Professor Triffin on International Liquidity and the Role of the Fund

IN HIS RECENT BOOK, Gold and the Dollar Crisis (New Haven, 1960), Professor Robert Triffin of Yale University draws a bold picture of the dangerous state of, and prospects for, international liquidity, and suggests expanding the role and changing the character of the International Monetary Fund to cope with these problems


IN HIS RECENT BOOK, Gold and the Dollar Crisis (New Haven, 1960), Professor Robert Triffin of Yale University draws a bold picture of the dangerous state of, and prospects for, international liquidity, and suggests expanding the role and changing the character of the International Monetary Fund to cope with these problems

IN HIS RECENT BOOK, Gold and the Dollar Crisis (New Haven, 1960), Professor Robert Triffin of Yale University draws a bold picture of the dangerous state of, and prospects for, international liquidity, and suggests expanding the role and changing the character of the International Monetary Fund to cope with these problems

1. The main part of this book (pp. 17–145) consists of reprints, with only minor changes, of two long articles published in 1959 in the Quarterly Review of the Banca Nazionale del Lavoro: “The Return to Convertibility: or Convertibility and the Morning After” and “Tomorrow’s Convertibility: Aims and Means of International Policy” (March and June issues, respectively). These articles drew upon his previous writings, notably Europe and the Money Muddle (New Haven, 1957) and his Wicksell Lecture for 1958, The Future of the European Payments System (Stockholm 1958, especially pp. 34–43). The book also contains a reprint of Triffin’s statement to the Joint Economic Committee of the Congress* of the United States in October 1959 and most of the questions and answers from the Committee’s Hearings (pp. 1–17 and 167–91, respectively). The few remaining pages of the book are devoted to comments from other sources, including the Report of the Radcliffe Committee.

The general thesis of these several studies was also treated in a paper, “The Gold Shortage, the Dollar Glut, and the Future of Convertibility,” delivered at the August 1959 meeting of the International Economic Association, and an article, “Improving World Liquidity,” in The Banker (January 1960). The most recent statements of Triffin’s position are given in his paper, “Le Crépuscle de L’ Etalon de Change-Or,” presented to the Belgian Royal Economic Society in June 1960 and in his testimony before the Joint Economic Committee in December I960.1 The printed record of both occasions includes his statement and the questions and discussion that followed.

2. Section I of this paper2 is an exposition of Triffin’s diagnosis of present and prospective difficulties in the international financial area and of his prescriptions for dealing with these difficulties. Section II is an analysis of five questions that are basic to Triffin’s proposals: first, the problems involved in obtaining a stable structure of international reserves; second, the need for changing the character of the International Monetary Fund (IMF) to provide additional international liquidity; third, the characteristics of the reserve structure contemplated by the Triffin Plan, including the role played by gold and by exchange guarantees; fourth, the operational and policy problems that an expanded IMF would have faced in recent years in assessing and managing the state of international liquidity; and fifth, some political and international implications of an expanded IMF.

It should be noted at the outset that Triffin has focused attention upon some major problems of international liquidity and international financial organization. His diagnosis of present and prospective difficulties, and his recommendations for meeting these difficulties, are acute and thought provoking. They reflect his extensive knowledge of financial history and his working contributions to the solution of postwar problems. His diagnosis and his prescription are, nevertheless, highly controversial. His proposals involve a number of far-reaching changes in financial structure and operation, and in working controls over them. They deserve intensive and systematic examination. The present article tries to do this, but it does not attempt, except in passing, to discuss other suggestions of a less comprehensive character. Neither does it deal systematically with the increasingly important role of the International Monetary Fund, and of adjustments in its operating policy—past, present, and potential—designed to make it even more useful. These self-imposed limitations have given this article a more unrelieved critical cast than was originally intended. To modify this, however, would make an overlong article even longer.

I. Professor Triffin’s Views


3. Triffin’s diagnosis is that

The most fundamental deficiency of the present system and the main danger to its future stability, lies in the fact that it leaves the satisfactory development of world monetary liquidity primarily dependent upon an admittedly insufficient supply of new gold and an admittedly dangerous and haphazard expansion in the short-term indebtedness of the key currency countries.3

4. The world’s normal requirements for increases to monetary reserves appropriate to the growth of trade and the maintenance of convertibility, principally by the major trading countries, will certainly be greater than prospective additions to reserves resulting from gold production.

In the ten years 1958–67, the shortfall will be at least $6 billion and it may be as much as $17 billion, depending upon whether trade grows during the period at the rate of 3 per cent or 6 per cent a year, and also upon the following assumptions:4 (a) the United States, Germany, Switzerland, and Venezuela will not increase their reserves; (b) in the next decade the United Kingdom and France will increase their reserves to 40 per cent of imports as of the end of 1957, or to $4.6 billion and $2.5 billion, respectively, and the reserve requirements of these two countries for the ensuing decade will involve an expansion of these base figures in line with the growth of their trade;5 and (c) sales of gold by the U.S.S.R. will average $200 million a year, that is, approximately the average of annual sales in 1956–58.

Triffin calculates the reserve deficiency for the next decade, over and above $7 billion of gold additions to reserves, on the basis of four assumed rates of growth of trade, ranging from 3 per cent to 6 per cent a year, but he would probably be inclined to accept a higher rather than a lower figure.6

5. The gap between required “normal” additions to reserves and prospective additions to reserves in the form of gold cannot be filled by the growth of dollar or sterling balances. The gap cannot—and what is more, it should not—be filled by larger international holdings of these or any other national currencies. It is contrary to the interests of the United States and the United Kingdom to permit much additional growth of dollar and sterling foreign balances, since growth of these balances weakens the net reserve positions of these countries. Moreover, it is not in the interests of the world to create international reserves in the form of exchange balances by unrequited capital exports to wealthy creditor countries; nor is it reasonable to tie the growth of international reserves to the needs of the domestic policies of the creditor countries whose currencies are so used.

6. The balance of payments deficit of the United States in the postwar period has increased markedly the official dollar holdings of all other countries, excluding those in the communist bloc The reserves of many countries are at comfortable levels, while some are large. On the other hand, many less developed countries have inadequate reserves. According to Triffin, it is clear that the present situation is unstable and dangerous because of the large amounts of dollars and sterling held as reserves.

7. The use of national currencies as international reserves has had bad results, notably in the late 1920’s and early 1930’s. Sooner or later—and sooner rather than later—it will have dangerous consequences.

The “net” reserve position of the United States deteriorated sharply in the 1950’s.7 Gold holdings decreased from a high of $24.6 billion in 1949 to $19.5 billion at the end of 1959 and $17.6 billion at the end of 1960; short-term dollar liabilities to foreign official and private holders (excluding international agencies) increased from $7.6 billion in 1949 to $16.2 billion at the end of 1959 and about $18 billion at the end of 1960.

The United States is now exposed to the hazards of gold outflows on a large scale. Its ability to follow a low interest policy for domestic reasons has been sharply limited by the possibility of large capital outflows in response to higher interest rates abroad. This would mean large losses of gold reserves. In October 1959, Triffin stated:

Our huge gold losses of last year [1958] were due in part to such a repatriation of foreign capital at a time when interest rates had fallen here well below the rates available in Europe. Those gold losses have [been slowed down this year by an extremely sharp rise of interest rates in this country, prompted by our domestic concern about creeping inflation.

In this case, external and internal interest rate policy criteria happily coincided, but they may diverge tomorrow.

If and when we feel reassured about our internal price and cost trends, we may wish to ease credit and lower interest rates in order to spur our laggard rate of economic growth in comparison not only with Russia, but with Europe as well.8

8. The present structure of international reserves has grown haphazardly. The use of gold as a major component of international reserves has a historical, but hardly an economic, justification. “Nobody could ever have conceived of a more absurd waste of human resources than to dig gold in distant corners of the earth for the sole purpose of transporting it and reburying it immediately afterward in other deep holes, especially excavated to receive it and heavily guarded to protect it.”9 The use of national currencies in international reserves economizes on the use of gold, but this use was “developed haphazardly under the pressure of circumstances rather than as a rational act of creation.…”10 This development created the risk of runs on currencies and of flights to gold.

The extension of the gold exchange standard in the 1920’s had tragic consequences. Based as it was upon the willingness of the key currency countries to allow their own net reserve positions to deteriorate, it “led, in 1931, to the devaluation of the pound sterling, to the collapse of the international gold exchange standard, and to the consequent aggravation of the world depression.”11 It magnified, if indeed it was not a primary factor in, the deflation that began in 1929. The large balances now outstanding of dollars and sterling are a warning that the present situation is very vulnerable. Yet, since the end of World War II, only the growth of foreign holdings of national currencies has tended to compensate for the inability of gold production to expand international liquidity in relation to the world’s requirements.

In the next decade, gold production will “admittedly” be insufficient to meet the world’s growing demand for reserves. Given the present monetary arrangements, this deficiency can be compensated for only by increased holdings of reserves in the form of national currencies, even though existing foreign holdings of national currencies are already too large for short-run stability. Increased holdings would increase this instability, and they might even “upset the apple cart.” The world must thus go forward between the Scylla of an inadequate amount of reserves and the Charybdis of a destabilizing composition of reserves.


9. The world needs an international organization to create a new type of reserve, one with an international character, to replace all exchange reserves held in national currencies. It needs a kind of reserve which can be expanded appropriately in accordance with increasing production and trade and yet not be overexpanded to create inflation. The job of providing this new kind of reserve in the required amounts should be assigned to an enlarged and amended IMF. It will take time to negotiate such a charter for the XIMF (to create an abbreviation in order to simplify the following discussion). The recent increase of IMF assets to almost $15 billion, following the increase of members’ quotas by more than 50 per cent, is therefore most opportune because it makes time available for such negotiations.

10. The key elements in this new charter would be that members should agree to keep none of their official reserves in the national currencies of other countries (i.e., in the form of demand and time deposits, short-term investments, etc.) and to keep all of their exchange reserves in the form of deposits with the XIMF. A fundamental part of this arrangement would thus be the agreement that members would hold reserves in only two forms, gold and XIMF deposits, and that, with “the possible exception of moderate working balances,”12 they would discontinue holding reserves in the form of currencies of other nations. Members would agree to accept transfers of these deposits in the settlement of their balance of payments accounts.

11. Members would keep a stated percentage of their gross reserves in the form of deposits with XIMF. Gross reserves would be defined as gold, net creditor claims previously accumulated on the IMF (which would be automatically transformed into XIMF deposits), and “other liquid or semi-liquid foreign exchange holdings, i.e., primarily dollar and sterling balances.” 13 If, as suggested, each country would initially be required to keep 20 per cent of its gross reserves in the form of deposits, XIMF would begin operations with about $11 billion of assets, consisting of $5 billion of gold, $3 billion of dollars, $2 billion of sterling, and $1 billion of other assets, based on reserves data as of the end of 1958.

