V Summary

Kyong Huh, Benedicte Christensen, Peter Quirk, and Toshihiko Sasaki
Published Date:
May 1987
  • ShareShare
Show Summary Details

The experience with flexible exchange arrangements in the specific form of floating exchange rates in developing countries since the advent of generalized floating by industrial countries in 1973 is relatively limited, although in recent years an increasing number of developing country Fund members have adopted such systems. The early experience indicates that floating exchange rate systems can function satisfactorily in developing countries with relatively diverse economic structures, despite the limited depth of their financial systems. However, the exchange arrangements have to be adapted to the institutional strengths and weaknesses of individual countries. Most particularly, these freely floating exchange market arrangements have to be supported by the sustained pursuit of appropriate domestic economic policies to ensure their efficient operation over time. Such arrangements are often also the only alternative to restrictions, arrears, and controls to insulate the balance of payments from domestic economic mismanagement. However, from an efficiency viewpoint, it must be stressed that a floating exchange rate cannot substitute for appropriate economic policies.

Developing countries have adopted floating rates for a variety of reasons, but most did so because of severe balance of payments difficulties that had resulted in external payments arrears. In fact, flexible exchange rate arrangements, including the floating exchange rate systems on which this report has focused, may be the realistic option for members with severe balance of payments difficulties reflected in low official reserves and persistent arrears and insufficient room for maneuver in domestic policies. In a number of the countries, protracted balance of payments problems had previously been addressed by extensive controls on foreign exchange transactions that had led to disintermediation. A major aim of the market-determined floating arrangements has therefore been to bring back into the official sector the extensive illegal or unofficial parallel markets in which exchange rates were substantially depreciated, as well as to encourage the repatriation of capital flight which, in a number of cases, had become a major problem. Another reason for the introduction of market-related floating in several instances was the desire on the part of the authorities to shift the determination of the exchange rate from the authorities to market forces. With the rate determined in an open market, the authorities were better able to focus decision making on other areas of economic management.

Arrangements that were in effect before the adoption of market floating ranged from a managed float, in which the exchange rate in relation to the intervention currency was changed frequently in an attempt to counteract the effects on competitiveness of rapid domestic inflation, to relatively fixed exchange rate arrangements against a major currency. Many arrangements in effect before free market floating also have included multiple exchange rates. In some, the introduction of the freely floating market for exchange rate determination involved transitionally the existence of a secondary market for certain transactions in which the rate was freely determined, followed by a progressive transfer of transactions to that market until complete unification of the markets was effected. Typically, where this approach was taken in the context of a Fund-supported program, it was understood that the rates would be unified within a relatively short time.

An important choice facing a developing country in instituting a floating exchange market is whether it should take the form of an auction or an interbank market. The experience to date has been that markets operated by the commercial banking sector have been less subject than officially operated auction markets to destabilizing intervention in the form of inappropriate official purchases and sales or ad hoc controls on access to the market. In addition, it has been possible in using interbank markets to build readily on existing expertise of banks and foreign exchange dealers operating in formal or informal parallel markets, even when the formal markets are very limited. Interbank markets also function on a continuous basis, while auctions are periodic by their nature and therefore less efficient as clearing mechanisms. The less frequent are the auctions (they are conducted daily in only one country), the less efficient and smooth will be the clearing process, as delays in obtaining foreign exchange will be longer, and uncertainty and risk involved in the exchange transaction will be greater, the longer the interval between supply to the central bank and the actual auction itself. Interbank market arrangements have been the more common setting for freely floating exchange rates in developing countries to date; nine developing countries have adopted interbank arrangements and six have instituted auction arrangements. In these and other instances, the Fund has assisted in formulating the systems by transferring and adapting experience among members concerning the design and implementation of specific market modalities.

A major consideration in setting up the floating markets has been the need to incorporate safeguards against destabilizing speculation and the establishment of monopoly positions. In several instances, limits have been put on the foreign exchange positions of commercial banks and other dealers, both to ensure that the market is not cornered and to limit excessive exposure to exchange risk. The limits have generally been set with reference to norms derived from experience in the management of foreign exchange working balances. Freedom of entry into the exchange market has been an important factor in ensuring its competitive operation, particularly where the number of commercial banks is small.

