Private Capital Movements and Exchange Rates in Developing Countries

This is a revised text of a paper presented to the International Economic Association, Round Table Conference on Capital Movements and Economic Development, Washington, D.C., July 28, 1965. It will also appear in the proceedings of the Conference published by The Macmillan Company, London, and St. Martin’s Press, New York.1


This is a revised text of a paper presented to the International Economic Association, Round Table Conference on Capital Movements and Economic Development, Washington, D.C., July 28, 1965. It will also appear in the proceedings of the Conference published by The Macmillan Company, London, and St. Martin’s Press, New York.1

THE ECONOMIC AND FINANCIAL literature of recent years has contained a considerable amount of discussion of the relationship between exchange rates and international differences in interest rates, on the one hand, and capital movements between the major industrial countries, on the other. Much less attention has been paid to the question of the effect of variations in exchange rates, or of the choice among alternative exchange rate systems, on private capital flows from industrial to developing countries. The present paper deals primarily with this relatively neglected topic. The principal question that will be considered is the following: how are private capital movements in any one country likely to respond to changes, or expected changes, in foreign exchange rates and in the country’s prices and costs, given the other factors which influence the incentive to invest in that country? Although the paper is chiefly concerned with flows of private capital to developing countries, the general analysis of the influence of changes in prices, costs, and exchange rates on movements of various types of capital applies also to international investment in industrial countries. The concluding section elaborates some implications of the analysis with regard to the broader question of the effect of alternative exchange rate systems (fixed or fluctuating exchange rates, unitary or multiple exchange rates) on capital flows.

These questions must be approached mainly through a priori reasoning. Countries that differ insofar as their exchange rate systems are concerned ordinarily differ also with respect to a number of other factors influencing the attractiveness of their economies to foreign investors. Many of these factors cannot easily be quantified. Moreover, investment responds not to the actual changes in the determinants of returns on capital assets but rather to the changes which potential investors expect; the historical record reveals the former but not the latter. For these reasons, the scope for statistical testing of theoretical propositions concerning the determination of the extent, timing, direction, and composition of foreign investment is very limited. The empirical material presented in this paper is intended only for illustrative purposes.

Exchange Rates, Inflation, and Capital Movements

If a country in which investment is contemplated is plagued by rapid domestic inflation, the exchange rate must be expected to depreciate sooner or later. It is generally agreed that the expectation of the depreciation of a currency tends to discourage foreign investment in assets with given prices and yields denominated in that currency. This is so because of the exchange loss which a potential investor has to expect whenever the proceeds of some of the earnings and of the ultimate liquidation of an asset have to be converted into the investor’s currency at a less favorable exchange rate than that at which the asset was originally purchased. These considerations, which concern a particular type of foreign investment, are sometimes applied to foreign investment in general. This leads to the view that inflation and exchange instability (i.e., repeated devaluations or a continuously depreciating exchange rate) in the investee country reduce earnings on foreign investments and thus discourage capital inflows and encourage capital outflows.2

It is, however, difficult to reconcile the presumed deterrent effect of inflation and exchange depreciation with the continued large inflows of private capital into some of the countries whose currencies show very rapid rates of loss of internal and external purchasing power. Net inflows of private capital into four less developed countries—Argentina, Brazil, Chile, and Colombia—during the period 1951-63 are shown in Table 1. These countries experienced rapid inflation (measured by the cost of living index) and exchange depreciation (measured both by the exchange rates mainly applicable to capital transactions and by those mainly applicable to general commercial transactions). They were, in fact, selected because, among the countries for which the relevant data could be compiled, they had the highest rate of inflation. During the 13-year period 1951-63, the net inflow of private long-term capital to the four countries totaled $4.6 billion. This inflow contributed to the foreign exchange earnings of these countries an amount equal to 10 per cent of their earnings from merchandise exports. They experienced a net outflow of short-term capital (including errors and omissions)3 of about $1.3 billion; however, the net inflow of private capital of all types was quite large ($3.3 billion), amounting to 7 per cent of their combined exports.

Table 1.

Selected Developing Countries with High Rates of Inflation: Net Private Capital Inflow and Changes in Cost of Living and Exchange Rates, 1951-63

article image
Sources: Capital movements, from International Monetary Fund, Balance of Payments Yearbooks; cost of living index, exchange rates, exports, from International Monetary Fund, International Financial Statistics, monthly issues and Supplement 1964/65.

Including net errors and omissions. A minus sign indicates a net capital outflow.

Direct and other long-term investment.

Annual averages; computed changes are from the average of 1950-51 to that of 1963-64, from the average of 1950-51 to that of 1957-58, and from the average of 1957-58 to that of 1963-64.

Exchange rates applicable to capital movements at the end of the year; computed changes are from year-end 1950 to year-end 1963, from year-end 1950 to year-end 1957, and from year-end 1957 to year-end 1963.

Exchange rates applicable to most commercial transactions (excluding principal export commodities) at the end of the year; see also footnote 4 above.

A comparison between two subperiods shows no consistent evidence—country by country—that long-term capital inflows were smaller in more inflationary periods than in less inflationary ones. In three of the four countries, prices rose much more rapidly from 1958 to 1963 than from 1951 to 1957. Yet during the 6-year period 1958-63, the net private long-term capital inflow into the four countries amounted to the equivalent of $3 billion (15 per cent of export earnings), almost one fifth of the net private long-term capital flow (including long-term private export credits) from the member countries of the Organization for Economic Cooperation and Development (OECD) to all less developed countries.4 This more inflationary period was also characterized by a substantial outflow of short-term capital from these four countries; it amounted to $1.1 billion, compared with only $150 million for the 7-year period 1951-57. This short-term capital outflow constitutes a loss of foreign exchange resources which these countries can ill afford. Nevertheless, what is surprising—in view of the financial instability of the four selected countries—is not the size of the net short-term capital outflow but rather the large magnitude of the net long-term capital inflow.