12. Members would have to deposit all official holdings of any national currencies that they did not convert into gold. These additional deposits could be used freely to make all international payments. If countries paid in all of their exchange holdings as of the end of 1958, the Fund’s deposit liabilities would rise from $11 billion to approximately $21 billion. Its gold assets would remain at $5 billion, but its holdings of dollars, sterling, and other foreign exchange would rise to $16 billion.

13. Deposits with the XIMF would be used along with gold and would be “fully equivalent” to it in international settlements. They might be considered as good as gold, or even better, since they would carry an exchange guarantee and earn interest.

They could be drawn upon by their holders to procure any currency needed in such [international] settlements or for stabilization interventions of central banks in the exchange market. The amounts withdrawn would be merely debited from [sic] the withdrawer’s deposit account and credited to the account of the country whose currency has been bought from the Fund.14

Deposits would be usable or convertible in the following sense:

(a) Subject to the limitation that a member could not reduce its deposits below the required percentage level, a member could use all of its deposits to satisfy the claims of other members or to acquire any national currencies for the same purpose. For example, a country with gross reserves of 100, of which 80 was gold and 20 was the required XIMF deposit, could meet a balance of payments deficit with gold, or with gold and XIMF deposits, but deposit funds could not constitute more than 20 per cent of the total payments.

(b) A member could use its “excess” deposits, i.e., deposits in excess of its calculated deposit requirement, to acquire gold in gold markets or from countries willing to buy and sell gold, or from the XIMF. Nevertheless, although countries could formally convert excess deposits into gold in the XIMF, Triffin describes a number of considerations which would make this unlikely, and he outlines various provisions, requirements, and escape clauses which are designed to cope with any large demands for gold.

(c) Deposits that were exactly equal to 20 per cent of a country’s gross reserves (or whatever other percentage was currently in force) could neither be used nor withdrawn. This arrangement implies that exchange balances transferred to the XIMF to satisfy deposit requirements are immobilized until they are released by a fall in gross reserves; and, specifically, while they are immobilized, they cannot be withdrawn or converted into gold.

14. XIMF deposits would enjoy an exchange guarantee. They would “be expressed in a gold unit, and escape therefore the foreign exchange risk always attached to foreign exchange holdings in national currencies.” 15 They would thus possess greater security than reserves kept in the form of dollars, sterling, and other national currencies. They would also earn interest, at a rate which would depend upon XIMF earnings from loans and investments. Correspondingly, the currency and investment holdings of XIMF would carry an exchange guarantee in its favor, in the same way that the IMF present holdings of national currencies are guaranteed against current depreciation.

15. These arrangements would eliminate the instability and danger that lurk in a gold exchange system and that permit runs on one currency in favor of another currency or gold. They would provide a stable structure of international reserves.

16. But it is also necessary to assure that the total of international liquidity should grow in accordance with requirements for it. All Triffin’s discussions of future liquidity requirements are to the effect that the total of international liquidity may have to increase at the same rate as world trade if contractionary forces and exchange crises are to be avoided. The Fund’s lending powers should be limited to the increases necessary to preserve an adequate level of international liquidity.

The IMF lending capacity would be based, as in the Keynes plan, on the accumulation of bancor accounts—in the form of deposits with the IMF—by member countries… . The overall lending capacity of the Fund can properly be limited to the creation of bancor amounts sufficient to preserve an adequate level of international liquidity. Various criteria could be retained for this purpose. The simplest one might be to limit the Fund’s net lending, over any twelve months period, to a total amount which would, together with current increases in the world stock of monetary gold, increase total world reserves by, let us say, 3 to 5 per cent a year. The exact figure could not, of course, be determined scientifically and would, in any case, depend in practice upon the compromise between divergent national viewpoints which would emerge from the negotiation of the new Fund Agreement. A reasonably conservative solution would be to retain a 3 per cent figure as definitely noninflationary, and to require qualified votes (two thirds, three fourths, and ultimately four fifths of the total voting power, or even unanimity) to authorize lending in excess of 3, 4, or 5 per cent a year.16

There are, of course, other guideposts which might be used. Triffin noted that “alternative criteria, more logical but also more difficult to define concretely, might be derived from the current trend of some international price index reflecting inflationary or deflationary pressures on the world economy.” 17

17. The XIMF would be empowered to create new deposits based on loans to, or investments in, its member countries. Loans to members would be analogous to drawings from the IMF, but they would be divorced from what is termed the artificial and arbitrary regimentation of Fund quotas.

The Fund’s lending operations, moreover, should be no more automatic than they are at present, and this discretion should enable it to exercise a considerable influence upon members to restrain internal inflationary abuses.

The experience acquired in the 12 years of operation of the Fund is extremely valuable in this respect. Fund advances should continue to require full agreement between the Fund and the member with relation not only to the maturity of the loan, but also to the broad economic and financial policies followed by the member to insure long run equilibrium in its international transactions without excessive recourse to trade and exchange controls.18

Indeed, “the normal procedures for Fund advances need not differ substantially from those gradually developed by the Fund over its twelve years of existence.”19

18. Stand-by arrangements, which have been used extensively by the Fund, would continue to be used by the XIMF, but they might

be supplemented by overdraft agreements, to be renewed at frequent intervals, and guaranteeing all members in good standing rapid and automatic Fund assistance in case of need, but for modest amounts and with short-term repayment provisions. These overdraft agreements would be primarily designed to give time for full consideration of a request for normal, medium-term, loans or stand-by agreements, and would be guaranteed by the country’s minimum deposit obligation.20

19. The XIMF could also make investments on its own initiative, but with the consent of the government concerned. These investments could take the form of the purchase of short-term or long-term government securities against XIMF deposits. Investments should be made in ways that will facilitate economic development. They should be made in countries that need development capital or to organizations (such as the International Bank for Reconstruction and Development) which make loans for economic development. Triffin also mentions the possibility that loans might be made to private business enterprises engaged in development. All investments would be subject to an exchange guarantee. Thus, the currency and investment assets of XIMF, as well as its deposits, would be guaranteed against currency depreciation.

20. Practically all exchange reserves at the present time are held in the form of dollars and sterling. Triffin considers such a pattern of investment economically unjustifiable. Sterling and dollar balances can be built up only by capital exports to the United Kingdom and the United States, which are already creditors and capital exporters. Short-term capital imports by these countries increase the amount of capital which they must export on long term in order to maintain balance of payments equilibrium for themselves and the rest of the world. The present practice increases the capital exporting problems of these creditor countries and is equivalent, in a phrase which is used again and again, to carrying coals to Newcastle. Triffin proposes that the process of building these balances be first halted, and then reversed, and that international short-term investments in dollars and sterling be liquidated over a period of time at, say, a maximum rate of 5 per cent a year. The proceeds should be invested in countries that really need capital imports.

21. In conclusion, Triffin proposes to give XIMF the power to create a new kind of money—an international money and an internationalized money—and to expand the quantity of this money in accordance with international liquidity requirements as determined by its members. Gold would remain a component of international reserves, and the XIMF could create new deposits according to the principle that the rate of new gold additions to monetary reserves plus additional deposits created by the XIMF should not exceed the rate of increase of trade, and “a resonably conservative solution would be to retain a 3 per cent figure as definitely noninflationary.…”21 His proposal would make XIMF a central bank for central banks, with the responsibility and the authority to do for central banks internationally what central banks now do for commercial banks nationally.22 The net result of his proposals would be to create a system that could be rationally managed to stabilize the structure of international reserves and to expand reserves in accordance with need.

II. Comment

22. Triffin’s proposals appear to be designed to provide, at little or n cost, benefits for everybody. It will be useful to summarize these proposed benefits, as follows: (a) The world would benefit by obtaining, at long last, a currency both international and internationalized, described as being beyond question in quality, elastic in quantity, and responsive to the needs of trade. “Hot” money would be cooled off, and runs on currencies eliminated, (b) The United States would benefit because the possibility of converting official dollar balances into gold would be reduced, if not eliminated. The United States would be free to set interest rate policy without fearing repercussions via the outflow of official dollar balances and their conversion into gold or other currencies, and it would know that a payments deficit was no longer needed to increase international liquidity, (c) The United Kingdom would benefit because it would be freed from the specter of a flight from sterling into gold or dollars, and from the necessity of subordinating an interest rate policy appropriate to its domestic needs to the defense of the external position of sterling, (d) Holders of dollar, sterling, and other currency balances would be able to exchange their holdings for Fund deposits, which would be of irreproachable quality because of an exchange guarantee, and which would probably earn interest into the bargain. (e) Less developed countries would benefit because the existence of the expanded Fund would make it unnecessary for them to transfer short-term capital to the United States or the United Kingdom in order to obtain exchange reserves. Instead, the flow could be reversed. A large part of the assets behind XIMF deposits could be invested directly, or through other agencies, to finance economic development, thus moving capital to the countries that need it most, (f) Even gold mining countries and the gold mining industry may be said to benefit in a backhanded way. Though gold as the basis for international reserves is held to be unnecessary, restrictive, and a waste of economic resources, it is nevertheless retained in the system, after its dangers have been neutralized, as a harmless example of cultural lag and as an inexpensive subsidy for the production of something that the world wants but does not need.

23. The following discussion attempts to determine whether Triffin’s proposals are necessary or appropriate to realize all these benefits. It considers whether present and prospective dangers are as great as described. It inquires whether existing financial arrangements and institutions are as inadequate as alleged to cope with present conditions as well as those that are likely to develop. It examines the effect on the assets and liabilities of the Fund of the implementation of Triffin’s proposals, both before and as a result of the expansion said to be needed during the next few years. Finally, it considers what fundamental policy changes are required for the proposed international financial mechanism, including further transfers of authority by countries to international agencies, additional responsibilities assumed by countries themselves, and new operating objectives of the expanded Fund.

Is the Structure of International Reserves Unsatisfactory?

24. The first claim of the Triffin plan is that it will result in a more stable structure of reserves. It is proposed first to examine the present structure, in order to evaluate its advantages, disadvantages, dangers, and costs.

25. U.S. liquidity is very high in terms of the ratio of gold to official dollar balances. Even a large-scale conversion of official balances would have moderate effects, in view of the size of the balances and the need to maintain at least part of them for working purposes. The problem of liquidity really arises when account is taken of other present or potential liabilities. Banking and all other private dollar balances totaled $6 billion at the end of 1958; a change in leads and lags applied to foreign trade of $35 billion would amount to additional billions; repatriation of foreign funds invested in U.S. securities would amount to many more billions; and speculation against the dollar financed by funds borrowed by both nationals and foreigners would provide additional billions. All of these potentialities have increased since 1958, and banking and all other private balances alone reached $7.5 billion in June 1960.

