likely to be evoked, through variations in the demand for drawings, even with constant drawing policies. Where some positive anticyclical action is called for, however, drawing policies are likely to prove too inflexible, and open market policies more effective. Investments for this purpose could be of a short-term though renewable character, and would normally be made predominantly in the industrial countries.

Long-term investment in central reserve countries might also be an appropriate means of counteracting any sudden and substantial decline in official holdings of reserve currencies. These investments should be on long term to cover the possibility that the decline might be a permanent one, but there could be arrangements for withdrawing the investments if official holdings of the currencies in question should recover.

Increases in Quotas

Increases in quotas affect countries’ liquidity in the manner discussed in an earlier section. They also provide the Fund with additional resources consisting of gold and “drawable” currencies, the former to the amount of 25 per cent of all quota increases, and the latter to the amount of 75 per cent of the additions to quotas of countries whose currencies are, at any given time, usable for drawings. The addition to the Fund’s holdings of gold and drawable currencies is intended to maintain the Fund’s liquidity, i.e., to enable it to honor the enhanced drawing facilities that accompany the quota increases without the necessity of paying out the currencies of countries that are in a weak balance of payments position or reserve position and might thus be embarrassed by the drawing. In the short run, before countries have had time to use their new drawing facilities, a rise in quotas will necessarily increase the Fund’s liquidity. In the long run, however, if the increase in quotas is matched by a corresponding proportionate increase in the demand, and need, for drawings, the additional resources provided by an all-round increase in quotas will only suffice to enable the Fund to maintain its original degree of liquidity. Thus, if the original country distribution of quotas, relative to the distribution of potential payments surpluses and deficits, was such that the resources of the Fund had to be supplemented by borrowing arrangements with certain countries, and if the distribution of potential surpluses and deficits remains unchanged, a quota increase as described above will normally have to be supplemented by a proportionate increase in the borrowing arrangements. Only insofar as the increase in quotas outstrips the increase in the demand, and need, for drawings will it be possible for the Fund to reduce its reliance on such arrangements.

Thus far we have dealt with general, all-round increases in quotas. Quotas, however, can be increased at different times for some classes or groups of members (e.g., members with “small” or “medium” quotas), to the exclusion of others. If the quotas increased are primarily those of countries that are seldom in a net creditor position and frequently in a debtor position, the increases will have effects in some ways comparable to the adoption of more liberal drawing policies—the level of outstanding drawings will be increased without any increase in the Fund’s resources and with some decline in the Fund’s liquidity. Conversely, if the quotas increased are mainly those of countries normally in a creditor position, the main—though not the only—effect will be to increase the resources and the liquidity of the Fund, and to reduce the need for borrowing.

Borrowing Arrangements

Under Article VII, Section 2, the Fund can borrow a member’s currency from that member, or (with the consent of both parties) from another source, whenever “it deems such action appropriate to replenish its holdings of any member’s currency.” Such borrowing may affect the liquidity of members directly, but much more important is its effect, or the effect of the arrangements under which it is carried on, on the liquidity or resources of the Fund itself.

The nature of the borrowing arrangements will determine (1) the nature of the resources made available to the Fund; (2) the conditions governing repayment and the other characteristics of the instruments of Fund indebtedness; and (3) the conditions under which the resources are made available to the Fund.

As regards (1), the proviso from Article VII, Section 2, quoted above, would seem to confine the currencies that might be borrowed to those suitable for being drawn from the Fund, though the Fund’s need for the currencies in question does not necessarily have to be immediate.

As regards (2), countries acquiring claims on the Fund by lending to it might be able to require the Fund to convert these, on demand, into currencies or gold; or the claims might be repayable at fixed dates but (like gold tranche drawing rights) convertible into foreign currencies whenever required to meet payments difficulties experienced by the lending country; or convertible in some proportion to the decline in that country’s other reserves; or repayable only on maturity. The duration of the claims might be short or long. They might be transferable generally, as between central banks, or not at all. They might be interest-bearing or otherwise.

The precise degree of liquidity to be accorded to a claim on the Fund arising out of an act of borrowing is thus variable, and should depend on the effect that the Fund wishes to exercise on world liquidity. In practice, it might be difficult to induce lending members to be content with a lesser degree of liquidity than that afforded by a gold tranche position in the Fund unless substantial interest were paid. On the other hand, it would be important for the Fund’s own liquidity that lenders collectively should not be in a position to encash too high a proportion of their claims on the Fund at any one time. Any system of deposits withdrawable (i.e., convertible into currency or gold) on demand would leave the Fund exposed to the possibility of “runs” that might deprive it of the power to maintain or expand world liquidity at a time when this was required. It would, therefore, seem necessary that repayment of the Fund’s indebtedness on the initiative of the creditor should be confined to situations in which the creditor was able to represent that its balance of payments created a need for such repayment. The precise degree of liquidity that could safely be given to instruments of the Fund’s indebtedness would depend in part, as will be shown, on the nature of the access to resources conferred on the Fund by the borrowing arrangements.