In attempting to explain why financial instability was not more of a deterrent to private foreign investment in these countries, the following section draws attention to the manner in which domestic prices and foreign exchange rates enter into the calculation of expected yields on foreign investment. To be sure, capital movements respond to a large number of factors which affect the yields expected by investors. Some of these, such as changes in the structure of the economy of the investee country, discoveries of mineral deposits, changes in exchange control or tax systems, the passage of investment laws, the amount of—and experience with—previous investments in the country, and, not least, the country’s potential for economic growth,5 are very likely to play a more dominant role in the investment decision than expectations of changes in prices or exchange rates. But where these changes are large, they must, after all other elements have been taken into consideration, exercise an important influence on the decisions of potential investors.

The Return on Foreign Investments

The expectation of changes in the foreign exchange rate and in domestic prices and costs influences private capital movements, other things being equal, by altering the expected rate of return on assets which foreign investors may hold in the investee country. The next section will deal with the formation of expectations; the present section sets forth the relation between changes in prices and exchange rates, on the one hand, and the rates of return on different types of foreign investment, on the other.

In some instances, depreciation of the investee currency will reduce, in proportion to the depreciation, the net revenues obtained from an asset, computed in terms of the investor’s currency, while domestic price increases in the investee country will not affect these net revenues. This is true in the special case of investment in a fixed-interest asset, such as a bond, denominated in the investee country’s currency. This special case has often served as a starting point for the discussion of the effect on capital movements of expected changes in the exchange rate. However, insofar as capital flows from industrial to developing countries are concerned, investment in local-currency bonds is relatively unimportant. In the quantitatively much more important case of direct investment, the relation between the return on capital and the exchange rate is more complicated. Depending on the circumstances of particular industries, the returns on direct investment capital may be either raised or lowered by a depreciation of the investee currency; similarly, they may be either raised or lowered by a rise in domestic prices in the country where the investment is made.

The analysis of the effects of price changes and changes in the exchange rate may be set out in general terms for any kind of foreign investment and with allowance made for different exchange rates for various classes of transaction (multiple exchange rates). In the following discussion, the effects of exchange control provisions are left out of account; in particular, it will be assumed that earnings and capital can be freely repatriated at exchange rates applicable to these transactions. It will also be assumed that prices in the investor’s country of residence remain unchanged and that the currency of that country is freely convertible at a constant par value. For simplicity’s sake, the investor’s currency will be called “dollar” and the currency of the investee country “peso.”

The rate of return, or yield, y, of an asset is the rate of discount which makes the sum of expected future net revenues, N, equal to the present price (or cost of construction), A, of the asset.6 The yield can be computed from the formula


For a given present price, A, the yield will become larger or smaller when future net revenues become larger or smaller. In order to determine the influence of changes in prices and exchange rates on the yield of the asset, it is necessary to ascertain the effects of these changes on net revenues or on the present price of the asset or on both.

As mentioned above, in the special case of a bond denominated in pesos, the net revenue expressed in dollars is simply the amount of the annual coupon earnings of the bond, and in the year in which it matures also the face value, divided by the exchange rate (defined as the price of one dollar in terms of pesos). In the general case, however, various components of the net revenue from a foreign asset may be differently affected by changes in prices and exchange rates. For instance, gross receipts from sales of an industrial or commercial enterprise may accrue partly in pesos and partly in dollars. Many companies will incur certain costs in pesos and others in dollars.

Consider, for instance, the following elements affecting the net revenue of a production enterprise (the peso values are separated into “real” and price components):

article image

Net revenues in dollars, N, for a particular year of operation can be expressed as


In general, domestic inflation or exchange depreciation will tend to affect in two ways net revenues expressed in dollars: (1) it will alter the valuation of given “real” transactions, such as S, C, S$, and (2) it will induce changes in these real transactions in response to the changed price or cost relationships. The argument set forth in this paper relies on the effects under (1) and leaves out of account those under (2).

In regard to changes in valuation, inflation in the host country will—if exchange rates are assumed to remain constant—tend to increase the dollar equivalent of receipts from local sales (Sps), of local costs (Cpc), and probably also of local taxes (Tpt). If, on the other hand, constant peso prices are assumed, depreciation of a unitary exchange rate (r=rx=rm=rc) will tend to reduce the dollar equivalent of local sales, costs, and taxes.8 Under a unitary exchange rate (or under multiple exchange rates which change in the same proportion), a depreciation of the peso will not affect the dollar equivalent of given transactions which are inherently denominated in dollars, namely, proceeds from exports or costs of imported materials or services.

The effects of changes in the price and cost relations on the real transactions, whose analysis would considerably complicate the presentation, can be left out of account for two reasons. First, the present paper aims only at an identification of the direction, and not of the magnitude, of the influence on asset yields of changes in prices, costs, and exchange rates. Except under unusual assumptions, the adjustments which a firm may make in its transactions in response to price changes9 are unlikely to alter the direction of changes in asset yields which would have resulted from such price changes in the absence of these adjustments. For instance, a rise in the price of materials produced in the investee country will, other things being equal, directly lower the net revenue of manufacturing enterprises that use these materials as inputs. To the extent that similar materials can be obtained from abroad, production costs may rise less than they would have in the absence of such substitution possibilities, but they will neither decline nor ordinarily remain constant.10 In general, the “impact effect” of a price change on the net revenue of a company will indicate the direction, though not necessarily the magnitude, of the ultimate effect after the company has adjusted its transactions to the altered relative prices.

