W. R. Gardner and S. C. Tsiang *


W. R. Gardner and S. C. Tsiang *

W. R. Gardner and S. C. Tsiang *

THE FUND AGREEMENT outlaws competitive depreciation. Article I (iii) states that it is a purpose of the Fund “to avoid competitive depreciation” and under Article IV, Section 4(a) “each member undertakes to collaborate with the Fund . . . to avoid competitive exchange alterations.” Since every exchange devaluation is designed to strengthen the competitive position of the devaluing country in the markets of the world, a question might be raised whether the Fund is basically opposed to exchange adjustments. It is obvious, however, that that is not so. The Articles of Agreement provide in some detail for changing par values in order to correct a fundamental disequilibrium. Clearly an adjustment of this character, even though it would subject other countries to more competition, could not be branded as competitive depreciation and, therefore, contrary to the purposes of the Fund. Only an excessive devaluation—one that over-corrected for the fundamental disequilibrium—could be regarded as competitive depreciation in the sense of the Fund Agreement. The remainder of this paper will be devoted to analyzing from the economic standpoint what constitutes such an overcorrection.

Exchange adjustment and reserves

An excessive devaluation must be measured in terms of its departure from a correct exchange rate. What are the criteria of a correct rate? It must be assumed in this discussion that devaluation helps a country to eliminate a deficit in its balance of payments, or is expected to do so; otherwise there would be no balance of payments reason for devaluing. If, given time, a devaluation would just remove the deficit, it would prima facie restore equilibrium; but merely eliminating the deficit would not correct the effects produced by the previous maladjustment. It might leave the country with depleted reserves. To obtain reserves sufficient to cope with the unpredictable fluctuations of its balance of payments, the country should pass from a deficit to a surplus in its international transactions. The surplus should be on a sufficiently modest scale so that the country can ease back into balance when its reserve position has been restored; but that a surplus may be needed to make up for the previous deficit is clear.

Providing the rate of accumulation is moderate, the process of building up reserves could properly be carried in most cases beyond the amount lost during the deficit period. Most countries outside the United States today appear to have inadequate reserves to cope with the potential deficits in their balance of payments. This is particularly true if reserves are ultimately to be sufficient to take care of deficits in a convertible currency world. They could never, of course, enable a country to finance a chronic deficit that it was taking no steps to correct. That is not their purpose. But they appear to be inadequate to deal even with the temporary deficits that are likely to develop in a convertible currency world during the course of a business cycle or during the period necessary for corrective measures to restore balance. Because of this general shortage of reserves, a moderate accumulation from year to year (if it could be achieved) would be constructive for most countries. Only if the rate of accumulation were so rapid and persistent that it threatened to bring a country more than its appropriate share of the world’s stock of reserves would there be direct evidence of competitive depreciation.

The use of the term “reserves” here involves some oversimplification. The full measure of the deficit which the country is endeavoring to correct is to be found not in its loss of reserves but in compensatory official financing in all its forms. Reserves are merely one element in such financing. Intergovernmental loans and grants from abroad may play a far greater role in covering a country’s deficit. To achieve equilibrium the country must readjust its balance of payments so as to eliminate the need for compensatory financing as a whole; and similarly, in building up a surplus, the country may not only acquire reserves but may repay its debts abroad in advance of maturity and make compensatory loans or grants of its own. It seems unlikely, however, that a country that has been struggling with a deficit would immediately start extending compensatory loans and grants once a surplus is achieved. Hence, aside from possible prepayment of debts, the full measure of the surplus in most cases is likely to be the reserves that the country accumulates. 1

In an inconvertible currency world, a country may have to adopt exchange rates that will balance its position in hard currencies or bring a reasonable addition to its hard currency reserves even though the rate leads to a simultaneous accumulation of soft currencies. The inconvertibility of soft currencies is evidence that the international position of the soft currency countries is maladjusted rather than that the exchange rate of the devaluing country is incorrect. To the extent that soft currencies are themselves convertible into hard (e.g., the dollar pool in the sterling area arrangements) they become part of the hard currency position.

The time period involved

The time period in terms of which the effects of the exchange rate must be considered is a period of years rather than of months. Even in a stable world it takes a considerable time for an exchange adjustment to work out its effects. A capital flight may be reversed quickly if it is believed that the new rate will hold; but the reorientation in the sphere of goods and services cannot be carried through so easily. Substitution for imports or larger production of exports may have to be achieved through shifts of manpower, resources, and capital equipment; and new channels of distribution and servicing may have to be developed abroad. Furthermore, it may be necessary to average out a world business cycle before the effects of the devaluation on commodity markets can be appraised. They cannot be appraised against a background of abnormal boom or depression. The test, then, is what the devalued rate will do over a substantial period of years.

By the end of this period of years a correct exchange rate should not have to be supported by (1) a persistent one-way movement of reserves or other forms of compensatory financing, or (2) direct balance of payments controls (i.e., trade or exchange controls imposed to adjust the balance of payments rather than to afford protection to domestic industry, raise fiscal revenues, or achieve other ends not associated with the balancing of international transactions). The new rate should, in fact, enable the country to add a reasonable amount to its reserves if they are unduly low.

