Journal Issue
Finance & Development, March 1979

Exchange rate policy:some current issues: An examination of some technical issues arising out of the recent instability of exchange rates

International Monetary Fund. External Relations Dept.
Published Date:
March 1979
  • ShareShare
Show Summary Details

Morris Goldstein and John H. Young

Under the Second Amendment of the Fund’s Articles of Agreement each member country undertakes to collaborate with the Fund and other members to assure orderly exchange arrangements and to promote a stable system of exchange rates. The Articles emphasize that stability at the national level is a necessary condition for a stable international monetary system and, subject to qualifications, each member agrees to foster orderly economic growth with reasonable price stability.

While substantial efforts have been and are being made to achieve greater stability, there is still a significant difference between the objectives of the amended Articles and the state of the world economy. With rates of inflation differing substantially from country to country, and with large current account imbalances persisting among a number of major countries, it is not surprising that exchange rates have at times been subject to large changes and substantial short-run fluctuations.

There is widespread agreement that the increased flexibility of exchange rates has been an important part of the adjustment process, but the size of exchange rate movements in recent years has nevertheless created difficulties for members. As might be expected, members have differing views on the benefits of exchange rate flexibility. Many of the less developed countries that peg their exchange rates to some currency or a combination of currencies, and some of the smaller members of the European common margins arrangement (the “snake”), do not, for example, share the attitude of some of the larger Fund members toward floating. These differing views are outlined and discussed in Chapter 2 of the Annual Report for 1978, and this article summarizes the analysis of some of the technical issues considered in that Report. In particular, consideration will be given here to the extent to which the world has a pegged or floating system, the dispersion of inflation rates, the short-run variability of exchange rates, the concerns of developing countries, and the adjustment process.

A pegged or floating system?

One question that is frequently asked is the extent to which the present system should be categorized as a floating system. After all, on October 31, 1978, the great majority of Fund members, 95 out of 134, were classified as having pegged rates. Sixty-four were pegged to a single currency, 13 to the special drawing right, and 18 to some other composite of currencies. On the same date, members with floating currencies included 6 that maintained common margins, 5 that adjusted exchange rates for their currencies according to a set of indicators, and 28 that maintained other types of exchange arrangements, including regimes that are described as floating independently. Thus, if the measure is in terms of Fund membership, the system is predominantly pegged.

If, however, the measure is based on the extent to which trade is conducted by countries with pegged or floating rates, the volume of trade of individual members becomes important. Since almost all of the largest countries are categorized under some form of floating, it turns out that less than one fifth of all trade is carried out by members that are classified as having pegged rates. This is, however, a rather crude method of measuring how much trade is carried out across pegged or floating rates. A more effective method is to measure the value of trade that is actually conducted at pegged and floating rates. For example, for those countries that are pegged to the U.S. dollar, only their trade with the United States, and with other members whose currencies are pegged to the U.S. dollar, will be conducted at fixed rates; their trade with other countries will be carried out at floating rates, as the U.S. dollar fluctuates vis-à-vis other currencies. In the same way, members of the European common margins arrangement conduct trade both with countries that are inside the snake and with others outside it. Although the proportion varies for individual countries, the result is that, on average, less than one third of the total exports of members of the snake is conducted at pegged rates and the rest is carried out at floating rates.

Analyzing world trade flows in this way shows that less than one fifth of all trade moves across pegged exchange rates—a proportion similar to that indicated by simply classifying the trade of countries by the exchange rate arrangements of the country. The similarity in the results arises from the fact that the floating rate trade of members classified as having pegged rates happens to be roughly equal to the pegged rate trade of members with some form of floating currency.

Dispersion of inflation rates

Aggregate price indices for the various groups of countries show that the exchange rate system has been operating in recent years in the context of high and variable inflation. The average rate of inflation for the 14 industrial countries, as measured by the deflators of gross national product, was about 12 per cent in 1974, 11 per cent in 1975, and was still close to 7 per cent in 1977 and 1978. This may be compared with an average inflation rate for these countries in the period 1961–64 of about 2.5 per cent and in 1965–69 of 3.7 per cent. The inflation experience for other groups of countries over recent years has been even less encouraging, with average inflation rates in the non-oil developing countries, for example, having been roughly twice as high as in the industrial countries.