In discussions of international liquidity, emphasis is often laid on transferability as an important attribute affecting the liquidity of claims on the Fund arising out of Fund indebtedness. Any value such transferability might have for the holders of such claims would depend on the willingness of other countries to accept transfers. Such willingness could certainly not be relied on in the absence of definite arrangements whereby countries would agree to accept transference of such claims up to specific amounts. These arrangements, in turn, would be likely to be at the expense of undertakings to provide resources directly to the Fund. Countries might even insist that claims on the Fund transferred to them be counted against any lines of credit which they might be extending to the Fund. The system is therefore not one which would provide additional facilities for Fund members at no cost to the Fund’s resources. The fact that the transferee would be, by definition, a willing holder of claims on the Fund would be an advantage, but bilaterally arranged transfers might conflict with those general understandings regarding an equitable distribution of liquid claims on the Fund which would probably be necessary if the Fund was to supply unconditional liquidity on a large scale. If the rights of the creditor to encash indebtedness to meet payments difficulties were made sufficiently automatic—in conformity with suggestions made earlier for further increasing the automatism of drawings in the gold tranche—the attractions of transferability to the lender might well become negligible.

It should be noted that if, as appears likely, the degree of liquidity attending to instruments of Fund indebtedness were not very different from that attaching to a gold tranche position, an act of borrowing by the Fund, taken by itself, would be likely to have only a minor effect on countries’ external liquidity. A member lending its own currency would probably be a country with a gold tranche position. The rise in the Fund’s holdings of its currency would involve a decline in that position, to offset which it would acquire a claim on the Fund of a degree of liquidity comparable to that afforded by a gold tranche position.²⁸ Conceivably the claim on the Fund, having the legal form of an asset, would be deemed by the member as more worthy of inclusion in its reserves than any kind of drawing right. The real gain in liquidity would, however, be slight or negligible. Again, if a member lent to the Fund gold or the currency of another member, the lending member would acquire a highly liquid claim on the Fund but would suffer a corresponding loss in its reserves. A country whose currency was lent to the Fund by another member would suffer a decline in its Fund position, probably a gold tranche position, but its other reserves might benefit from the reduced supply of its currency in the international exchange market. However, if the act of lending to the Fund involved a decline in the holding of foreign exchange reserves, the net effect would be to reduce international liquidity.

The real significance of an act of borrowing is that it would increase the Fund’s resources without having much direct or immediate effect on members’ liquidity. (In this respect it contrasts with a quota increase, which, as we have seen, would not only increase Fund resources but also immediately enhance members’ conditional liquidity in the form of drawing facilities in the credit tranches.) However, as soon as the Fund used the resources it had borrowed, whether for additional drawings or investment, a net increase in international liquidity of the sort described in earlier pages would occur.

As regards (3), resources might be made available to the Fund (a) in the form of deposits made on the initiative of the lender, though on conditions laid down by the Fund and possibly subject to quantitative limits established by the Fund; (b) under lines of credit, calls on which could be on the Fund’s initiative subject to conditions agreed in the credit arrangement; or (c) as the outcome of an agreed rule of a more or less automatic statistical character.

The Fund could not, at present, accept deposits entirely at the will of the lender and in any currency that the lender chose to deposit. It could accept only currencies potentially needed for drawings, and then only if the country whose currency was deposited consented. However, the Fund could reasonably declare its willingness to accept deposits in any currency of which its holdings fell below a certain low percentage of quota (say, 25 per cent or 20 per cent), provided that the deposits were made by or with the consent of the country whose currency was deposited. Interest rates could be such as to make such deposits attractive. While depositors might hesitate to incur the loss involved in depositing at par strong currencies standing at a premium in the market, they might hope to recoup this loss by drawing premium currencies at par when the time came to withdraw their deposits. One attraction of this technique would be that, since the deposits would be made on the initiative of the depositor, no negotiation would be necessary between the lender and the Fund.

On the other hand, the system of voluntary deposits would seem to leave the Fund dangerously exposed to net withdrawals of its resources. This would be true even if the right to withdraw were confined to countries with a payments need, for there would be no assurance that surplus countries would necessarily be induced by the mere attraction of the interest paid, the gold guarantee, etc., to add to their deposits at a time when deficit countries were being compelled by payments difficulties to withdraw theirs. If deposits were encashable on demand, the situation would be even more precarious, since a general nervousness on world exchange markets, or a lack of confidence regarding the Fund’s own liquidity, might lead to a net withdrawal of deposits just at the time when the Fund was anxious to extend itself in assisting member countries.