Second, under conditions prevailing in many less developed countries, the error introduced by judging the magnitude of the ultimate effect of a price change from its impact effect may not be very large. Substitution of foreign for domestic materials, intermediate products, or services in the production process is in some instances not possible at all; in others, it is limited to a narrow range of cost items. Moreover, companies producing for the local market are often sheltered from foreign competition by quantitative import controls and are not, themselves, competitive in international markets. As a result, changes in the relation between the exchange rate and domestic prices will not ordinarily induce much substitution of competing imports for the products of local foreign investment enterprises or vice versa, nor large shifts of their sales between home and foreign markets, at any rate not in the short run.11

For these reasons, the following discussion will consider the effects of changes in prices, costs, and exchange rates on the net revenue of an asset (enterprise) under the simplifying assumption that the real transactions, the volume of sales, the employment of labor, the purchase of domestic or foreign materials, etc., remain unchanged. It may also be assumed for the moment (an assumption made only for expository convenience and relaxed presently) that domestic inflation raises all relevant peso price and cost indices in the same proportion (p=ps=pc=pt). Under a unitary exchange rate, formula (2) can then be simplified to


where Np represents the local net revenue in pesos (S–C–T), N$ stands for the net revenue earned in dollars (S$C$), and V$ is, as before, the working capital valuation adjustment.

Whether inflation or exchange depreciation will tend to raise or lower net revenues will depend, under the simplifying assumptions of equation (3), on the difference, Np between peso revenues and costs (including taxes) and on the working capital valuation adjustment. Companies producing chiefly for the local market will ordinarily have positive net peso revenues, while enterprises whose output is mainly exported will tend to have negative net peso revenues, i.e., their peso costs and taxes will exceed any peso revenues which they may derive from a small volume of local sales. An illustration of the relative magnitudes of net revenues accruing in local currency and in dollars—Np and N$, respectively, in equation (3)—is given in Table 2 for all U.S. direct investment enterprises in the Latin American Republics in 1957, the most recent year for which these survey data are available. In that year, net revenues accruing in local currency were positive in manufacturing and in public utilities, but negative in the petroleum industry, mining, and agriculture,12 and for all industries taken together.

Table 2.

U.S. Direct Investment Enterprises in Latin America:1 Sales Receipts, Costs, and Net Income in Local Currency and in Foreign Exchange, 19572

(In billions of U.S. dollars)

article image
Source: Department of Commerce, U.S. Business Investments in Foreign Countries, Supplement to the Survey of Current Business (Washington, 1960), pp. 90 and 147.

Twenty republics.

Items may not add to totals because of rounding.

Imports other than capital equipment and fees paid abroad. The import data are incomplete; costs in foreign exchange may therefore be understated, and costs in local currency overstated.

In the report from which the data for this table have been taken, depreciation and depletion allowances are shown as U.S. dollar equivalents of amounts reported by companies in local currency. To the extent that proper allowance for price increases of equipment or exchange devaluation is not made in the reported figures, total net revenues after depreciation and depletion allowances may be overstated.

No separate allowance is made in this table for the working capital valuation adjustment, V$, except to the extent that this item may be included under another heading in the reported accounts of the individual companies summarized in this table.

In the absence of exchange rate changes, inflation will raise the net revenue—and thus the rate of return on the invested capital—of enterprises selling mainly in the local market (for instance, manufacturing companies), provided that their selling prices rise at least in step with their wage and other costs and their tax payments. This effect will be particularly pronounced if an appreciable portion of their current cost items is imported from areas in which stable prices prevail. Net revenues of export-oriented firms, on the other hand, will tend to be reduced by inflation. Their local costs will rise, but their sales prices will remain constant.

The effects of exchange depreciation are just the reverse of those resulting from inflation. The dollar equivalent of the net revenue accruing in pesos will be reduced. If this net revenue is positive, for instance in a manufacturing subsidiary, the return on capital will decline, but if it is negative, e.g., in a mining company selling its output at a given world market price, the return on capital will be increased. Moreover, the portion of the working capital subject to exchange valuation adjustment of an export-oriented enterprise will ordinarily be smaller than that of a company selling in the local market. Its accounts receivable are presumably denominated in dollars, and its inventories consist of products awaiting sale for dollars or of materials to be used in production which will be sold for dollars; in fact, if the company has accounts payable in pesos (e.g., tax liabilities), the valuation adjustment resulting from exchange depreciation may well be negative (indicating a valuation gain) rather than positive.

Regardless of whether a company is oriented toward exports or local sales, inflation accompanied by an exchange depreciation in the same proportion as the increases in those domestic prices and wages which are relevant for the operations of the enterprise will tend to leave net revenues expressed in dollars, and rates of return on capital, unaffected.13 If prices rise faster than exchange rates, so that the peso becomes increasingly overvalued, yields on investment in companies selling in the local market will tend to increase; if the opposite trend prevails, export-oriented enterprises will be favored.

The assumption that all relevant domestic prices and costs change in the same proportion was made merely for expository convenience. Equation (2), or an even more detailed formula that could easily be set out, permits the relaxation of this assumption. For instance, public utilities are oriented toward the local market, but their net revenues are not so likely to rise during inflationary periods as those of manufacturing enterprises, since price regulation prevents their prices or service charges from being increased in step with their costs in local currency.

The assumption of a unitary exchange rate can also be relaxed in the framework of equation (2). Under a system of multiple exchange rates, the investor’s dollar earnings resulting from net revenues which accrue directly in dollars may be affected by differential rate changes. A depreciation in the exchange rate applicable to the company’s exports, other rates remaining unchanged, would raise net revenues, while a depreciation of the exchange rate applicable to imports of goods and services or of that governing the repatriation of earnings and capital would lower it. To the extent that, as is usually the case, exchange rates for imports and the repatriation of earnings are higher (i.e., more depreciated) than those for a country’s principal exports, returns in export industries will be lower than they would be under a unitary exchange rate. As a result, foreign investment in these industries will be discouraged.

In order to give an indication of the magnitude of changes in the rate of return on foreign investments brought about by changes in prices and exchange rates, the results of a few calculations based on data from Tables 1 and 2 are shown in Table 3. These calculations refer to entirely hypothetical cases: they indicate the changes in returns from direct investments in selected industries in the four countries shown in Table 1 under the assumptions (1) that the distribution of revenues and costs accruing in dollars and pesos in these industries corresponds to that found in Table 2 for U.S. direct investment enterprises in the 20 Latin American Republics in 1957 and (2) that this distribution was not affected by the relative changes in prices and exchange rates between the dates shown. As indicated above (p. 8), the assumption under (2) is not likely to correspond fully to reality. To the extent that foreign investment companies can improve their position by adjusting their transactions in view of altered price relations, their net earnings will rise more, or decline less, than is indicated by the computation in Table 3.