26. At the end of 1958, short-term dollar liabilities totaled $14.6 billion, of which $8.7 billion was in official hands and $5.9 billion in the hands of banks and other private holders. On the same date, sterling liabilities23 amounted to $9.4 billion, of which $6.7 billion was in official, and $2.7 billion in nonofficial, hands. Nonofficial holdings of dollars and sterling therefore totaled $8.7 billion, while official holdings totaled $15.4 billion.24 Though Triffin’s proposals refer to all official reserves, he recognizes that part of these will have to be held in liquid form. Countries

have, in any case, to retain some working balances in a form other than gold in order to avoid repeated gold sales or purchases each time they wish to sell or buy [sic] foreign currencies in the market to stabilize their own exchange rate. These working balances will have to be held either as excess deposits with the Fund or directly in the key currencies actively traded on the exchange markets. Either of these two alternatives would reduce substantially the danger of an excessive gold drain from the Fund. Working balances equal to only 5 per cent of annual imports, for instance, would absorb as much as $5 billion.25

Not all of the official holdings of exchange therefore involve the threat of conversion into gold.

Private holdings are, however, considerably more mobile than official ones. They are more responsive to interest rate differentials and speculative possibilities. Private dollar holdings cannot be converted into gold in the United States, but they can be used to buy gold abroad, or to buy any other key currencies at home or abroad. Either use could result in an outflow of gold. If nonofficial holders of dollars should switch to sterling to take advantage of higher interest rates in the United Kingdom, the United States would have to support the dollar-sterling exchange rate by selling gold, unless the United Kingdom was willing to hold these dollars in the form of additional XIMF deposits or working balances. Similar considerations would apply for the United Kingdom and purchases of dollars with external sterling.

27. Under present arrangements, the United States could support the dollar with its gold reserves, and it could also draw other key currencies from the IMF, consistent with IMF policy. To draw these currencies the United States would pay dollars to the IMF; it would use these other currencies to buy dollars in the foreign exchange markets. The United States would thus relieve the pressure on the dollar, whether this stemmed from official or nonofficial sources.

The ownership of dollars would be transferred from their present holders to the IMF, and spot liabilities would be converted into term liabilities. The countries whose currencies were drawn from the Fund would automatically acquire credit balances, if it is assumed that the IMF initially held their currencies to the extent of only 75 per cent of their quotas, or lower debit balances, if it is assumed that the IMF held more than 75 per cent of their quotas. By using drawings to help counter pressure on the dollar, the United States would slow down the decrease in its gold holdings. Its gross reserve position would therefore be better than if it met all of the pressure on the dollar by selling gold. The effect upon its net reserve position would depend upon what liabilities were considered as contra items against gold reserves. Very short-term dollar liabilities—dollars being offered in the foreign exchange market—would be changed into dollars that had to be repurchased in three to five years (or, if there were no specific repurchase agreement, as gold reserves increased). The spot net position of the dollar would be improved immediately, while its five-year net position would remain unchanged.

28. The case for a full and open-end guarantee of dollar and sterling balances, given a realistic appraisal of present and prospective dangers to these currencies, and the resources that could be used to meet these dangers, must be considered as not proved. Perhaps some lesser kind of assurance may be useful, though even this is not clear. But if this is the case, only a few countries need be involved, since dollar and sterling balances, official and private, are highly concentrated.

29. In all probability, the risks of exchange instability have been overstated by Triffin, whose fears for the future of the dollar are colored by the large balance of payments deficits of the United States since 1958, deficits which must in any event be reduced or eliminated. On the other hand, the ability of the key currency countries to meet these risks has been understated. The resources of the Fund, supplemented by country credits and supported by appropriate domestic measures, were quite sufficient to halt heavy drains on sterling in 1956 and 1957. With much larger resources and with a larger assortment of convertible currencies, the Fund would be able to help even more in any future crisis. The United States has never drawn on the Fund, but there is no reason why it should not under conditions and for purposes consistent with Fund policy. The United States at the end of 1960 had a net creditor position in the Fund of $1.6 billion; after exhausting this, it could draw an additional $4.1 billion before Fund holdings of dollars would reach 200 per cent of quota. If the Fund needs more resources, it may obtain these by borrowing.26 It is quite unlikely that sales of foreign currencies against dollars, up to as much as $6 billion, added to U.S. sales of its own gold, coupled with the determined attitude that such actions would proclaim, could fail to halt a run on the dollar.

Is the XIMF Needed to Expand International Liquidity?

30. The other leg of the dilemma described by Triffin is that the prospective growth of reserves will fall short of the requirements associated with the growth of trade. This shortfall over and above additions to gold reserves is calculated for the next decade and is declared to be far in excess of any safe expansion of dollar and other exchange reserve holdings. This conclusion rests principally upon Triffin’s estimate of reserve requirements, since there is general agreement about prospective gold additions to monetary reserves.27

31. Triffin’s view of the required growth of international reserves is essentially a mechanistic one based on a rough relationship of reserves to imports;28 yet it is simply not true that the need for monetary reserves can be read off arithmetically from a table relating reserves to trade. The fact that this ratio—or more accurately, this family of ratios, since both trade and reserves can be thought of in many different ways—has been falling in the postwar period is no proof that reserves have been becoming less and less adequate or that they have reached some dangerous minimum. It is questionable that the level of reserves in 1957 or in 1958, the years used by Triffin as bases for reserve calculations, represented the minimum reserves required by the noncommunist world, and that in the future these reserves must grow as fast as world trade.

32. Equally, there is no reason to believe that in any period, and particularly in the next ten years, the demand for reserves will grow at the same rate as trade.

The demand for reserves is a result of the policy decisions by a relatively small number of countries. There was a great, but highly concentrated, demand during the 1950’s. In this period, the major part of the increase in reserves was for the account of Germany, Austria, Italy, and Japan, countries which had been stripped of reserves by the war and had a problem of rebuilding them. If the demands for reserves for this particular purpose had not existed, dollar balances would in all probability not have grown as much as they did. If these countries (and a few others) had not been willing to add to their assets in the form of money (gold or dollars) rather than goods, the large and continuing deficits in the U.S. balance of payments in 1950–59 would not have occurred. The accumulation of reserves for this purpose must sooner or later come to an end, with important consequences for the future demand. Some of these countries have reserves which will be in excess of their needs for some years to come; thus they can be sources of reserves for other countries.

Many of the less developed countries, such as India, accumulated large reserves during World War II—only because they could not buy goods. These countries have already spent most or all of these reserves and show little sign of wishing to recapture reserve levels of ten and fifteen years ago. If appropriate reserves figures were available, some countries would show a negative net reserve position. The reserves of the less developed countries were remarkably stable during the decade of the 1950’s, though this stability was the arithmetic result of a number of differing trends. The reserves of oil producing countries (with the total largely influenced by Venezuela) increased until 1957 and declined sharply thereafter; those of the U.K. colonies increased steadily, doubling from 1950 to 1954 but increasing by only 10 per cent in the rest of the decade; and for all of the other less developed countries (with the total influenced largely by India), reserves decreased during the ten-year period.

Gold reserves of the United States have decreased by $7 billion from their all time peak of $24.6 billion at the end of 1949; but even so, they are at the level immediately preceding World War II. Reserves of the United Kingdom at the end of 1960 were $3.2 billion, or 6 per cent less than reserves at the end of 1950 and 20 per cent below those at mid-1951, the high for the decade.

33. The trend in gross reserves since World War II is the resultant of these and many other factors, and the future movement is not predictable on a statistical basis related to the growth of trade. Furthermore, the adequacy of the reserve position of any country is influenced not only by its gross reserves, but also by its debtor or creditor position in the Fund, the size of its unused Fund quota, the amount of its outstanding and overdue short-term commercial debt, official short-term borrowings, gold pledges, banks’ holdings and other private holdings of exchange, and the like. Failure to take such factors into account, for example, leads Triffin to classify Argentina in 1953–55 in his group of high reserve countries.29 Countries that are in debt can, in the future, improve their real reserve positions by reducing their obligations rather than by increasing their gross reserves. This would not create any demand for reserves, and the world total of reserves would not need to grow, although it would become more adequate to meet demands. Finally, there is little reason to assume, as Triffin does, that the United Kingdom is actively pursuing policies compatible with the reserve target of $5 billion that has been talked about for a number of years.30

34. There are other difficulties in determining reserve requirements by applying a growth-of-trade factor to a base of present international reserves.

For example, in his Wicksell Lecture (1958), Triffin said that “world trade and production are estimated to have increased in volume, over the last ten years, at a pace of 6 per cent a year. A parallel rate of increase in the world’s monetary reserves ($62 billion at the end of last year) would require their expansion by about $3.7 billion annually.” 31 This figure of $62 billion includes all the assets of the IMF (gold, convertible currencies, and inconvertible currencies), as well as country holdings of inconvertible currencies and balances under payments agreements. Would this imply that there is some requirement to expand reserves of inconvertible currencies, or to expand a total of which some components are much less useful than others? If a dollar’s worth of inconvertible reserves becomes convertible, does it not become more efficient?

A similar kind of question is posed by the treatment of claims on the European Payments Union (EPU). The $62 billion total of world reserves referred to in the Wicksell Lecture included claims on EPU totaling $1.3 billion. At a compound growth rate of 6 per cent a year, this inclusion created a demand for $1 billion of reserves over the next decade. But when claims on EPU were funded at the end of 1958, they vanished from the statistics on reserves. Triffin’s calculations in 1959 of the required rate of growth of reserves for the next decade were reduced correspondingly. The reason for excluding these bilateral claims from reserves statistics was presumably that they represented longer-term assets; many of these funded claims had terms of a year or two, and a few had terms as long as seven or ten years. Germany was the largest net claimant on EPU and was interested in showing lower, rather than higher, reserves. Moreover, Germany’s net bilateral claims of $1 billion in January 1959 had been reduced to $480 million by October 1960. The arbitrary nature of the classification of reserves is suggested by the subsequent rate of payoff as well as by a comparison with overseas holdings of sterling. At the end of 1956, approximately one third of sterling liabilities had maturities of more than five years; yet they were included in country reserves.

The currency holdings of the IMF present another problem in the calculation of requirements. Reserve requirements in 1958 were calculated by applying trade growth factors to a world total of reserves that included the currency holdings of the IMF. Since then, member quotas in the IMF have been increased by more than 50 per cent, and IMF holdings of currency enlarged by more than $4 billion. How, then, should requirements for 1960 be calculated? Should the procedure be the same as for 1958? Should the requirements for 1960–70 include an element for expanding this additional $4 billion at the assumed rate of growth of trade? At a 3 per cent rate, this involves a reserve requirement of almost $1.5 billion during the decade. Or should reserve requirements be calculated on the basis of 1960 figures that exclude this $4 billion, which in turn might be treated as a deduction from calculated requirements?

As these examples suggest, to calculate future reserve requirements by applying figures on the growth of trade to published figures on reserves may lead to results that vary considerably from one date to another. Arithmetic calculations of reserve requirements in relation to trade are simple—this would appear to be their outstanding virtue—but they are not necessarily convincing.