Effects on other countries

Even an appropriate devaluation by one country affects the competitive position of others. Before devaluation, these other countries enjoyed a competitive advantage of which they are deprived by the elimination of the overvalued rate. They must readjust to the new situation in world markets. This readjustment is not forced upon them by competitive depreciation, if the devaluing country has merely shifted from an overvalued to a correct rate. It is forced upon them rather by the removal of an artificial support, which must be withdrawn if the world is to move toward a realistic pattern of exchange rates and currency convertibility. Significant readjustments will be needed, however, only on the part of countries that are materially affected by the devaluing country’s action, and that would otherwise experience a deficit in their balance of payments or a surplus substantially less than the reserves they should be accumulating. The adjustments may take other forms than exchange depreciation. Even if exchange depreciation is involved, it would normally be less than that of the initial devaluing country except where comparable balance of payments deficits are being experienced.

The adjustments of other countries will in turn affect the milieu in which the initial devaluing country finds itself. Ideally, that country’s devaluation should be such that it will achieve an appropriate accumulation of reserves in the face of the responses forced on other countries, if these responses in each case involve a correct adjustment. In fact, however, these other countries may refrain from making any adjustment or may try to overadjust their rates or impose direct controls. Unless the Fund can induce such countries to adopt correct measures, it will have to judge the proposal of the initial devaluing country against the background of controls and maladjusted rates that will in fact prevail.

In short, in order to judge whether a proposed exchange adjustment is correct, the Fund must have a broad concept of (1) the appropriate distribution of reserves in the world, (2) the effects on the world pattern of exchange rates and the nexus of direct controls that the proposed adjustment would introduce, and (3) the extent to which, in the light of this new milieu, the new exchange rate would enable the devaluing country over a period of years to acquire reserves.

Some special cases

There are several cases in which a country may not acquire reserves (at least immediately) and yet may be guilty of excessive depreciation. Some of these are discussed below, together with others of a different type in which a country that is acquiring reserves at a reasonable rate may devalue without competitive depreciation.

  • (a)A rate may involve competitive depreciation even though a country’s reserves are depleted and it is clear that the new rate will not increase those reserves in the immediate future—particularly if the country’s direct exchange and trade controls are lifted. Few European countries in the immediate postwar period could have adopted any rate that would have enabled them to drop direct controls, do without the compensatory financing provided by the United States, and still add to their reserves within the next year or two. But it would have been possible for some of these countries to adopt rates that, in their ultimate working out over a longer period of years, would have added excessively to their reserves. Such rates would have involved competitive depreciation at the time they were adopted, since the test lies in the ultimate effects.

  • (b)A country that is losing reserves because the market for its main export commodity (e.g., coffee) has been drastically curtailed by depression abroad may be guilty of competitive depreciation even though devaluation fails to stop the immediate loss of reserves. In devaluing its own rate, it may force depreciation on its competitors; and the depreciated rates of the whole group may diminish, rather than increase, the value of coffee sales on the depressed world market. If the world market for coffee were permanently contracted, the depreciation might be in order because it would then be necessary to stimulate other types of exports and curb imports in the countries that had hitherto placed their main dependence on coffee. Since, however, the market has been contracted by a merely temporary depression abroad, this basic reorientation of the economies of the coffee-producing countries is not needed. With the upswing of the business cycle, the demand for coffee will return and the old exchange rates will prove correct (if they had been correct initially). Hence an exchange devaluation that would be futile during the depression because it would not be effective in the short run, and that would prove to be an overadjustment in terms of the long run, would constitute competitive depreciation.

  • (c) In the two cases discussed above, the evidence of competitive depreciation lies in the ultimate effects of the exchange adjustment as contrasted with the immediate effects. Ultimately the over-devaluing countries would add excessively to their reserves. It is possible, however, to conceive of a situation in which an over-devaluation would not be accompanied, even ultimately, by an increase of reserves. Other countries in defending their positions might themselves devalue, or deflate internally, or apply exchange controls and quotas to the exports of the devaluing country, thus preventing it from pulling reserves from them. The fact that they offset an excessive devaluation in this way would not make it any the less excessive. 2 The devaluation would constitute competitive depreciation on the part of the originating country.

  • (d) On the other hand, it might be possible for a country to devalue without having a balance of payments deficit and without being guilty of competitive depreciation. Such a country might have a program for absorbing its unemployed or for expanding production through greater investment. It might anticipate that the program would increase domestic income and create a deficit in its balance of payments unless precautionary steps were taken. If the program were being effectively implemented, a corresponding exchange adjustment, even though it somewhat anticipated the expansion of income, would not constitute competitive depreciation. It would not result over a period of years in an undue expansion of the country’s reserves.

  • An anticipatory move of this character would not be justified, however, if the country were merely expanding along with the rest of the world. Nor would it be in order as a device for stimulating employment in a depression at the expense of other countries, for the rate would then prove to be undervalued in the long run.