Lines of credit could be established by the Fund with individual lenders or, as in the General Arrangements to Borrow, with a group of lenders. The line of credit would have one very important advantage over the mere acceptance of deposits, in that it would ensure the Fund of access to additional resources for the lifetime of the stand-by arrangement. Voluntary deposits, once they had been withdrawn (e.g., to meet a payments deficit of the creditor) might not be reconstituted if the interest paid was not attractive or if the Fund’s own liquidity was suspect. In the line-of-credit technique a similar difficulty would arise only when the credit arrangement lapsed and had to be renewed. It might, moreover, be possible to have permanent credit arrangements or lines of credit which, like subscriptions, would give the Fund a permanent increase in reserves.

The nature of the tension between a lending country’s right to secure payment of its loan at need, and the Fund’s right to draw on the line of credit at its need, is broadly analogous to that described earlier as existing between a member’s right to draw on a gold tranche position and the Fund’s right to sell the member’s currency and thus to increase or restore that gold tranche position. Some conventional rule is required to determine to what extent the Fund should draw on any given country’s line of credit. The rule might be, for example, that the Fund would borrow from any given country to whatever extent was required to prevent its holdings of that country’s currency from falling below some (low) percentage of its quota. The Fund’s policy with respect to currencies to be drawn and used in repurchase would be determined so as to secure an equitable distribution of net IMF positions among creditor countries, including in the concept of net IMF positions not only gold tranche positions but also amounts lent to the Fund. Lending countries incurring payments deficits would be expected to take advance repayment of their loans before drawing their gold tranche positions below the aforesaid percentage of quota.

A good example of the third technique mentioned above, whereby countries lend to the Fund according to some statistical criterion, would be an arrangement under which many or all Fund members would hold claims on the Fund in some proportion to, or as a function of, their total gold and exchange reserves. These claims would be acquired by depositing gold or currencies needed by the Fund. This would provide for a long-term growth in Fund resources pari passu with the growth of official reserves and, if used to finance an expansion in investment by the Fund, or in drawings relative to quotas, would make possible a long-term growth in the liquidity provided by the Fund. Such an arrangement would, however, make great demands on the willingness of countries to undertake important commitments—though not necessarily sacrifices—for an international end. Moreover, a once-for-all commitment would not suffice. If, as is probable, it became necessary, in order to secure a higher proportionate rate of expansion of world liquidity than that of monetary gold stocks, to raise periodically the proportion of reserves held in the form of claims on the Fund, a new agreement would be called for on the occasion of each increase.

One disadvantage of a simple system under which countries would hold a proportion of their reserves in the form of Fund deposits would be that, especially when that proportion was still small, they could use such deposits to cover only a small part of any payments deficits that they might have. This would make the holding of a deposit less attractive than the holding of a gold tranche drawing right, and might constitute a real hardship for the majority of countries whose own currencies would not be acceptable for deposit and who would therefore have to establish deposits at the expense of other reserves. The hardship would be greater if a part of the gold and foreign exchange holdings of the countries in question were pledged or otherwise unusable. A possible way round this difficulty might be a provision under which deposits with the Fund could be withdrawn pari passu with a decline in the depositor’s other reserves and reconstituted pari passu with any recovery in these reserves until the required proportion was restored. Regulations might be required establishing priority between such reconstitution and repurchases of any drawings outstanding.

The three modes of Fund borrowing discussed above are merely illustrative. The borrowing instrument is very flexible. Many different combinations can be imagined. For example, the following system—involving a combination of the system of voluntary deposit with that of lending to the Fund in relation to a statistical criterion—might be of use for maintaining the level of international liquidity if that were threatened by some temporary or permanent decline in the propensity of monetary authorities to hold reserves in the form of foreign exchange, and if no adequate arrangements or understandings existed among these authorities with respect to the holding of reserve currencies.

The Fund would stand ready to receive deposits (i.e., to borrow) in reserve currencies held by monetary authorities to the extent that the latter wished to reduce their holdings thereof. Currencies thus acquired by the Fund would immediately be invested (i.e., held in a form apart from its ordinary currency holdings) in the countries whose currencies were deposited, and would thus leave unaffected the usual drawing facilities available to these countries. Such investments—which would, like other Fund investments, carry a gold guarantee—would be amortized in gold or currency acceptable to the Fund over a lengthy period (e.g., 20 years), and would also be redeemable on demand in domestic currency at the Fund’s request. If depositing countries should subsequently desire to increase their holdings of the currencies formerly deposited, they would be under obligation to do so in the first instance by withdrawing their deposits from the Fund, which would in turn encash its investments. As long as the transaction had not been reversed according to this procedure, the deposit, though convertible into needed currency if required to meet payments deficits of the depositor, would have to be reconstituted when the other reserves of the depositor increased (or possibly according to some repayments schedule).

The main effect of this arrangement would be to enable countries losing confidence in a reserve currency to substitute a gold-guaranteed claim on the Fund. If the loss of confidence should turn out to be permanent, the system would permit the euthanasia of the reserve currency in question, but this would not be assumed in advance to be inevitable. Instead, the fact that a temporary gold guarantee could be obtained whenever necessary would probably strengthen the desire to hold reserve currencies.