35. These comments should not be interpreted to mean that, because there is question about Triffin’s rate of 3 per cent, or 5 per cent, or 6 per cent, some other rate is necessarily better. On the contrary, the search for any rate is vain. No mechanistic solution will serve. There are many roads to liquidity, many ways of increasing it, and many ways of judging when it is more or less adequate. The activities of the IMF, and the recent increase in its resources, affect the demand for reserves; also, the contracyclical activities of the industrial countries affect the demand.

An appropriate and expanding volume of world trade may be carried on with many different amounts of reserves, total reserves of many different compositions, and many different rates of growth of reserves. Any arbitrary determination that reserves should increase at some given rate may well lead to a lowering of credit standards and to inflation. In the end, stability of the reserve structure can be obtained only if the major countries really want it, and if they have access to credit to preserve it. Bagehot’s doctrine that the Bank of England should be prepared always to be the lender of last resort has its counterpart in the international field. Access to IMF drawings and stand-bys has formalized and internationalized the granting of credit. This marks a great advance over the cooperation among central banks that characterized the period of the 1920’s and has made it much more unlikely that a crack in the exchange structure can start, or, having started, that it can continue.

Reserve Structure Proposed in the Triffin Plan

36. The foregoing discussion has shown that the existing structure of international liquidity is less vulnerable than Triffin’s diagnosis suggests. Let us turn now to a consideration of his remedies for its weaknesses. He has claimed, as noted earlier, that the Triffin plan will result in a more stable structure. This proposition is based upon the transformation of the IMF into a central bank to hold, in the form of deposits, part or all of the reserves of its members. This power is separable from the power to create deposits, a function also proposed for the IMF. The latter is based upon the assumption that the XIMF and the central banks of all member countries constitute a closed banking system, essentially similar to the central bank and commercial banks within any one country.

37. As a prelude to this discussion, it should be noted that adoption of Triffin’s proposals would make no difference to the serious problem associated with the present balance of payments deficit of the United States and the recent increase in gold speculation and hoarding. On the surface, uneasiness about the dollar reflects the present state of reserves, liabilities, and rate of outflow of gold. But basically this uneasiness reflects trends. It reflects the series of large deficits which continue to reduce gold reserves and increase short-term dollar liabilities. Judgment about the adequacy of the reserve position of the United States and the strength of the dollar is highly colored—and correctly so—by fear that these trends will persist, by fear that the balance of payments deficit cannot or will not be eliminated.

Adoption of Triffin’s plan will not eliminate this balance of payments deficit. It will not balance any country’s external accounts. It should not relieve any country from keeping its balance of payments in order. Adoption of his proposal would probably reduce the proportion of the deficit currently settled in gold, but it would not affect the size of the deficit itself. To the extent that such a reduction of gold flows moderated present uncertainties and was considered as extending the time available for corrective measures, it might even result in a larger total deficit. Recent increases in gold speculation basically reflect the same set of problems. Speculation is not against the dollar in favor of other currencies; it is based on fear that the dollar—as well as other currencies—will be devalued. This is not a problem of a shortage of international liquidity. It is a problem arising out of the U.S. balance of payments.

38. To illustrate how the proposed arrangements would operate, let us apply the proposals to the data for the end of 1958, which were used by Triffin.32 (The use of later data would not materially affect this analysis.) At the end of 1958, the United States had gross reserves of $22.5 billion, consisting of $20.6 billion of gold and $1.9 billion of net claims on IMF; its deposit requirement with the XIMF, at 20 per cent, would be $4.5 billion and would require an additional payment of $2.6 billion to the XIMF. Since the United States does not include exchange assets in its gross reserves, this deposit requirement would be met by paying gold. All other member countries would have deposit requirements of $6.7 billion. On the reasonable assumption used by Triffin “that all countries would initially prefer to hold onto their gold assets,” 33 the member countries would meet their deposit requirements with $600 million of present claims on the IMF, $800 million of gold, and $5.3 billion of exchange. The exchange payments probably would consist of approximately $1.9 billion of sterling and $3.4 billion of dollars. Dollar balances held officially by all countries would be reduced from something over $9 billion (including bonds and notes) to $6 billion. As far as the United States is concerned, these payments of $3.4 billion of dollars would be rendered permanently inconvertible into gold unless the gross reserves of the rest of the world should fall below their total at the end of 1958, i.e., $33.7 billion. Such a development was very unlikely even in 1958, and it has become even more unlikely with the subsequent increase in reserves.

39. The effect so far upon the net reserve position of the United States would be the same as if the United States had redeemed $3.4 billion of dollar balances and paid out $2.6 billion of gold. But these transactions are only a first step. Under Triffin’s proposals, all remaining official exchange balances would have to be deposited in the XIMF or converted into gold. Therefore, the $6 billion of dollar balances and more than $3 billion of sterling balances left after the required transfer to the XIMF of 20 per cent of gross reserves, which may be termed “excess” dollar balances, would have to be turned into gold or deposited at the XIMF. To the extent that countries chose this opportunity to buy gold which they might be unwilling to acquire under existing arrangements, they would create the pressure on the dollar and the outflow of gold that Triffin fears so much. In such circumstances, the help that could be provided by the XIMF to the United States to relieve the pressure would be no greater than the help that could be provided by the present IMF. Triffin is probably right, however, in thinking that XIMF deposits could be given such advantages over present exchange reserves in key currencies, notably an exchange guarantee and the prospect of earning interest, that the conversion would be made smoothly, with little additional demand for gold.

XIMF Deposits, Gold, and Exchange Guarantees

40. The XIMF would therefore begin operations, based on the data at the end of 1958, with $21 billion of deposits, of which $11 billion was required, in accordance with the 20 per cent deposit ratio, and the remaining $10 billion was “excess” in the sense that it was over and above the deposit requirement. The assets corresponding to these deposit liabilities would be $5 billion of gold and $16 billion of exchange; the latter would consist of more than $9 billion of dollars, more than $5 billion of sterling, and $1 billion of other exchange assets. (This ratio of gold to deposits subject to conversion is lower than the ratio of U.S. gold holdings, whether these are defined as total gold holdings or only those in excess of legal reserve requirements, to balances held officially by foreigners.)

41. Members could at any time use their “excess” deposit balances to buy gold or to convert them into gold through the XIMF. Thus, expanding the functions of the IMF would not necessarily rule out future switches into gold. If members should attempt to convert “excess” deposits into gold on a substantial scale, the XIMF would then have to make gold harder to get, call for more gold from its members, or both. It could raise the level of required deposits to 25 per cent or 30 per cent of reserves, or to even higher percentages. The higher deposit requirements would reduce the amount of “excess” deposits subject to conversion and simultaneously make the high gold reserve countries pay in more gold. For example, an increase of 5 per cent in the deposit requirement would call for $1 billion in gold from the United States. It would even be possible, as Triffin suggests, to make the gold reserve countries finance the XIMF gold position to a larger extent by imposing “higher deposit requirements upon that portion of each member’s reserves which exceeds the average ratio of world monetary gold to world imports.” 34 Thus, a desire to convert “excess” dollars into gold rather than to deposit them in the XIMF would lead to a large gold drain from the United States, while an attempt at conversion of the same sums after they had once been deposited in the XIMF could be countered—to the extent that it was not financed by drawing gold from the United States—only by making a larger proportion of XIMF deposit balances inconvertible into gold.

42. Before pursuing the implications of these alternatives, we may follow through the development of XIMF over the next ten years, as contemplated in the Triffin plan. We have seen35 that he suggests that the Fund’s net lending should be limited “over any twelve months period, to a total amount which would, together with current increases in the world stock of monetary gold, increase total world reserves by, let us say, 3 to 5 per cent a year.” If it is assumed that monetary gold stocks increase at the rate of $700 million or $800 million a year, the annual increase in XIMF deposits would be about $800 million based upon a 3 per cent rate, $1.4 billion at a 4 per cent rate, and $2 billion at a 5 per cent rate. Triffin comments that “These estimates would rise gradually, but slowly, with further increases in world reserves. They could decrease as well as increase, on the other hand, with future fluctuations in the current additions to the world monetary gold stock.”36

43. How are annual net increases of such a size to be achieved by the expanded Fund?

It is unlikely that they could be significantly achieved by increasing short- and medium-term loans to member countries, for there is no suggestion that the lending policy of the IMF should be changed. Indeed, “the normal procedures for Fund advances need not differ substantially from those gradually developed by the Fund over its twelve years of existence.” 37 The XIMF would continue to make short-and medium-term loans to debtor and creditor countries to meet temporary balance of payments difficulties. But it is highly doubtful that the net increase in deposits, calculated as being required, or as being obviously noninflationary, could be achieved by self-liquidating loans to meet temporary balance of payments difficulties. Unchanged lending procedures should not result in a larger net volume of lending after an expansion of the functions of the IMF.

The major part, if not all, of the net increase in deposits would necessarily depend upon investments by the XIMF, and these would have to be medium- and long-term investments. The criteria for these investments would have to include, as a major element, the desired increase in international liquidity. As a matter of policy, they could not be made to any substantial extent in creditor countries that were already exporters of capital. Sending capital to them would merely increase the size of their capital-exporting activities—carry coals to Newcastle. On the contrary, the investments would have to be made in countries that were net importers of capital, that is, less developed countries. This prospect would obviously have great political appeal for such countries.

44. It is assumed that monetary stocks of gold in the hands of countries would increase during the next decade by $7–8 billion; thus about the same amount of additional international liquidity in the form of XIMF deposits would have to be provided by the Fund. Accordingly, Fund deposits would increase by some $8 billion in the next decade, and in the process its total assets would increase to $29 billion. Of this total, perhaps $15 billion would represent required deposits, based upon a 20 per cent factor, while the remaining $14 billion would constitute “excess” deposits, with respect to which convertibility into gold might be requested. Under the assumptions already discussed, gold holdings of the XIMF would increase little, if at all, since exchange would be deposited to the maximum extent possible to satisfy deposit requirements. However, the quality of Fund deposits, and the liquidity of the Fund as measured by gold, could, of course, be increased by raising reserve requirements. This would result in a substantial increase of the Fund’s gold holdings; at the same time, the legal convertibility potentialities represented by Fund deposits would be reduced.

The $8 billion increase of Fund assets would consist of foreign exchange and securities: loans to members, direct investments in member countries, and investments in the securities of organizations such as the International Bank for Reconstruction and Development. These investments would be made on the initiative of the XIMF and would also be agreed by the monetary authorities of the country concerned. Such agreement would be necessary to secure for the investments a guarantee against exchange rate changes and inconvertibility risks.

The $8 billion of additional investments in the next decade would be in addition to the $16 billion of assets (almost wholly dollar and sterling balances) with which the XIMF began operations. The gradual change in the composition of XIMF assets during this period, and the probable structure of its assets at the end of the period, are matters of considerable interest.