Table 3.

Selected Countries and Industries: Proportionate Increase or Decrease (—) in Net Earnings in Dollars of Hypothetical Foreign Investment Enterprises as a Result of Changes in Domestic Prices and Foreign Exchange Rates1

(In per cent)

article image

Calculated in accordance with equation (2), page 7, assuming a (constant) distribution of revenues and costs like that given for corresponding industries in Table 2. It is assumed that local revenues, costs, and taxes rise in proportion to the cost of living. For Argentina and Brazil, dollar revenues and costs are converted into pesos by the index of exchange rates applicable to commercial transactions, and total earnings in pesos are then converted into dollars by the index of exchange rates for transfers of capital and earnings (see Table 1). For Chile and Colombia, a somewhat modified version of equation (2) was employed which reflects exchange regulations with respect to mining and petroleum companies, respectively, and makes use of the special exchange rates effective for the purchase of local currency by foreign companies in these industries in the early 1950’s.

Changes are based on data for December 31 of the first and last years of each period.

Chiefly owing to the very unfavorable exchange rate in 1950, at which foreign mining companies in Chile had to convert those portions of their export proceeds which are needed to meet local expenses, application of the price index and exchange rate for 1950 to the assumed cost and revenue distribution of 1957 results in a negative net revenue for 1950. The proportionate increase in the net revenue from 1950 to 1957 and to 1963 can, therefore, not be shown. (The exchange rate for conversion into local currency of the export proceeds of large mining companies increased—i.e., depreciated—by 3,500 per cent between the end of 1950 and the end of 1957, while the cost of living index rose by somewhat more than 1,000 per cent over the same period.)

In the comparison of hypothetical earnings at the end of 1963 with those at the end of 1950, the joint effect of changes in prices and exchange rates is positive in each of the selected cases. For instance, in 1963 the dollar equivalent of the earnings of a manufacturing company in Brazil would be 30 per cent larger than it would have been if—with its transactions in real terms remaining unchanged—domestic prices and exchange rates had remained at their 1950 (year-end) level. The earnings of a mining company in Chile at the end of 1950, calculated on the basis of the revenue and cost distribution given in Table 2, would have been negative; between 1950 and 1957 the position of foreign mining companies would have improved considerably, chiefly as a result of the discontinuation in 1956 of the particularly unfavorable exchange rate at which foreign mining companies had to obtain Chilean pesos to meet local operating expenses. In Argentina, the earnings of a foreign-owned manufacturing company would have increased both from 1950 to 1957 and from 1957 to 1963. In two of the other three countries, the direction of the net effect of changes in prices and exchange rates on earnings differs between the two subperiods. For instance, earnings of the export-oriented mining industry in Chile declined between 1957 and 1963 because domestic prices increased more rapidly than the exchange rate for commercial transactions, whereas this industry had been favored by a reverse development between 1950 and 1957. The dollar equivalent of earnings of a manufacturing company in Brazil decreased from 1950 to 1957 by virtue of an excess of the rate of exchange depreciation over the rate of inflation, and increased from 1957 to 1963 because of the reverse relationship between these rates of change. While these results indicate general tendencies in earnings on foreign investment in the cases examined, the hypothetical character of the calculations should be borne in mind.

In the preceding discussion of the expected rate of return on a capital asset, it was assumed that the price or construction cost of the asset, in terms of dollars, was constant. But the cost in dollars, A, of a given investment program will, itself, depend on the price level in the investee country and on the exchange rate. Suppose that A$ represents the cost of the equipment to be imported into the investee country plus any working capital to be supplied by the investor, and that Ap stands for the estimate in a base period of the local construction costs and other local expenses connected with the investment (A$+AP=A). Some time after this base period, the expected dollar cost of the entire investment project will be


where pc and rc are the expected index figures, relative to the base period, of local costs and of the exchange rate applicable to capital transfers.

The expectation of exchange depreciation may, by lowering the expected dollar cost, lead to a postponement of the capital expenditure until after the anticipated depreciation materializes (or until the expectation is revised).14 On the other hand, the expectation of a constant exchange rate and rising construction costs in the investee country may cause the investor to advance expansion programs which would otherwise have been undertaken at a later date. The extent to which such changes in timing are induced will, of course, also depend on the effect of the expected changes in local costs and exchange rates on the net operating revenue from the completed production facility.

The discussion of this section is summarized in Table 4. Nonequity investments can be treated, in the calculation of rates of return, as special cases of formula (2) above by setting the appropriate items equal to zero. The expectation of exchange depreciation brings about changes in expected yields which are likely to result in a reduction of the inflow of foreign fixed-interest capital and in an increase of the outflow of domestic capital from the country whose exchange rate is expected to depreciate. The net effect on equity investment is not certain and will depend on the distribution of suitable projects between export-oriented and local sales-oriented industries. There is a similar uncertainty with respect to the influence on foreign equity investment of the expectation of inflation—abstracting from the effect of inflationary expectations on expected exchange rates—but the net outflow of domestic capital of the inflating country is likely to be reduced.

Table 4.

Effects of Inflation and Exchange Depreciation on Rates of Return on Foreign Investment and Residents’ Investment Abroad

article image

Assuming a uniform exchange rate or a multiple rate system in which the relevant rates change in the same proportion.

It is assumed that companies are able to raise local sales prices approximately in proportion to increases in costs resulting from inflation.

Inflation will not affect domestic residents’ returns on foreign investment, but it will raise the (nominal) returns on alternative domestic equity investments.

Two points should be re-emphasized. First, the effects of inflation and of exchange rate changes on yields of various assets were separately stated without regard to the likelihood that inflation would induce the expectation of exchange depreciation and vice versa. The interdependence of expectations with respect to a currency’s internal and external purchasing power is briefly discussed in the following section. Second, the analysis was based on the assumption that exchange control is absent or, at any rate, that it does not impinge in a material way on the transactions dealt with in this paper except by prescribing the exchange rates applicable to various transactions.