Of perhaps even greater relevance for the operation of the exchange rate system is the fact that the period 1974–78 has been one where rates of inflation, growth of real output, and monetary expansion have differed substantially from country to country. It is the intercountry differentials in these variables, rather than their behavior within each national economy alone, that are fundamental in determining exchange rate and reserve changes. An indication of the size of recent and past intercountry inflation differentials can be obtained from Table 1, where a measure of the dispersion of inflation among the seven major industrial countries is computed for the years 1960–73 and 1974–78. Two observations can be made about the figures given. First, the dispersion of inflation among the major industrial countries has been much larger (by a factor of about three) in the period of floating rates than during the period 1960–73. Second, the experience of the years since 1973 suggests that intervals of high average rates of inflation, which are also likely to be periods of large inflation differentials, carry a greater probability of large changes in the structure of exchange rates.

Short-run variability

The structure of exchange rates was expected to alter in response to the dispersion of inflation rates, but some exchange rate changes in the recent period have been very abrupt, reversible, and difficult to relate to underlying economic conditions. Average daily changes in selected currency rates against the U.S. dollar in the period 1976–78 are shown in Table 2. The influences operating on exchange markets are too numerous and diverse to permit a single explanation for recent exchange rate variability. It is possible, however, to identify some causal factors. For countries with well-developed money and capital markets, conditions in the financial markets are probably more important than those in the goods markets for determining short-run exchange rate movements. Indeed, the foreign exchange market can be thought of as an asset market, and the exchange rate between two currencies regarded as a relative asset price that moves with changes in the relative supply and demand for assets denominated in those currencies. As with other assets, current rates of return, risk factors, and expected future rates of return are important in determining the current price, and when the factors affecting these returns and risks fluctuate substantially, so too does the current price, that is, the exchange rate. In other words, periods of rapidly shifting interest rate differentials, sudden imposition or relaxation of capital and exchange controls, and changing exchange rate expectations are apt to be periods of large short-term exchange rate variability, even when relative prices move only slowly.

The influence of exchange rate expectations or, more correctly, changes in exchange rate expectations are perhaps worthy of special note because of the many factors influencing these expectations, and because the factors themselves are subject to frequent change, especially in an environment of high inflation and irregular economic growth. While the factors affecting exchange rate expectations are not directly observable, it is known that they include, inter alia, monetary and fiscal policies, relative cyclical positions, current account and trade account imbalances, inflation differentials and relative competitive positions, political uncertainties, official intervention in the exchange market, and, of course, the change in the exchange rate itself. The reason changes in expectations are particularly important for exchange rate variability is that the current exchange rate may already fully reflect all publicly available information. Hence, it is only new and unexpected information that will cause market participants to change their evaluation of the future exchange rate, and thus the present exchange rate as well.

Table 1Seven major industrial countries: 1 dispersion of Inflation rates, 1960–70 average and annually for 1971–78
Mean rate of

inflation 2
Dispersion of

inflation 3
1960–70 average3.61.5
1960–73 average4.21.5
1974–78 average11.04.6
Source: IMF, International Financial Statistics.

Canada, France, the Federal Republic of Germany, Italy, Japan, the United Kingdom, and the United States.

The (unweighted) average of annual rates of change of consumer prices in these countries.

The measure of dispersion used is the standard deviation of inflation rates among the seven countries.

Source: IMF, International Financial Statistics.

Canada, France, the Federal Republic of Germany, Italy, Japan, the United Kingdom, and the United States.

The (unweighted) average of annual rates of change of consumer prices in these countries.

The measure of dispersion used is the standard deviation of inflation rates among the seven countries.