45. The XIMF would have no immediate need to modify its initial investments, i.e., the dollars and sterling initially deposited with it, but “should be empowered to do so, in a smooth and progressive manner, insofar as useful for the conduct of its own operations. This purpose would be served by giving the Fund an option to liquidate such investments at a maximum pace of, let us say, 5 per cent a year.” 38 This option would be noncumulative. This percentage applied to initial XIMF dollar and sterling holdings of $15 billion would correspond to an amortization requirement for the United States and the United Kingdom of $750 million a year. These countries could buy back the securities represented by this amortization requirement by paying gold to the XIMF or by transferring XIMF deposits, i.e., by reducing their reserves or by increasing their balance of payments surpluses. Amortization would not necessarily have to be undertaken at once, nor would it necessarily be required every year. On the contrary, if the United States or the United Kingdom should have balance of payments deficits in the early years of the XIMF, holdings of dollar and sterling assets might even increase. But the logic and the language of the Triffin proposals—the many animadversions on the policy of carrying coals to Newcastle—require a persistent, though orderly, liquidation of dollar and sterling balances.

The resources derived from their progressive liquidation, however, would normally be reemployed in other markets whose need for international capital is greater than in the United States and the United Kingdom. A portion of such investments might even be channelled into relatively long-term investments for economic development through purchases of IBRD bonds or other securities of a similar character.39

46. If this policy were carried out at a rate of 5 per cent a year, the $15 billion of sterling and dollars initially turned over to the XIMF would be reduced to perhaps $8 billion at the end of a decade. Moreover, a policy of amortizing initial dollar and sterling assets would imply that new investments in such assets should be avoided. Instead, investments would be made in the securities of other countries. Investments in less developed countries would increase even more markedly than would otherwise be necessary. Some investments might be made through intermediaries (such as the International Bank for Reconstruction and Development) and some by encouraging the financial centers in Europe (Brussels, Frankfurt, etc.) to borrow on short term and lend on long term.40 The change in the XIMF portfolio of assets may, in broad categories, be outlined as shown in Table 1. Thus, in a decade, gold holdings would be reduced from one fourth to one sixth of total assets. Obligations of less developed countries might equal more than half of the total assets of the XIMF. A substantial part of these obligations would necessarily be direct, and on medium and long term to the countries themselves; some would be obligations of international agencies.

Table 1.

Assets of XIMF, 1958 and 1968

(In billions of U.S. dollars)

article image

47. The change in the character of the structure of international reserves to be accomplished by the XIMF is outlined by this numerical illustration. The present structure consists of gold and short-term obligations of net creditor countries. The proposed structure will consist of gold and XIMF deposits; and the deposits will increasingly rest upon the medium- and long-term obligations of the net debtor countries. (If the reserve situation in 1960 had been taken as a starting point, XIMF assets and deposits would be larger than shown in the table, and the composition of assets would be somewhat different. Gold would be a smaller, and dollars a larger, proportion of total assets. The point of the calculation would, however, remain unchanged.)

48. If this transformation is not to impair member countries’ confidence in the soundness of XIMF deposits, it would be essential to reinforce these deposits by appropriate guarantees. Such guarantees play a crucial role in Triffin’s proposals. All XIMF deposits would carry an exchange guarantee, expressed in a gold unit of account. If the price of gold were raised, the amount of XIMF deposits standing to the credit of each member would be increased correspondingly. If any currency were devalued relative to gold—if the gold content of any currency were reduced while the world price of gold remained the same—an XIMF deposit would then buy more units of that currency. Such arrangements involve corresponding guarantees of XIMF assets and are the same as those applicable to IMF holdings of members’ currencies under Article IV, Section 8, of the IMF Articles of Agreement. Under this provision, each member is required to maintain the gold value of the IMF holdings of its currency. If the par value of its currency declines, a member must pay to the IMF additional currency sufficient to restore the original gold value of the Fund’s holding. This provision is uniform for all members. It applies to members under Article XIV and whose currencies are not convertible within the meaning of the Articles of Agreement; it applies to members under Article VIII and whose currencies are considered to be convertible. It applies whether a member does, or does not, stand ready to buy and sell gold for official purposes at stated prices.

49. The significance of these guarantees would change, however, as the XIMF operations proceeded. With the passage of time, an increasing proportion of its assets would consist of direct obligations of less developed countries, and obligations of investing intermediaries, such as the International Bank for Reconstruction and Development (IBRD), “older financial centers” in Europe, and the like. Whereas XIMF’s investments at the beginning would be in key currencies which are widely used and which are completely convertible on external account, investments later would be in national currencies having no international circulation and which may be subject to more or fewer exchange restrictions. Though all investments would be guaranteed against devaluation, additional quantities of various currencies paid in to compensate for devaluation would hardly have the same acceptability.

There are difficulties, also, in the concept of guarantees by the investment intermediaries. Bonds issued by the IBRD carry neither an exchange nor a gold guarantee. The bonds are issued in particular currencies and are repayable in those currencies on a legal tender basis, regardless of what their gold or exchange equivalent may be at the time of repayment. Members of the IBRD guarantee to maintain the value of their capital subscriptions in terms of a gold equivalent, but the IBRD itself does not issue bonds with a gold equivalent. Nor is it clear why, under these arrangements, the older financial centers would be prepared to lend long and borrow short. They have traditionally done this in the past, but they did not have to operate then with one-sided guarantees. It seems improbable that they would be content with the arrangements proposed by Triffin, under which their obligations to the XIMF would have to be guaranteed against exchange depreciation whereas their foreign investments would be subject to all the risks of devaluation and restrictions on exchange and capital movements.41

But even if it is assumed that guarantees of the kind envisaged could be secured by the XIMF, an important question remains: Would not such a prospective change in the quality of its assets, first, encourage members to keep their excess reserves in gold rather than in XIMF deposits, and second, make it more difficult to secure agreement on an increase in reserve requirements? To the extent that it had these effects, it would limit XIMF banking operations and reintroduce elements of instability resulting from sizable gold flows.

50. An alternative to these far-reaching proposals is that, if the United States and the United Kingdom wanted to, and could, avoid any threat to the dollar and the pound by a comprehensive guarantee, such a guarantee could be negotiated specifically to meet the degree of danger apprehended. It could apply to all countries or to only a few; it could apply to all official reserves or to only that part in excess of regular working balances; it could apply to existing balances or to additions. The Triffin proposals offer no such flexibility and are much more extensive. They formally apply only to official holdings of exchange. In the event of a run, however, nonofficial holdings are quickly converted into official holdings. Moreover, official holdings may come to include speculative flights from a currency financed by bank credit. The exchange guarantee under the Triffin proposals is thus likely to be more far-reaching than any exchange guarantee that might be negotiated by any individual country acting alone.

Even after the IMF was expanded to hold the exchange reserves of its members, private exchange holdings would remain outstanding and sensitive to profit opportunities. Interest rate policies that might be pursued by the United States, the United Kingdom, or any other key currency country would continue to affect short-term capital movements. To this extent, a differentially low interest rate adopted for domestic reasons might still be put under pressure by the outflow of short-term funds. To argue that such developments would in no instance result in the outflow of gold would be to assume that countries would be willing to accumulate XIMF deposits without limit and that they would abandon their views with respect to maintaining conventional proportions of their official reserves in gold.

51. There are, in fact, only two ways to eliminate the threat of a demand for gold. One is to maintain a 100 per cent gold reserve for deposits. The other, as Keynes recognized, is to establish one-way convertibility, so that gold can be used to buy “bancor,” but “bancor” cannot be used to buy gold.42 A fractional gold reserve system is always subject to disturbances, and both its proximate and its ultimate stability depend upon confidence. Confidence will, in turn, reflect the opinion that the system, whether it be a gold-exchange standard or the XIMF, will be able to honor its commitments.

Furthermore, the acceptability of exchange assets (even with a country guarantee) depends upon maintaining balance of payments discipline, which limits the quantity of national currency paid out to foreigners. Gold flows may be disturbing, but they are a rude disciplinarian. Would the XIMF do as effective a disciplinary job? And could it do such a job without acting in much the same way as gold flows do? There may well be a conflict between the requirements for stability of the exchange structure in the short run and in the long run, and particularly if one raison d’être for an expanded IMF is to increase the growth of international liquidity.

52. There is no doubt that the calculated net increases in deposits could be obtained by a Fund whose objective would be to expand international liquidity (the money supply) even if it had to acquire large amounts of long-term obligations to do so. But three points should be noted. First, adherence to a lending policy based upon self-liquidating short-term borrowings will not necessarily expand the money supply at any predetermined rate, while adherence to a policy of expanding the money supply at a predetermined rate can hardly be based upon short-run borrowings. Second, the amount of additional investment required to achieve a given expansion of XIMF deposits is not necessarily in the ratio of 1:1. It may be less than this ratio if the key currency countries have balance of payments deficits. It may be more than this ratio if they are using balance of payments surpluses to reduce currency liabilities to foreigners—specifically, dollar and sterling assets held by the XIMF. This would be the situation whenever the XIMF exercised its option to have the United States, or the United Kingdom, amortize its liabilities to the XIMF at the rate of (say) 5 per cent a year. Finally, central banks have never been willing to adopt a policy of consistently and automatically expanding the domestic money supply in line with production or national income; they have invariably used a combination of criteria; and they have sometimes considered that increases in the money supply are not required for high levels of economic activity. An expanded IMF with responsibilities to attain the international liquidity that Triffin considers adequate would be put in the position of attempting to do internationally what no central bank has been willing to attempt nationally.

International Liquidity Since 1956

53. To put Triffin’s proposals in better perspective, and to see what they could imply for the international management of financial operations, a few comments may be in order on the course of international liquidity and on the use of IMF resources since 1956. It would not be unreasonable to expect, as a first hypothesis, that an inadequate rate of growth of international liquidity would be accompanied (with perhaps some lag) by one or more of the following: increased use of IMF resources, pressures upon world prices and trade, deflationary policies to gain reserves, and increased resort to impediments to trade. On the other hand, it would be difficult to conclude that, if most of these indications were not present, international liquidity was growing at an inadequate rate. The three-year period beginning with 1957 has been chosen for analysis partly because Triffin suggested a version of his present proposals in 1957, in his Europe and the Money Muddle, and partly because 1957 and 1959 appear to represent periods in the business cycle which are not too dissimilar; but there is no question that a longer period would have advantages. In commenting upon international liquidity and the effect of Fund operations, it may be noted that drawings on the IMF are statistically equivalent to increases in international liquidity, and that repayments are equivalent to decreases; however, liabilities (repurchase obligations) resulting from drawings are not deducted from statistics on country reserves.

54. In the three years 1957–59, the volume of world imports expanded at an average rate of 4.3 per cent a year. Imports of developed areas increased at a rate of 5.4 per cent a year, and those of less developed areas at the rate of 1.3 per cent.43 The rates of increase are considerably greater if imports are measured through the second quarter of 1960, but the differences are probably attributable mainly to cyclical influences.

Table 2.

Indices of Volume of Imports, 1955, 1956, and 1959

(1958 = 100)

article image
Source: United Nations, Monthly Bulletin of Statistics, April 1961, Special Table A.