Price and Exchange Rate Expectations

Contrary to the assumption made in the preceding section, expectations with respect to the price level and those regarding exchange rates are not mutually independent. Moreover, expectations with respect to changes in particular prices or costs, such as labor costs, materials costs, or competitors’ prices, are closely related to expectations with respect to the general price level and the exchange rate. It is not intended in this section to elaborate on the factors which influence businessmen’s expectations of changes in the general price level or of the future strength or weakness of the investee country’s balance of payments and its exchange rate policy. But something must be said about the interdependence of expectations.

In countries where inflation has persisted for many years, foreign investors would, in the absence of good reasons to the contrary, expect a continuation of inflationary trends. Balance of payments deficits caused by moderate domestic price increases can be financed, for a more or less limited period of time, through a reduction of external reserves. In some cases, the effect of inflation on the balance of payments can be counteracted by exchange control measures. But it is neither desirable nor possible to maintain the exchange rate at a constant level in the face of rapid and prolonged inflation. Over the long run, potential investors must therefore expect that the exchange rate will be depreciated more or less in correspondence with the rise in domestic prices.

For short-run periods, however, the investor cannot safely base his forecast of exchange rate changes on this necessary longer-run correspondence between exchange rates and price levels. For one thing, the balance of payments is affected by factors other than relative prices. Grants of aid from abroad, private capital inflows, favorable crops of export products, or changes in the exchange control system may make it possible for the authorities of a country—even if external reserves are not large—to maintain the exchange rate constant for some time while domestic prices increase. In countries experiencing rapid inflation, exchange rate adjustments are generally made at considerable time intervals even though domestic prices increase continuously. As a result, the currencies of these countries may at times be overvalued for several years; and after adjustment of the exchange rate, they may for some time be undervalued. The extent of variation in the relation of domestic prices to the foreign exchange rate in the four countries selected for this study is shown in Table 5.

Table 5.

Selected Developing Countries: Ratios of Cost of Living Index to Exchange Rate Index,1 Lowest and Highest Annual Values in Indicated Period


article image
Source: Based on data from International Monetary Fund, International Financial Statistics.

Exchange rate applicable to most capital movements.

While rapid and persistent domestic price increases make it likely that the exchange rate will sooner or later be depreciated, the reverse line of causation must also be considered. Exchange depreciation results in a rise of prices, in terms of local currency, of imported commodities and services. An exchange depreciation by x per cent will therefore have the direct result of increasing domestic prices by a proportion of x per cent which corresponds to the weight of imported goods and services in the total national expenditure on all goods and services. Experience has shown that, beyond this direct effect, there will be further induced price and cost increases resulting from the depreciation of the exchange rate, as well as wage adjustments made necessary by the induced rise in the cost of living index. For instance, if imports of goods and services amount to 20 per cent of the gross national product, one should ordinarily expect domestic prices to rise by more than one fifth, and perhaps by as much as two fifths, of the increase in the price of foreign exchange.

Beyond these considerations concerning the relation between the exchange rate and the general price level, expectations with respect to changes of particular prices or costs, such as competitors’ prices and labor or materials costs, play an important role in the decision of the potential foreign investor. For instance, he must consider the relation between changes in the general price level, on the one hand, and changes in wage and other costs and the prices at which the company’s products can be sold, on the other. During an inflationary period, wage costs often lag behind changes in prices, though in rapid and persistent inflation this lag is likely to be reduced, and it may disappear almost completely. To the extent that such a lag appears or is lengthened, real wages, and the producers’ wage costs in terms of dollars, would be reduced. The real purchasing power of tax payments may also decline in the course of accelerated inflation, unless the applicable tax structure is, on balance, progressive. Reductions in real costs resulting from such lags are likely to be more important when inflation begins or when it is accelerated; where prices rise at a nearly constant and predictable rate, these lags will tend to diminish and real costs will catch up with their long-run normal level.

The price of the product or service sold by the foreign investment enterprise in the local market can often, though not always, be raised in proportion to the company’s costs. It is particularly difficult for companies in regulated industries, such as public utilities, to maintain satisfactory real earnings during inflationary periods.

Under multiple exchange rate systems, it is necessary for the investor to form a view not only of the future strength or weakness of the balance of payments but also of the probable relative changes of exchange rates applicable to various transactions. For instance, in some circumstances experience may lead investors to expect that, whereas the basic export rate would remain unchanged for some time, the rate applicable to capital transactions and the repatriation of earnings would tend to depreciate. This is likely to occur particularly when, during an inflationary period, foreign exchange for capital and earnings transactions is bought and sold at a free or relatively uncontrolled exchange rate while the export rate is maintained constant or is depreciated by less than the free rate.

Expectations are rarely held with complete certainty. An increase in the degree of uncertainty with which a particular outcome is expected ordinarily reduces the response that it evokes. Price and exchange rate instability as such may, therefore, discourage the inflow of foreign capital. This deterrent effect will be most pronounced if the relevant prices are expected to fluctuate widely and in an unpredictable manner. The instability of certain commodity prices in world markets may give rise to greater uncertainty on the part of potential investors about future yields than does the more or less steady increase in domestic prices and costs in some countries with a long history of inflation. However, as has been outlined in the preceding section, the foreign investor is concerned primarily with the ratio of the investee country’s prices and costs to its exchange rate. Even in countries where domestic prices have risen steadily, this ratio has often shown wide fluctuations in the short run. The data may not permit conclusive demonstration of the deterrent effect of uncertainty on the total amount of private capital inflow into these countries over a longer time period, but fluctuations of the magnitude indicated in Table 5 in this relative-price factor, which is basic to the profit calculations of foreign investors, must be presumed to constitute an important obstacle to foreign investment, particularly where investors rely on a relatively short pay-out period.