Concerns of developing countries

The short-run fluctuations in exchange rates in recent years have caused problems for the less developed countries, despite the fact that most of them continue to peg their exchange rates. For those countries that peg to a single currency—and most of the less developed countries are in this category—greater exchange rate variability between the intervention currency and other currencies is likely to result in increased variability in both the country’s effective exchange rate and in the local currency price of its imports and exports. Increased short-run fluctuations among the major currencies also may mean that a less developed country’s exchange rate (vis-à-vis countries with which it conducts an important part of its trade) responds to factors more closely associated with the external position of the country issuing its intervention currency than to its own domestic or balance of payments needs.

Some less developed countries have attempted to minimize these problems by switching from a unitary peg to a peg based on a basket of currencies, but many countries find this solution administratively inconvenient, particularly when there is a single dominant currency used in trade and exchange transactions. In addition, a pegged rate is regarded in some less developed countries as providing, in their case, a viable framework for encouraging price stability, for fostering confidence in the authorities’ financial policies, for helping their planning process, and for promoting investment. In brief, the choice of an appropriate peg has become a difficult one for less developed countries in today’s exchange rate environment.

The increase in exchange rate fluctuations has also caused problems of portfolio management for the less developed countries, most of which hold nearly all of their foreign exchange reserves in a single currency. While the fluctuations in exchange rates have diminished the store-of-value function of some of the major currencies, the rise and variability in import costs have led to a demand for higher and more assured levels of reserves. Those less developed countries that peg to a single currency whose future value is uncertain may therefore face the dilemma that they need to hold larger working balances in that currency, yet they may also wish to diversify their reserves.

The adjustment process

There is considerable empirical evidence that relative price changes have a strong influence on the volume of imports and exports. Time is needed, however, for the requisite changes in demand and production to take place, so that only some fraction—say, one fourth to one half—of the ultimate volume effects will be observed over a period as short as a year. It is likely to take time for some consumers to switch from traditional suppliers with whom they may have long-term contracts to producers located in a country that has devalued its exchange rate. In addition, it may be neccessary for producers to set up new marketing networks in countries to which they have not previously exported. In contrast to these volume effects, exchange rate changes can affect a country’s terms of trade rather rapidly. As a result of this asymmetry of timing, the trade balance generally deteriorates before it begins to move steadily in the expected direction, leading to a “J” curve of the kind described on page 10.

Fund staff estimates of the short-run effects of exchange rate changes on the terms of trade for industrial countries indicate that the first few months following a depreciation are generally marked by a much faster rise in import prices in local currency than in export prices in local currency. Essentially, these changes in the terms of trade result from the fact that many countries do not face a given world price for their exports. When such world prices are of dominant importance, as they are for many less developed countries, changes in the terms of trade as a result of an exchange rate change are likely to be small. Changes in the terms of trade are generally more pronounced for industrial countries, the export products of which are differentiated from those of other countries.

Thus, while a country that imports products with prices fixed in terms of foreign currency experiences an immediate increase in the local currency price of its imports following a depreciation, exporters may raise their prices in terms of local currency only partially or with a delay. On the import side, a switch from foreign-produced goods to those produced by the country with a depreciated exchange rate takes time if, as noted above, sources of supply are established and contracts are in effect. Production conditions may be affected fairly quickly by the increased cost of imported inputs, but wage responses to changes in import price rises may be slow, depending on how rapidly import prices affect consumer prices and on the extent of indexation. Even one year after a depreciation of a country’s currency, a discrepancy between the rise in import and export prices has frequently been observed. Staff estimates show that such changes in the terms of trade often have a substantial effect on trade balances, at least for the major industrial countries.

Seen in the light of the preceding discussion, recent trade and current account developments appear less surprising. The snake currencies, the Swiss franc, the Japanese yen, and the pound sterling appreciated against the U.S. dollar and other currencies during 1977. These appreciations were generally associated with larger trade surpluses (or smaller trade deficits) for the appreciating currencies and larger trade deficits for the United States and some of the other depreciating countries during 1977, although other factors, including volume effects from earlier relative price changes, offset this to some extent. Given the sharp movement of exchange rates in 1977 and the early months of 1978, it would not be surprising if perverse effects of the terms of trade again dominated the initial volume effects and produced a widening of trade imbalances in the short run.