Rates of increase are calculated on a compound basis.

55. The increase of world imports at the rate of 4.3 per cent a year in 1957–59 was accompanied by some price reductions. World export prices fell by 3 per cent from 1956 to 1959; export prices of developed areas decreased by 2 per cent, and those of less developed areas by 8 per cent (Table 3).

Table 3.

Indices of World Export Prices, 1956–60

(1953= 100)

article image
Source: United Nations, Monthly Bulletin of Statistics, March 1961, Special Table A, and April 1961 Special Table A.

Third quarter.

56. The total and the ownership of international reserves were affected by many events in the period 1957–59. Fund quotas were increased from $8.9 billion at the end of 1956 to $14.0 billion at the end of 1959, with one quarter of this increase payable in gold. The major countries of Europe took far-reaching convertibility decisions at the end of 1958. At that time, more than $1 billion of reserves vanished from reserve statistics overnight, as short-term multilateral EPU credit balances were converted into longer-term bilateral balances. There was a substantial redistribution of reserves. Gold holdings of the United States fell sharply, and its short-term dollar liabilities increased substantially. Reserves of a number of other countries, notably India, Venezuela, and Egypt, decreased; although these decreases were much smaller than the decline in U.S. gold reserves, they were very significant for the countries concerned. Western Europe, including the United Kingdom, was the beneficiary of practically all of the gross increase in reserves and of the redistribution of reserves. Measuring the useful increase in reserves therefore presents some difficulties, while evaluating the increased effectiveness of reserves presents even greater ones. A number of separate indicators of changes in world reserves are given in Table 4.

Table 4.

International Reserves

(In billions of U.S. dollars; data pertain to end of year)

article image
Source: International Monetary Fund, International Financial Statistics, January 1961.

57. The stock of monetary gold held by countries and international institutions increased by 5.5 per cent during the three-year period, or at the rate of 1.8 per cent a year. Countries, because of payments to the IMF in connection with enlarged quotas, increased their gold reserves at a lower rate. Aggregate exchange reserves of all countries decreased by $400 million during the period and totaled $19.0 billion at the end of 1959; the amount of Fund drawings included in exchange reserves was $450 million larger at the end of 1959 than it had been three years earlier. According to one measure of over-all reserves, namely, world holdings of gold plus country holdings of foreign exchange, reserves increased from $57.5 billion at the end of 1956 to $59.2 billion at the end of 1959, or by less than 1 per cent a year.44

58. From 1956 to 1959 there was substantial easing in the international financial situation. Reserves of the countries in continental Western Europe increased to adequate, and in some cases admittedly high, levels. The major European currencies were made externally convertible at the end of 1958 and have been further liberalized since that date. Exchange restrictions in the world at large were reduced. A number of countries, including France, improved their reserve positions notably.

59. The Fund approved $3.4 billion of drawings from the beginning of operations through the end of 1959; $1.3 billion of this amount was outstanding at the end of the period (Table 5). In addition, it approved $2.0 billion of stand-by arrangements under which $208 million remained available at the end of 1959 for drawings. The years 1956 and 1957 were particularly active ones, accounting for one half of all the drawings and three fifths of all the stand-by arrangements agreed by the Fund from the beginning of operations to the end of 1959. In contrast, drawings and stand-bys in 1958 and 1959 were much less, while repurchases and expirations of stand-bys were substantial. Drawings outstanding at the end of 1959 were $450 million more than they had been three years earlier, but amounts available under stand-by arrangements were $900 million smaller. For the purpose of assessing the ease or the tightness of international liquidity, the most appropriate measurement of Fund assistance is probably drawings outstanding plus amounts still available under stand-by arrangements. By this measure, Fund assistance decreased from $1.9 billion at the end of 1956 to $1.5 billion at the end of 1959, and decreased further in 1960 to $1.25 billion. It should not be overlooked, however, that transactions with the United Kingdom are responsible for a substantial part of this trend.

Table 5.

Summary of Transactions of the International Monetary Fund, 1955–60

(In millions of U.S. dollars)

article image
Source: International Monetary Fund, International Financial Statistics, January 1961.

60. The immediately preceding discussion may be summarized as follows: In the three years 1957–59, the volume of imports increased 4.3 per cent a year, a rate of increase which was substantial, though somewhat lower than in earlier postwar years; export prices fell slightly, but were characterized by a stability which had been regarded only a few years before as most difficult to attain; the most important European currencies were made externally convertible; the over-all level of trade restrictions was reduced; and members’ use of the Fund’s resources decreased after 1957. The United Kingdom, Japan, and most of the industrial countries of Western Europe improved their reserve positions, in many cases markedly, without any of the severe deflationary measures usually associated with recovery from an inadequate level of international reserves. While all of these developments were taking place, international reserves increased at a slow rate, certainly less than 2 per cent a year. If the assumption is made that international reserves should increase as fast as trade, the events of 1957–59 would appear to be inconsistent with such a low rate of growth of international liquidity.

61. This three-year period nevertheless saw a number of disquieting developments. Perhaps the major one was the emergence in 1958 of the large balance of payments deficit of the United States, a deficit which has not yet been eliminated. But no one has ever suggested that this deficit was caused by an inadequate rate of expansion of international liquidity—nor that it can be cured by an increase in international liquidity. The balance of payments deficit of the United States did swell the reserves of the rest of the world, and to this extent may perhaps be considered as offsetting what would otherwise have been an inadequate rate of growth of reserves. But this is only to argue against the proposition that the total of world reserves is the only measure of adequacy and the only basis for financial policy. The distribution of reserves among countries has on many occasions been extremely important, and a number of other elements, particularly the quality of reserves and the substitutes for or supplements to reserves, also play important roles. In any event, the increase of reserves of all countries other than the United States during 1957–59 was somewhat smaller than the increase of imports, which was itself very highly concentrated in industrial countries.45

62. The Fund at all times has had enough resources to operate on a substantial scale. Even the large operations in 1956 and 1957 did not rule out sizable additional operations, while the smaller operations in 1958 and 1959 clearly left room for additional drawings and stand-by arrangements.

This raises several questions: Why was it that the Fund was not more active? How could the Fund have engaged in additional operations if its policy had been directed toward pumping additional liquidity into the international system in order to increase the rate of expansion of reserves to 3 per cent a year, or even to higher rates closer to the actual rate of growth of trade? Was the lending policy of the Fund too “tough” and did it leave unsatisfied a large amount of appropriate demands for drawings and stand-by arrangements? Such a charge against Fund policy has not been made. There would certainly be no agreement that Fund policy was too “tough” if its objective was to make short- and medium-term (up to five years) advances. But if the policy objective was to increase further the rate of growth of international reserves, Fund policy undoubtedly was too “tough.” The required increase of reserves for the three-year period 1957–59 at a rate of only 3 per cent a year was $5.2 billion, when the smallest reserve base, i.e., country reserves of gold and exchange, is used (Table 6). Reserves of all countries increased by only $1.4 billion. On this basis the Fund should have increased net drawings outstanding (plus undrawn stand-bys?) by an additional $1.3 billion a year. At a rate of 4.3 per cent a year, i.e., the rate of growth of imports during 1957–59, country reserves of gold and exchange should have increased by $7.5 billion instead of the actual $1.4 billion. This would have required an increase of $2.0 billion a year in international reserves.

Table 6.

Required Growth of Reserves at Various Rates, 1957–59

(In billions of U.S. dollars)

article image

63. A management of the IMF operating during 1957–59 on the proposition that international reserves should increase pari passu with imports, and that an annual rate of increase of 3 per cent was “definitely noninflationary,” would have tried to increase reserves by at least an additional $1.3 billion a year. The IMF could have achieved this objective by making additional loans to nonindustrial countries, and it could also have engaged in investment operations. A large proportion of such funds would inevitably have been spent on imports from industrial countries, probably for economic development. Whether such an expansion in this period at this rate would have been desirable and noninflationary, and whether this procedure would have been the most appropriate for attaining the desired goal, are questions that would require extended consideration.

64. Alternatively, XIMF lending policy might be related to the growth of reserves, and the determination of reserve adequacy, as measured for countries plus all international monetary organizations (IMF, EPU/EF, and BIS).

On this basis, reserves increased by $6.7 billion in 1957–59, of which the major part ($4 billion) represented currency payments (i.e., tendering of notes payable on demand in currency) by IMF members in connection with their quota increases of 1959. The increase of $6.7 billion was more than enough to meet requirements based upon a growth of imports of 3 per cent a year, but it fell short of meeting the actual rate of 4.3 per cent by almost $800 million.46

The large increase of reserves was achieved by enlarging Fund quotas. Why the delivery to the Fund of $4 billion of demand notes, which may not be cashed for many years, improves world liquidity immediately, may be puzzling. The explanation probably is that the usual meaning of the term “reserves” is too restricted, and that it would be useful to introduce such a term as quasi-reserves into the analysis of the statistics. Possibly published statistics on country reserves of Fund members would be more meaningful if their gold contributions to, their creditor positions in, and their drawing rights on, the IMF were considered along with their reserve holdings of gold and exchange.

65. International liquidity increases when the ability to borrow increases, whether from international, national, or private sources.47 It also follows that, with a given amount of borrowable resources, liquidity is greater when countries are willing to borrow than when they are not. When countries can borrow, and are willing to borrow, it is unnecessary for an international organization to force-feed liquidity into the system and into the reserve statistics, in accordance with calculated requirements based upon the growth of trade. Nor is it meaningful after the fact to compare the actual growth of reserves with the actual growth of trade. It is sufficient to ensure that the amount of quasi-reserves is large—that the lender of last resort always has some resources to lend—and to let the reported total of reserves take care of itself. Countries will then rely on reserves and quasi-reserves in the proportion that seems to them most appropriate.

Some Political and International Implications

66. Triffin has remarked that “there is no doubt that these reforms could be viewed as a first step toward the setting up of a supranational monetary authority, to which central banks and governments are understandably reluctant to yield any parcel of their cherished national sovereignty and independence.” 48 This is a consideration which is partly real and partly fictitious, but it is not one that can be put aside. It raises a number of complex issues, of which four are briefly discussed here: the role of the XIMF in national money markets; the role of the XIMF in international investment; some political aspects of exchange guarantees; and the present and future control of the Fund.

67. The XIMF would inevitably have to operate in many of the world’s money markets, particularly those in the United States and the United Kingdom. The XIMF would begin operations with more than $9 billion of dollar assets and more than $5 billion of sterling assets, based on data as of the end of 1958.49 These assets consist of demand and time deposits, and short- and long-term government securities. Management of these portfolios would require many Fund decisions: to keep deposits in some banks rather than others, to hold assets in cash or to invest, and to vary the proportion of investments between short-term and long-term securities. Decisions of this sort could affect both the level and the structure of interest rates; they would be reinforced by exercise of the option to demand the annual liquidation of up to 5 per cent of XIMF holdings of dollars and sterling, equal to $500 million of dollars and $250 million of sterling.