Summary and Conclusions

The expectation of changes in exchange rates and in domestic prices and costs is only one of a number of elements affecting the decision to invest in a foreign country. Some of the other factors that investors must consider when deciding upon the amount and timing of foreign investments are probably given a greater weight in the decision than the price elements to which attention has been drawn in this paper.

It is, however, sometimes maintained that, other things being equal, inflation and exchange depreciation tend to discourage capital inflows and encourage capital outflows. If this argument were generally valid, countries that have experienced very rapid rates of inflation and exchange depreciation should be found to have had a net capital outflow or, at any rate, only a small capital inflow compared with more stable countries, unless they were unusually attractive to private investors on other grounds. The data for the four countries studied here—i.e., countries where the rate of inflation exceeded that in other countries for which the necessary data could be compiled—raise some doubts about this proposition.

It is argued in this paper that, with respect to the quantitatively most important component of the capital flow from industrial to developing countries, namely, direct and other equity investment, expected domestic price changes and expected changes in the exchange rate do not enter separately into the decision to invest. It is rather the ratio of domestic prices or costs to the exchange rate which determines whether a contemplated investment will be more or less profitable. In the longer run, investors have reason to expect movements in the exchange rate which follow more or less closely the changes in domestic prices in the investee country relative to prices in the industrial countries. Where the investor can take a long view of the profit potential of an investment, changes in the ratio of prices to exchange rates which are expected to be transitory will not affect the investment decision to any appreciable extent, though they may induce the investor to make particular outlays earlier or later than he would have done if stable prices and exchange rates had prevailed.

Investment projects which are undertaken with the expectation of a short pay-out period will tend to be affected to a much greater extent by the relation between expected changes in domestic prices and exchange rates. During a relatively short period (e.g., a few years), investors will not find it possible to rely on the expectation of parallel movements of prices and exchange rates. Their decisions to invest will, therefore, have to be made under some uncertainty, particularly in countries which have a history of rapid inflation and periodic exchange depreciation. This uncertainty may tend to reduce the total volume of investments of this relatively short-term kind. In addition, periods in which the ratio of prices to the exchange rate in the investee country is expected to rise will favor investment in industries supplying mainly the local market, whereas in periods when this ratio is expected to fall investment in export-oriented industries will be stimulated. If opportunities for foreign investment of both types are present in a particular country, the expectation of changes in this ratio may alter the composition of the capital inflow but, apart from the deterrent effect of the uncertainty with which the expectations are held, it will not necessarily reduce the total amount invested in the country.

These considerations may help to explain the continued large inflow of long-term capital into some countries with rapid inflation and exchange depreciation. Although the exchange rates in those countries were not always kept in line with domestic prices, investors appear to have expected that a rough correspondence between relevant domestic prices and costs and the price of foreign exchange would obtain in the long run. Moreover, in countries—for example, Argentina and Brazil—where exchange rates did not depreciate in proportion to domestic price increases in recent years, a change in the industrial composition of new foreign equity investment toward industries producing chiefly for the local market may have occurred.

Residents of the investee country, too, are likely to consider the relation between domestic prices and the exchange rate an important element in their investment decisions. Especially under conditions of rapid inflation, the expected ratio of prices of domestic assets to the price of foreign exchange will indicate whether the purchasing power of their investment will be best maintained by investing at home or abroad. When domestic prices have risen for some time relative to the exchange rate, residents may expect devaluation in the not too distant future and will, therefore, tend to shift funds abroad. Shortly after a devaluation, residents may expect an increase in domestic prices and, for a considerable period, a constant exchange rate and will tend to repatriate their funds in order to invest them in domestic real assets. Similar considerations apply to foreigners with respect to some of the short-term funds which they employ in the country in connection with their principal business enterprises. Here again, however, numerous other factors determine the movements of short-term funds, and the mechanism just described merely causes a general tendency for short-term capital movements which may be offset by other influences.

The considerations advanced in this paper give rise to some conclusions with respect to exchange rate policy, though these are by no means novel. In comparison with the effect of exchange rate policy on the balance of payments as a whole, its impact on private capital movements is, in the long run, likely to be small. Exchange rate policy should, therefore, be determined primarily with a view toward its effect on the current account balance. Countries that, on the whole, maintain domestic price stability but have to depreciate their exchange rate at a particular juncture to correct a fundamental disequilibrium which has arisen over the years need not fear that the exchange rate alteration will have an adverse effect on investment from abroad. Provided that investors expect the new exchange rate and existing price relationships to be maintained for an indefinite period, such investment is unlikely to decline permanently, though it may shift to some extent from industries supplying the local market to export-oriented industries.

Where the exchange rate is allowed to fluctuate in response to market forces in a general environment of domestic price stability, as in Canada between 1950 and 1962, the rate is likely to remain within fairly narrow limits. Such fluctuations as do occur in these circumstances do not give rise to a large revision of the expectation of exchange rates ruling in the more distant future. Nor do they necessarily make these expectations more uncertain than they would have been under a fixed exchange rate (which is, of course, subject to possible alteration). Private long-term capital movements are in this case not likely to differ very much in amount or composition from the pattern which would have been observed under a fixed exchange rate. Short-term capital will move in and out of the economy in response to short-term expectations with respect to exchange rates, but these movements will tend to be offsetting over a longer period.

Where domestic prices are not stable, the considerations worked out earlier in this paper come into play. A policy of maintaining an overvalued exchange rate through trade and exchange restrictions for as long as it is feasible, and of changing the rate periodically whenever the pressure on the control system and the distortions in the allocation of resources become unbearable, will not only be detrimental to the efficiency of the economy but may also, by increasing the degree of uncertainty as to future exchange rates, tend to reduce the net inflow of foreign capital. While the best solution would be to eliminate the causes of domestic instability, the persistence of inflation in many developing countries makes it necessary to consider also the “second best” solution of minimizing the harm done by domestic instability. The exchange rate policy which is indicated from the viewpoint of maintaining reasonably orderly payments conditions on current account, and of permitting the inflationary economy to partake as much as possible of the gains from international trade and specialization, will also assure the least damage to the flow of private foreign capital into the economy. This policy is one of adjusting the exchange rate with as much continuity as possible to its long-run equilibrium level, which will, under conditions of rapid inflation, move in close correspondence with domestic prices and costs. As far as the effect on capital movements is concerned, this policy will leave the relation between various domestic prices and costs, on the one hand, and the exchange rate, on the other, as free of disturbance as is feasible under the circumstances. As a result, incentives to foreign investors will not be impaired as much as they would be under an exchange rate policy which permits substantial, and widely fluctuating, discrepancies between the internal and external purchasing power of the currency.