Table 2Average daily changes in selected currency rates against the U.S. dollar 1(In percent)
Germany, Fed. Rep. of0.
United Kingdom0.240.460.280.550.

Average percentage change from previous day in spot exchange rates against U.S. dollar (New York noon quotations).

Average percentage change from previous day in spot exchange rates against U.S. dollar (New York noon quotations).

Staff estimates of the volume effects of an effective exchange rate change of 10 per cent for the 14 industrial countries indicate that the volume of imports and exports is always altered in the expected direction, but that the size of the response differs from country to country. For example, the response of the volume of imports in the 14 industrial countries to a change in the exchange rate of 10 per cent varies from 1 to 6 per cent. The response of the volume of exports varies from about 7 per cent in France, Japan, Sweden, and the United States to only 3 to 4 per cent in the remaining countries, where either export demand or supply is less price elastic, or the degree of openness of the economy leads to large feedbacks from exchange rate changes to domestic costs and prices.

The implication of those estimates is that, other things being equal, the exchange rate changes in 1977 and the early months of 1978 should lead to a substantial improvement in the pattern of trade balances among industrial countries during the next two to three years. For example, if changes in relative prices adjusted for exchange rates up to the middle of 1978 are considered, staff estimates suggest that by 1980 these will lead to a significant improvement in the U.S. trade balance and a major reduction in the surpluses of Japan and the Federal Republic of Germany.

The J-curve

The term “J-curve” refers to the shape of the adjustment path frequently followed by the trade balance of countries in response to an exchange rate devaluation. The initial impact of a depreciation is often negative because import prices rise more rapidly in local currency than export prices, and there has not been time for the volume of trade to adjust. After a lag, however, the trade balance improves with a reduction in the rate of growth of imports, a parallel rise in the rate of growth of exports, and a reduction in the gap between the price indices of imports and exports. As a result of these factors there is an initial decline in the trade balance, but this adverse movement will be first checked and reversed, leading the trade balance to follow the rising portion of the “J.” The economic factors that lead to this result are discussed in the article in the section on the adjustment process.

The chart illustrates the cumulative quarterly effects of a 10 per cent devaluation in the exchange rate of a country on the unit values and volumes of its exports and imports and on its trade balance over a period of three years. The following arbitrary assumptions are made: (1) export unit values in local currency respond to the devaluation at a rate of 1.5 per cent per quarter over a period of four quarters and thereafter at a rate of 0.25 per cent, adding to 8 per cent three years after the devaluation. (2) Import unit values in local currency respond at a rate of 7 per cent in the first quarter following the devaluation and at a rate of 3 per cent in the second, reflecting a complete pass-through after two quarters. (3) After half a year the volumes of both exports and imports begin to respond at a rate of 1 per cent per quarter with the cumulative response extending over two and a half years and adding to 10 per cent three years after the devaluation.

J-curve: an illustration of the effects of a 10 per cent depreciation on the trade balance1

1 The price and volume effects of the depreciation are shown as cumulative quarterly per cent changes (left scale). The price and volume effects of the depreciation on the trade balance are expressed in units of local currency at an annual rate {right scale).

Illustrative calculation of “J” curve: price and volume effects of a 10 per cent depreciation(In billions of local currency units)
QuartersAssumed rate of change in import prices(Cumulative per cent)Assumed rate of change in import volume(Cumulative per cent)Value of imports(L/C units)1Assumed rate of change in export prices(Cumulative per cent)Assumed rate of change in export volume(Cumulative per cent)Value of exports(L/C units)1Trade balance(L/C units) 1
III10.0–567.76.75569.6+ 1.9

Local currency units.

Local currency units.

Other Resources Citing This Publication