  • (e) A more difficult question is raised if the devaluing country has no unemployment or industrial slack, but has a tendency toward inflation. Instead of an orderly growth in production, as in the previous case, there is a threat of a disorderly rise in prices. Conceivably the devaluation would set in motion a train of events that would drive costs correspondingly higher without materially changing the competitive position of the devaluing country. There would then be little evidence of competitive depreciation.

    The devaluation would, however, be quite useless and the Fund might hesitate to concur in it since it would subject the country to an unproductive inflation. Moreover, in practice, a rise in domestic costs all along the line, just sufficient to offset the effect of the devaluation on individual export prices, would be highly improbable. Unless there were a continuous series of devaluations, they could not keep just abreast of the developing inflation; and, even if they did so in their over-all effects, there would almost surely be changes in the competitive position of the country in one commodity market or another. Thus a confused situation would exist in which the country might be gaining a competitive advantage at one period or in one direction and losing it in another. There would be unsettling effects on other countries. Since these unsettling effects could not be justified as a necessary accompaniment of progress toward a valid economic objective, such as expansion of production or employment, the exchange rate adjustment might well be regarded as excessive. The degree of competitive depreciation involved in the adjustment, however, would clearly be only partial and uncertain. Theoretically, as noted in the previous paragraph, it might not exist at all.

  • (f) A case somewhat similar to that of a devaluation matched by the inflation it helps to set in motion would occur if a country adopted an export tax (or other export deterrent) sufficient to offset the effect on exports of the exchange adjustment. Such a combination might be used if the country wished to employ devaluation only to curtail imports. If the curtailment of imports brought the country back into balance of payments equilibrium (or brought it a surplus justified by its inadequate reserve position), it would be difficult to charge the country with competitive depreciation. As long as the tax and other conditions remained the same, the exchange rate would be reasonable even though without the tax it would stimulate exports and create an excessive rate of acquisition of international reserves. While adoption of such a pair of offsetting measures might enable the country thereafter, without having to seek the concurrence of the Fund, to reduce its tax and thus render the devaluation excessive, and while this possibility might affect the Fund’s attitude toward the proposal in the first place, the depreciation could not in fact be regarded as competitive unless this second stage was reached.

Dependence on broad approximations

General though the preceding discussion has been, it has undoubtedly implied greater exactitude than can ever be attained in arriving at working decisions. The effects of a devaluation on trade, services, and capital movements of the devaluing country and of the world; the period relative to which the effects should be considered; the appropriate distribution of reserves—all these must be matters of judgment to a high degree. Only partial light can be thrown upon them by the data and statistical analysis now available. This paper is aimed not so much at developing a precise formula as it is toward suggesting the general framework within which judgment must be applied. It is evident that it will always be difficult to prove that a proposed depreciation is competitive. Perhaps the greatest safeguard is that there is no real advantage to a country in indulging in it—except possibly during a world depression. Few will wish to turn the terms of trade against themselves in order to accumulate relatively unneeded reserves.

Note on multiple rates and competitive depreciation

The discussion of competitive depreciation in this paper has been on the assumption of a single rate. A single rate involves nondiscrimination among countries and commodities and, therefore, raises only the basic issues. With multiple rates (or broken cross-rates), it is evident that competitive depreciation might be practised via one rate or another without any effect on reserves even though no defensive measures were taken by other countries. A relatively low rate for one class of commodity (or country) might be counterbalanced by a relatively high rate for another. That was the essence of the Schachtian system in the 1930’s, which is often cited as the classic example of competitive depreciation.

This type of competitive depreciation, however, is merely one aspect of the problems created by discriminatory rates and can best be treated in connection with an over-all discussion of these problems. Such a discussion will not be attempted here.


Mr. Gardner was Chief of the International Section of the Federal Reserve Board before becoming Chief of the Balance of Payments Division of the Fund. Recently he has been appointed Assistant Director of the Research Department.

Mr. Tsiang, economist in the Balance of Payments Division, is a graduate of the London School of Economics, and was formerly Professor of Economics in the National Peking University and the National Taiwan University. He is the author of several articles in the Economic Journal and Economica.


Some acquisition of foreign assets by private residents of the devaluing country may be subject to the control of the monetary authorities to such an extent that they should be considered as equivalent to reserves. Furthermore, foreign monetary authorities may make some use of their accumulated reserves in the devaluing country. While such use would be an evidence of balance of payments surplus in the devaluing country, it would take the form of a repayment of its foreign debt. An exchange adjustment that would permit such repayment of short-term debt, as well as moderate prepayments of the country’s long-term foreign debt and a reasonable accumulation of reserves, would appear to be justified, even though a surplus financed by loans and grants from the devaluing country would indicate an excessive exchange adjustment.


This case should be distinguished from that discussed earlier in which a correct exchange adjustment was followed by various readjustments abroad. It was there noted that, whether or not these responses were appropriate, the devaluing country would have to accept the situation as it was and adjust its rate accordingly. Presumably there would be enough freedom left in the markets of the world for its devalued rate to be effective. That assumption had been made at the outset of the discussion. There was no assumption, however, that an excessive devaluation would be allowed to work out its full effects.