It is true that maintenance of an overvalued exchange rate under inflationary conditions may for some time lessen pressures for wage adjustments, since the (official) local prices of imported commodities will fail to rise in step with prices of domestic goods and services. However, since imports cannot be allowed to enter freely under these circumstances, shortages of imported goods tend to add to inflationary pressures on the demand side and to undo the mitigating effect of the overvalued exchange rate—via the official consumer price index—on wage demands. Moreover, regarded in this way an overvalued exchange rate is not an anti-inflationary device but merely an instrument for the repression of certain symptoms of inflation. Its use for this purpose adds to the inflationary tendencies by distorting the allocation of resources and diminishing the efficiency of the economy. Therefore, not much weight should be given to this particular argument against adjusting the exchange rate in accordance with changes in domestic prices.

It could be argued that, in principle, persistent overvaluation of the exchange rate would encourage foreign investment in industries oriented toward the local market—a tendency which, as such, may be considered desirable. In practice, however, it is not possible to maintain a significantly overvalued unitary exchange rate for a sufficiently long period to induce the expectation that it will continue to the time horizon of investors, because of the adverse effect of an overvalued rate on other balance of payments components. In order to attract in each period enough foreign investment to offset a given deficit in other balance of payments transactions, the degree of overvaluation would have to increase in successive periods so as to provide an additional incentive to foreign investors, and progressive overvaluation would result in a rising deficit on other balance of payments accounts.

Foreign investment could also be attracted by maintaining an over-valued exchange rate applicable to the repatriation of earnings in a multiple exchange rate system, while other rates were set so as to produce balance in the country’s external transactions. Such a system of multiple exchange rates is a form of differential taxation (or subsidization) of various industries or activities. Multiple exchange rates are, however, easily altered by administrative decision. This adds to the uncertainty with which expectations regarding the relationship between the relevant domestic prices and various exchange rates are held. For this reason, it would be preferable—provided that it is feasible—to introduce any differential taxation which is thought to be desirable through tax legislation and to eschew the administrative flexibility which a multiple exchange rate system may seem to promise. Shifts in resource allocation are time consuming and costly; they must be permitted to take place, and even encouraged, when necessary. But it should not be supposed that a multiple exchange rate system furnishes the authorities with an instrument for rapid and frequent alterations of income distribution and resource allocation, whose flexibility is not purchased at a price.

An exchange rate policy which, under inflationary conditions, ensures a reasonably close correspondence between the internal and external purchasing power of the currency will not, by itself, be able to provide for an economy all the advantages of growth in a stable environment; but it will remove at least one source of maladjustments involving the relations between export industries and industries producing for the local market (including import-competing industries)—maladjustments which are particularly difficult to rectify once they have been allowed to persist for some time. Furthermore, such a foreign exchange policy will provide the basis for—though, of course, no assurance of—an uninterrupted flow of imported capital goods and materials necessary for continued development. Finally, it will ensure that access of the economy to private foreign resources obtained through foreign investment will not be needlessly abridged.

Mouvements de capitaux privés et taux de change dans les pays en voie de développement


Cet article passe en revue les effets de l’inflation et de la dévalorisation du taux de change sur les rendements de divers types d’avoirs acquis par les investisseurs privés étrangers. En ce qui concerne le type d’investissements extérieurs le plus commun dans les pays moins développés, à savoir les investissements directs, les rendements augmentent sous l’effet de la dépréciation du taux de change et diminuent sous l’effet de l’inflation intérieure si la production de l’entreprise où les capitaux sont investis est en grande partie exportée; l’inverse se produit si l’entreprise écoule sa production ou vend ses services surtout sur le marché local. Si, toutes choses étant égales, on s’attend à ce que les prix et coûts intérieurs augmentent par rapport au prix des devises, les investissements étrangers directs dans le secteur manufacturier en seront peut-être stimulés, mais les investissements dans les industries extractives, dont la production est souvent destinée aux exportations, pourront se ralentir. Toute augmentation prévue du prix des devises en fonction des prix et des coûts intérieurs tendrait à avoir l’effet inverse sur les mesures destinées à stimuler les investissements. Ces incidences dépendent naturellement de la mesure dans laquelle les prix de vente et les coûts de l’industrie en cause subissent l’influence de l’inflation générale des prix, et aussi, dans le cas des régimes de taux de change multiples, des variations relatives des taux de change applicables aux divers types de transactions.

Dans une situation d’inflation, et lorsque le taux de change est maintenu fixe, les investisseurs sont souvent extrêmement incertains du rapport qui pourra s’établir entre les prix et les coûts intérieurs d’une part, et le taux de change de l’autre. Cette seule incertitude peut réduire le volume d’investissements étrangers que pourra recevoir un pays en voie de développement. On peut réduire la marge d’incertitude de ces prévisions en adoptant sans trop de retard une politique tendant à ajuster le taux de change aux variations des prix et des coûts intérieurs.