68. The Triffin proposals would greatly increase the amount of internationalized investment. The decisions to invest, and the directions and the terms of investment, would be taken by the governing authorities of the XIMF. If the XIMF did not wish to make all of its investments directly in member countries, it could invest indirectly, through such other international agencies as the IBRD. The investments (apart from reinvestments) would be made with credit newly created by the XIMF; the real resources that would be bought with these funds would come from countries, very largely from developed countries. Injections of bank credit might be helpful if countries were in a recession, and inflationary if their resources were fully employed or if there were important bottlenecks in production. In either case, countries would have to readjust their ratios of domestic to overseas availabilities, and of savings to investment, in order to meet these additional export demands.

69. The political problems involved in negotiating an exchange guarantee of the scope proposed by Triffin are probably enormous in creditor countries with short-term foreign liabilities, particularly the United States and the United Kingdom. The long-term foreign investments of these countries are valued at billions of dollars. They are several times as large as foreign holdings of dollars and sterling. All of these investments in securities and business enterprises are exposed to risks: devaluation, changes in business policy and taxation, and, in some cases, the more than negligible possibilities of expropriation and “intervention.” Countries with large foreign investments may well be unwilling that their own investments should not be guaranteed, but that external holdings of their currencies should be.

70. Finally, brief mention should be made of the membership of the Fund, now and in the future. The significance of stipulated voting majorities (½, ⅔, ¾, etc.) depends very much upon the size of the XIMF, the number of its members, and voting rights once quotas are abolished. The membership of the Fund has increased substantially since 1946. It may well continue to increase. The Soviet Union attended the Bretton Woods Conference, and its quota then agreed was slightly less than that of the United Kingdom and almost 15 per cent of the total. Czechoslovakia and Poland were original members of the Fund with combined quotas equal to almost 3 per cent of the total. These three countries could conceivably be admitted to membership in the XIMF under political conditions which cannot now be predicted. Other communist countries also could become members; and less developed countries that already have attained independent status, or that may do so in the future, may join the XIMF. These potential additions to membership should be carefully considered when it is proposed that the XIMF be given the power to create credit, to make international investments, and to exercise important powers in money markets.

III. Conclusion

71. The implications of Triffin’s proposal are far-reaching. An international agency would be authorized to manage and systematically expand the international money supply. The character of the Fund’s operations would be changed to facilitate expansion of international liquidity at what is conceived to be an adequate rate. The expanded Fund would have authority to create new money in the form of deposits, which would to an increasing extent rest on investments in underdeveloped net debtor countries. The process of creating deposits would involve drafts on the real resources of other members, largely the developed countries, and it could readily become inflationary. The character of the assets of the expanded IMF would increasingly be changed from currency holdings and three- to five-year loans, conceived to be self-liquidating and revolving, into medium- and long-term loans which would have a different character. These changes would rest on important exchange guarantees. They could involve intervention of the XIMF in the money markets of the United States, the United Kingdom, and perhaps other countries. They would involve changes in the sources of funds for, and in decisions over, long-term investment. The authority to require gradual liquidation of dollar and sterling balances would reduce either the gold reserves or the long-term capital exporting ability of the United States and the United Kingdom. The expanded IMF would inevitably be less liquid than the present one, and this might raise questions about its ability to create deposits as good as gold and to provide the key currencies that members may need, unless reserve deposit requirements are raised and the gold contributions of a few members increased far beyond the level initially proposed.

72. These far-reaching changes, and others which can be foreseen only dimly, are based upon findings of the dangerous state of, and prospects for, international liquidity. The proposals are based upon a simplified view of the statistics on reserves and trade that does not reflect such important factors as the distribution of reserves, the change in the quality of exchange assets, the state of balance of international trade and exchange rates, and the growing role of the IMF. At best, these findings of serious reserve deficiency are unproven; at worst, they are incorrect, at least for the next five or ten years. It may be argued that ten years is not a long time, and that the world should now anticipate developments over a much longer period. As to this, opinions differ. Since 1900 there have been many forecasts of reserve requirements based upon the rate of growth of trade. With practically no exceptions, these forecasts concluded that reserves would soon become inadequate, and that this inadequacy would cause deflation. There have been many forecasts since World War II of growing inadequacy. The remedies for this inadequacy have ranged from proposals to increase liquidity by raising the price of gold to Triffin’s proposal to create a new kind of international deposit money. Yet for a large part of this period the problem was inflation, not deflation. Reasonable price stability, if it has indeed been achieved, is of recent date. Such threats as there are to this stability hardly stem from the side of deflation.

73. These comments in no way imply that international liquidity does not present many difficult problems. However, there is reason to believe that, in view of the recovery already made, the lines of development pursued in the postwar period offer the best chance for further progress. The international financial machinery has proved to be flexible and adaptable. Within its terms of reference, or with relatively small changes in them (compared with the large and controversial changes proposed by Triffin), this machinery is capable of further growth and adjustment. The IMF with its enlarged resources and its proven type of operation provides the guidance and monetary support that should facilitate further expansion and stability of the structure of international reserves.


I. Rate of Growth of Trade as a Guide to XIMF Operations

The rate of growth of trade may be used in three ways: as an ex ante indicator of reserve requirements, as an ex post measure of the adequacy of the growth of reserves, and as a current guide to the operations policy to be followed by an expanded IMF.

Triffin clearly uses the rate of growth of trade as an ex ante indicator of future reserve requirements. The conclusion that the prospective rate of growth of reserves in the next decade will be inadequate rests on this base. He uses trade as an ex post measure much less clearly. His judgment that the present level of international liquidity is inadequate because reserves have failed to increase in the postwar period as rapidly as trade is not clear-cut. Such a judgment inevitably must rest on a number of elements, including the base line from which calculations should start. This difficulty will also be present a decade hence if an attempt is then made to measure the adequacy of reserves solely with respect to growth rates.

The use of the rate of growth of trade as a current guide to the operations policy of the XIMF, though perhaps clear at first thought, becomes increasingly cloudy on further reflection. Triffin proposes50 to expand the IMF into a central bank for central banks, so that it can be concerned with the adequate growth of international reserves and with the stability of the structure of reserves. The adequate growth of reserves is consistently taken to be growth at the same rate as the volume of trade, and all calculations of reserve requirements and shortfalls are based upon this definition. It might, therefore, be expected that he would set up the growth of trade as the guide to, and the determinant of, the operations policy of the expanded Fund. But this is not the case. Triffin nowhere lays upon the XIMF the responsibility of increasing international liquidity pari passu with trade. In fact, he does not explicitly lay upon the XIMF the responsibility of increasing international liquidity at an adequate rate. On the contrary, the growth of trade is used only as a limit to XIMF lending. Though “a reasonably conservative solution would be to retain a 3 per cent figure as definitely noninflationary,” 51 the appropriate rate of expansion of XIMF deposits is left to the judgment of the Executive Directors; rates of expansion greater than 3 per cent a year would require progressively larger majorities.

It may be interesting to speculate why it is necessary to read Triffin so carefully on this question of using the rate of growth of trade as a guide to, and determinant of, the operations policy of the expanded Fund. The path of his argument through prospective reserve requirements and the imminent reserve deficiency is so straightforward that it is easy to conclude that the operations policy of the XIMF should be directed toward achieving an adequate rate of growth of reserves, one equal to the rate of growth of trade. Many commentators have done this. For example, The Economist (London) stated that the expanded IMF would “arrange that its net lending in a period of twelve months, together with current increases in the supply of gold, would increase total world reserves by say 3–5 per cent a year. Thus world liquidity would keep up automatically with world trade.” 52 An article in The Banker (London) speaks of “the Fund’s basic lending, the amount of which would be determined on an arithmetical basis linked to world trade. Thus the Fund’s basic lending over any twelve months might be fixed at such an amount as would, together with current increases in stocks of monetary gold, increase world reserves by 3–5 per cent a year.” 53 One highly qualified European observer put the matter this way:

The Fund shall so handle its lending operations that international monetary reserves can grow at the necessary rate. A growth of total reserves, including gold, of around 3 per cent annually should be regarded as normal. But in case of need the Fund, with a qualified majority, shall be able to bring about a growth of world currency reserves in excess of this rate.

Professor Brian Tew described the operations policy of the XIMF as follows:

The aim would be to increase the total amount of Bancor in existence at a rate sufficient to ensure that the world’s stock of Bancor plus gold should grow pari passu with legitimate requirements for international liquidity in an expanding world economy. In order to meet the “inflationary bias” danger raised against Keynes’s proposals, Triffin suggests very simply that a presumptive ceiling on the Fund’s loans and investments be agreed to in advance, and that qualified majority votes (⅔, ¾, or even ⅘ of the total voting power) be required to exceed this ceiling… .54

The writer’s initial impression, while reading Triffin, was that the expanded Fund would, without question, direct its lending and investment operations to assure an adequate rate of growth of the world’s reserves, and that a 3 per cent rate was a self-evident policy target. This is strongly implied on many occasions, but nowhere stated in so many words. The expanded Fund is nowhere clearly directed to provide an adequate rate of growth of reserves, the prospective inadequacy of which is responsible for its being, nor is it told what an adequate rate of growth would be. The reader has to conclude that Triffin proposes to expand the functions of the IMF and give it a full armory of central bank instruments, but that he is unwilling to provide it with an explicit statement of policy objective, let alone any specific guides to operations policy.

II. Adequacy of Reserves in the United Kingdom

The balanced and carefully considered views expressed by the Radcliffe Committee55 are very suggestive on this point. The Committee does not come down on the side of increasing reserves unequivocally or as a grave matter of overriding priority. Its views are contained in a number of separate observations, of which the most important are indicated here. The world total of international reserves is scarce—certainly no more than barely adequate—but it should not be augmented by an increase in the price of gold.56 The resources of the Fund should be augmented and its scale of activities enlarged.57 It is impossible to “to disentangle the problem of international liquidity from the problem of international balance… . So long as no one country is a large and persistent debtor or creditor, a succession of deficits and surpluses can be handled with remarkably small liquid reserves.”58 These remarks, in addition to those bearing on the responsibilities of the United Kingdom to provide liquidity for, and capital exports to, the sterling area, provide the context for evaluating the following two most important paragraphs bearing on the desirable level of reserves:

662. We regard the right course of action as one calculated to add to reserves or reduce liabilities out of a current surplus sufficiently large to leave room also for long-term investment abroad. We do not suggest that the improvement in reserves and liabilities should be brought about precipitately. Repayment of sterling balances, so far as they constitute the central reserves of other countries or the working balances of overseas commercial banks, would tend to reduce the liquidity of overseas monetary systems. If sterling outgoings were restricted in order to force these countries to draw on their sterling resources, some of them might be faced with a sterling shortage and a general contraction of activity would be precipitated in a group of countries which includes some of the United Kingdom’s principal export markets. The same kind of situation might arise in other markets if the United Kingdom tried to restrict imports in order to add to the reserves. Quite apart from the fact that such action would be inappropriate in a period of trade recession, efforts by the United Kingdom to improve her own liquidity are bound to reduce the liquidity of other countries and aggravate any shortage of international liquidity. We should therefore refrain from seizing too eagerly on the opportunity of extinguishing short-term debts as a means of strengthening, if only temporarily, the pound sterling.