Los tipos de cambio y los movimientos de capital privado en los países en desarrollo


Este artículo analiza los efectos que la inflacion y la depreciación cambiaria ejercen sobre el rendimiento de varias clases de activos adquiridos por inversionistas privados extranjeros. En tratándose de la clase de inversión extranjera que es más frecuente en los países menos desarrollados, es decir, la inversión directa, si se exporta la producción de la empresa en que la inversión ha sido hecha, la depreciación cambiaria aumenta el rendimiento y la inflación lo disminuye; lo contrario ocurre si la empresa vende la mayor parte de su producción o de sus servicios en el mercado interno. Si, dadas, por lo demás, las mismas circunstancias, se espera que los precios y costos internos aumentarán relativamente al precio de divisas extranjeras, ello pudiera servir de estímulo a la inversión directa extranjera en el sector manufacturero, pero es posible que la inversión en las industrias extractivas, las cuales a menudo están orientadas hacia la exportación, sufra retardo. La expectativa de un aumento en el precio de las divisas extranjeras en comparación con los precios y costos internos, pudiera afectar en un sentido contrario a los incentivos para las inversiones. Todos estos efectos dependen, por supuesto, de la medida en que una inflación general de los precios influya sobre los de venta y sobre los costos de la industria de que se trate, y donde existan sistemas de tipos de cambio múltiples, dependen, además, de las alteraciones relativas de los tipos de cambio aplicables a varias clases de transacciones.

En una situación de inflación y de tipos de cambio mantenidos constantes, los inversionistas frecuentemente hállanse ante una gran incertidumbre respecto a la futura relación entre los precios y costos internos, por una parte, y el tipo de cambio, por otra. De por sí dicha incertidumbre puede disminuir el volumen de las inversiones extranjeras que un país en desarrollo habría de recibir. El grado de incertidumbre en cuanto a esas expectativas puede reducirse mediante una política que, sin demora indebida, ajuste el tipo de cambio a las variaciones en los precios y costos internos.


Mr. Rhomberg, Chief of the Special Studies Division of the International Monetary Fund, is a graduate of the University of Vienna and of Yale University and has been a member of the faculty of the University of Connecticut. He has contributed chapters to several books on economic subjects and articles to economic journals.


The author has benefited greatly from comments made at the Round Table Conference by Professor Wolfgang F. Stolper and other participants.


This view finds expression, e.g., in the following statement: “Balance-of-payments problems, inflation, and consequent currency depreciation have for many years been a way of life in most Latin American nations, with a resulting adverse effect on earnings in terms of U.S. dollars, as well as an erosion of U.S. dollar investments in Latin American enterprises, particularly manufacturing” (“Proposals to Improve the Flow of U.S. Private Investment to Latin America: Report of the Commerce Committee for the Alliance for Progress,” Private Investment in Latin America, Hearings Before the Subcommittee on Inter-American Economic Relationships of the Joint Economic Committee, 88th Cong., 2nd Sess., p. 79).


Net short-term capital flows can be obtained in Table 1 by subtracting long-term from total capital inflows. It is here assumed that errors and omissions reflect mainly unrecorded capital flows of a short-term character. This item does, however, also contain the net amount of any unrecorded current account transactions. Furthermore, some clandestine capital transactions may fail to be reflected in the entry for errors and omissions.


See Organization for Economic Cooperation and Development, The Flow of Financial Resources to Less-Developed Countries, 1956-1963 (Paris, 1964), Table IV-4, p. 134. Lack of data for the years before 1956 prevents a comparison of the ratios for the two subperiods of private investment in the four selected countries to total OECD private investment in all less developed countries. However, annual U.S. private long-term investment in Latin America was less in 1958-63 than in 1951-57. Even if 1957, a year of extremely high U.S. foreign investment in Latin America (mainly in the petroleum industry) is omitted, it turns out that U.S. investment was slightly more in 1951-56 than in 1958-63. The increase in private long-term investment in the four countries, from $1.6 billion in 1951-57 to $3 billion in 1958-63, is thus very likely to represent a rise relative to the supply of private long-term capital available to Latin American countries in general.


See Graeme S. Dorrance, “Inflation and Growth: The Statistical Evidence,” pp. 82-102, below.


By “net revenue” is meant the sum of gross receipts from the asset minus the costs incurred in realizing these receipts, such as labor and materials costs and taxes; depreciation allowances are not deducted, and the initial outlay on the asset is thus assumed to be realized over the life of the asset.


V, equals the exchange loss, resulting from a depreciation of the peso, on net assets denominated in pesos, such as bank deposits in the investee country, or accounts receivable minus accounts payable; where conditions permit it, such exchange losses can be avoided by raising working capital, or that part of it which is tied up in assets denominated in pesos, in the local market.


In addition, there may be an exchange valuation loss, i.e., V, may be positive.


The term “price” is here used in the general sense, including prices of cost items and of foreign exchange.


With an unchanged exchange rate, production costs would remain constant only if (1) the supply of foreign materials were perfectly elastic with respect to price and (2) the elasticity of substitution between domestic and foreign materials were perfect. But if these conditions were met, the price of domestic materials would presumably not have risen in the first instance.


To be sure, in the long run, adaptation to altered price and cost relationships may be possible through changes in the capital stock. The extent to which such long-run adaptation can occur will vary from industry to industry in accordance with technological characteristics.


U.S. investment in Latin American agriculture is in heavily export-oriented enterprises.


This statement must be qualified with respect to the working capital valuation adjustment: there may be an exchange loss on net financial assets denominated in pesos (though not on inventories, since the inventory valuation gain resulting from inflation offsets the exchange loss on the capital tied up in inventories).


Postponement of an investment project will be advantageous if the reduction of the cost of the investment achieved thereby exceeds the difference between the net revenue which the asset would have earned and the return earned on an alternative short-term investment of the funds set aside for the project. Let the peso cost of an investment be a fraction q of the total cost, A, of the project; let its expected yield (after proper allowance for expected changes in prices and exchange rates) be y and the yield on an alternative short-term investment, e.g., Treasury bills in the investor’s country, i. Other things being equal, it will be worthwhile to postpone the investment for one year if pc/rc, < 1—(yi)/q. For example, if the project yields 12 per cent per annum, the Treasury bill rate is 4 per cent, and 25 per cent of the outlay for the project is in pesos, a 50 per cent expected increase (i.e., depreciation) of the exchange rate in the course of the next year will, assuming unchanged peso costs, just warrant postponement of the investment (pc/rc = 0.67; 1—(yi)/q = 0.68).