663. So long as there is no special danger of a general shortage of international liquidity, an increase in reserves should, in our view, have priority. No doubt the United Kingdom has not much freedom of choice as to whether to strengthen her position by increasing reserves or reducing liabilities; the decision is largely governed by the willingness of other countries to hold sterling. An increase in reserves, however, marks more unambiguously than a decline in sterling balances a strengthening of the position of sterling and would be so interpreted by financial opinion. If the decline were in the balances of sterling countries it would do little to reinforce the liquidity of the sterling area as a whole, although it would improve the position of the United Kingdom. What the United Kingdom can do to add to her reserves depends, however, on the state of international liquidity, and this also has a direct bearing on the reserves which she requires.


Mr. Altman, Advisor in the Research and Statistics Department, is a graduate of Cornell University and the University of Chicago. He taught economics at Ohio State University and was on the staff of the National Resources Planning Board and of the French Supply Council. He was Director of Administration of the Fund until 1954. He is the author of Savings, Investment, and National Income and of a number of papers published in technical journals.


Comptes Rendus des Travaux de la Société Royale D’Economie Politique de Belgique, No. 272; and Current Economic Situation and Short Run Outlook, Hearings Before the Joint Economic Committee of the Congress of the United States (86th Congress, Second Session, December 7 and 8, 1960). The latter will hereafter be referred to as Hearings, 1960.


An earlier and somewhat differently arranged version of this article was published in Hearings, 1960, pp. 175–207.


Gold and the Dollar Crisis, p. 100.


Ibid., pp. 49–50.


Reserves of the United Kingdom and France at the end of 1957 were $3.1 billion; reserves equal to 40 per cent of their imports in 1957 would be $7.0 billion, leaving an initial shortfall of $3.9 billion. Reserves of the United Kingdom alone at the end of 1957 were $2.4 billion. Reserves corresponding to 40 per cent of imports would be $4.6 billion; and if growth at the rate of 3 per cent a year is assumed, they would in 1967 be $6.1 billion. Under the same assumptions, French reserves in 1967 would be $3.3 billion.


The illustration of the world’s reserve deficiency for a decade in the Wicksell Lecture was that “world trade and production are estimated to have increased in volume, over the last ten years, at a pace of about 6 per cent a year. A parallel rate of increase of the world’s monetary reserves ($62 billion at the end of last year [1957]) would require their expansion by $3.7 billion annually.” As for a low rate of growth, he stated that “the 3 per cent rate assumed by the Fund [in International Reserves and Liquidity] becomes plausible only when ‘normal’ peacetime experience is diluted with the abnormally low, and in fact predominantly negative, growth rates of wartime years and of the 1930’s world depression” (Gold and the Dollar Crisis, p. 48).


By the end of 1960, the “net” reserve position of the United States, as measured by gold reserves in relation to short-term foreign official and private dollar holdings, was negative. If short-term dollar liabilities are understood to include U.S. Government bonds and notes with an original maturity of more than one year, the deficit in “net” reserves was more than $2 billion. This is the basis for the comparison made by Triffin in Hearings, 1960, p. 224.


Employment, Growth, and Price Levels, Hearings Before the Joint Economic Committee of the Congress of the United States (86th Congress, First Session, October 26–30, 1959), Part 9A, p. 2931 (hereafter this will be referred to as Hearings, 1959); Gold and the Dollar Crisis, p. 9.


Hearings, 1959, Part 9A, p. 2911. The same statement is in Gold and the Dollar Crisis, p. 89.


Hearings, 1959, Part 9A, p. 2911.


Gold and the Dollar Crisis, p. 9. The same point is referred to on pp. 88–89 in terms of “the policies adopted to readjust the country’s balance of payments and arrest the gold outflow.” Success of these policies entails the cessation of the current contributions of the key currency countries “to the solution of the international liquidity problem. Their failure, on the other hand, may have far worse consequences still by stimulating large-scale liquidation of outstanding foreign held balances, and triggering an international financial panic involving other currencies as well. This is precisely what happened in 1931… .”


Ibid., pp. 113–14.


Ibid., p. 107. The meaning of the term “semi-liquid” is not spelled out, but it could hardly include balances under payments agreements, holdings of inconvertible currencies, or the $2.5 billion of foreign currencies obtained by the United States in connection with the sale of agricultural surpluses.


Hearings, 1959, Part 9A, p. 2911; cf. Gold and the Dollar Crisis, p. 117.


Gold and the Dollar Crisis, p. 105.


Ibid., pp. 103–04. See also Appendix I, below.


Gold and the Dollar Crisis, p. 104, footnote 1.


Hearings, 1959, Part 9A, p. 2912; cf. Gold and the Dollar Crisis, p. 115.


Gold and the Dollar Crisis, p. 115.




Ibid., p. 103


Hearings, 1959, Part 9A, pp. 2938–39


At the end of 1956, approximately 40 per cent of sterling liabilities was held in the form of bank deposits and Treasury bills. The remainder (60 per cent) was held in the form of government securities, and of this, half had maturities longer than five years.


Total exchange assets held officially are estimated at $19.2 billion as of the end of 1958 (International Monetary Fund, International Financial Statistics, September 1960).


Gold and the Dollar Crisis, pp. 113–14. It is not clear why the same consideration does not apply to official balances at the present time.


Under the present Articles of Agreement, the Fund can sell gold, or borrow, to replenish its currency holdings. It may be prudent, as E. M. Bernstein has suggested, to arrange for borrowing in advance of need by issuing to a small number of countries debentures which could be called for payment to meet large emergencies; see International Effects of U.S. Economic Policy, (Joint Economic Committee of the Congress of the United States [86th Congress, Second Session], Study Paper No. 16, January 25, 1960), pp. 84–86.


On this point, Triffin has generally approved the estimates in Oscar L. Altman, “A Note on Gold Production and Additions to International Gold Reserves,” Staff Papers, Vol. VI (1957–58), pp. 258–88.


Gold and the Dollar Crisis, p. 36. See also Appendix I, below.


Gold and the Dollar Crisis, p. 43.


Ibid., p. 50. For purposes of calculating reserve requirements, Triffin assumes a figure of $4.6 billion as of the end of 1957 (ibid., p. 50). For the views of the Radcliffe Committee on the U.K. reserves, see Appendix II, below.


The Future of the European Payments System (Stockholm, 1958), p. 34.


These data were taken from International Financial Statistics, October 1959.


Gold and the Dollar Crisis, p. 110.


Ibid., p. 114


Par. 16.


Gold and the Dollar Crisis, p. 104.


Ibid., p. 115.


Ibid., p. 110. The Economist (London) commented that “Professor Triffin proposes an orderly liquidation of such balances, say by 5 per cent a year …” (January 9, 1960, p. 134), and a number of other commentators drew the same unqualified inference. Triffin replied in a letter (The Economist, January 23, 1960, pp. 304–05) describing the proposal stated here. In this letter he also wrote, “I would indeed expect the Fund to accumulate instead additional sterling and dollar balances in the initial years… .” This expectation does not appear in Gold and the Dollar Crisis, nor in earlier writings.


Gold and the Dollar Crisis, pp. 117–18; Hearings, 1959, Part 9A, p. 2912.


“Altman on Triffin: A Rebuttal,” in Hearings, 1960, p. 212.


Cf. Hearings, I960, p. 212.


This was the arrangement suggested for the International Clearing Union; see Proceedings and Documents of the United Nations Monetary and Financial Conference, (Washington, 1948), Volume II, p. 1565, par. 29.


As is evident from Table 2, in 1957–59 imports by North America and Western Europe increased at almost the same rate. Rates of increase based upon export data tell a very different story, since they were affected by the events at Suez. World exports in 1957–59 increased by 3.8 per cent a year. Exports by regions increased as follows: 3.4 per cent for developed areas including North America (– 2.8 per cent) and Western Europe (7.0 per cent); 3.6 per cent for less developed areas.


As already noted, country holdings of exchange reserves were affected by the elimination of $1.3 billion of claims on EPU after the end of 1958. If adjustment is made for this nonrecurrent event, world gold plus country exchange reserves increased by $3 billion from 1956 to 1959, or by 1.7 per cent a year.


In 1957–59, the volume of imports for the whole world increased by 4.3 per cent a year. Imports for the world excluding North America increased by 4 per cent. (The corresponding increases for exports were 3.8 per cent and 6 per cent, respectively.) Reserves of all countries other than the United States, including the contribution to reserves made by the balance of payments deficit of the United States, increased by 3.8 per cent a year.


On the basis of world reserves (Table 4), the growth of reserves required to match world imports increasing at a compound rate of 4.3 per cent a year would have totaled $8 billion.


When the United Kingdom made advance repayments to the Fund in 1960, Treasury officials are reported to have observed that these repayments did not in any way weaken the United Kingdom’s reserve position. The repayments were considered to be merely a transfer from front-line reserves to secondary reserves, since they enlarged U.K. drawing rights on the Fund (The Wall Street Journal, September 6, 1960). Governor Cobbold of the Bank of England stated in a speech in October 1960, “I hope that we shall all get more used to regarding the International Monetary Fund as a second line of reserves. Too little importance has been attached to the very large increase in their facilities which was arranged last year, and to the part they can play in offsetting these [short-term money] movements… . I should like to see countries draw on these facilities as a matter of ordinary business when they need to reinforce reserves, and repay when reserves are rising… .”


“The Twilight of the Gold Exchange Standard and the World Dollar Crisis,” in Hearings, 1960, p. 238.


It may be noted, by way of comparison, that the Federal Reserve System owns $27 billion of U.S. Government securities and all commercial member banks another $27 billion (data are for November 1960, from the Federal Reserve Bulletin, December 1960).


Hearings, 1959, Part 9A, p. 2938; Hearings, 1960, p. 238.


Gold and the Dollar Crisis, p. 103. In a later statement, this point was made as follows: “One could, for instance, adopt a highly conservative estimate of such needs—allowing, let us say, for a maximum reserve increase of 3 or 4 per cent annually—provided that additional lending be made permissible, in case of need, by special voting majorities assuring a proper control of such decisions by the major creditor countries in the Fund” (Hearings, 1960, p. 238).


January 9, 1960, p. 134.


Fred Hirsch, “Development Aid and World Reserves,” The Banker, March 1960, p. 148.


International Monetary Cooperation, 1945–60, (London, 5th ed., 1960), p. 184.


Committee on the Working of the Monetary System, Report, (Cmnd. 827, London, 1959).


Ibid., pars. 672–75.


Ibid., pars. 675–78.


Ibid., par. 680.