Under auction arrangements where foreign exchange must be surrendered to the central bank, often the volume of foreign exchange transactions in the market has been sharply diminished as a result of the accumulation of reserves at the market-determined rate or extra-market allocations of foreign exchange by the central bank. These latter are largely for official uses or for the reduction of external arrears (including payments for external debt obligations). In some countries, the exchange requirements of public enterprises have been met outside the auction, and retention privileges have been granted to private sector concerns, also with the aim of guaranteeing the availability of foreign exchange. As a result of the setting aside of foreign exchange for these purposes, the proportion of foreign exchange receipts accruing to the economy as a whole that has actually been auctioned by the central bank has been as low as one fourth. This has created problems for the effective functioning of the market and for the provision of sufficient exchange for orderly discharge of current import demands.

Steps to introduce forward exchange markets have also been taken by a few countries with floating spot markets. Forward market cover is important in developing countries to facilitate external financing, and to insulate the productive sector from some of the effects of market volatility. However, the development of forward exchange market facilities in these developing countries is at a relatively early stage, with only a very limited volume of forward transactions. Further work is needed to determine institutional arrangements suited to conditions in developing countries—including possible brokering arrangements by the central bank or more limited forward market arrangements for short-term facilities, primarily for trade cover.

In most cases, floating exchange rate systems have been introduced with Fund assistance in formulating technical aspects, as part of Fund-supported adjustment programs. In fact, in several instances, the implementation of the floating system was one of the first steps taken in such programs and set the stage for the formulation of the underlying macroeconomic framework and other supporting policies. Although it can be argued that such a sequence could in principle lead to an early overshooting of the rate and subsequent reversal of the initial depreciation, in practice this has not happened. On the other hand, the new realism in the exchange rate policies in the countries adopting floating rates has helped to create a more favorable climate for comprehensive adjustments to domestic policies as well as for discussions with foreign donors and creditors, particularly in the period leading up to discussions on debt rescheduling. It is, of course, critically important for the effective functioning of a floating market exchange rate that it be supported by effective implementation of appropriate policies of demand restraint and structural change and by increased external resource inflows, especially of a kind that can be readily fed into the exchange market. To assist in providing as stable an environment as possible for the introduction of floating, and with the aim of minimizing the initial depreciation of the floating rate, sometimes arrangements have been made for bridging finance from donor countries or commercial banks ahead of drawings under a stand-by arrangement with the Fund and rescheduling of existing external obligations.

The successful operation of an exchange rate float will be helped by strengthened confidence in the economy and more particularly in its external payments outlook. In this context, the outlook for increased foreign exchange receipts on current and capital accounts will be critical. Usually the measured effects of the initial depreciation associated with floating on the output of tradable goods (exports and import substitutes) have been subject to lags, but the absorption of black market inflows and the incentives created for reducing capital flight have led to some relatively early significant beneficial effects on the capital account. The curbing of capital flight, and the encouragement of greater repatriation of foreign exchange earnings and capital reflow have, in fact, been a major aim of some authorities in adopting floating systems. This implies, for countries where the onset of balance of payments difficulties seems imminent—as often indicated by a rising premium on foreign exchange in the parallel market and foreign exchange cash flow problems—that the adoption of a floating rate may serve to prevent the difficulties from intensifying into an actual accumulation of arrears.

To the extent that exchange controls are retained following the exchange market reform, incentives will remain for evasion through, for example, underinvoicing of export receipts and overinvoicing of goods and services imports. In fact, most members adopting floating rates have substantially liberalized their exchange and trade systems, and from available data on capital flows it seems that liberalization and floating have led to decreased capital flight in all surveyed countries, and to some short-term reflows. However, the data on which these observations are based are yet weak. In addition, given the prolonged balance of payments problems experienced by the countries adopting floating systems, confidence is understandably slow to rebuild. It requires a perception that the adjustment policies are likely to be lasting and that there are increased resource availability and improved external balance. The lack of such confidence has probably limited the extent of the capital reflows observable to date. In any event, the adoption of floating exchange rate systems should normally take place in the context of a broad program of exchange and trade liberalization, which has been the case with those members that have introduced such systems to date.

A concern that has been voiced about the adoption of floating exchange rate systems by developing countries—at least before the more extensive use of such regimes in recent years—is that they could lead to a free fall of the exchange rate, thereby contributing to a cumulation of inflationary forces. Another concern has been that floating rates would be relatively unstable because of the limited depth of financial markets. Such concerns, however, are misplaced in that they attribute the free fall or the instability of the exchange rate to the particular modality of exchange rate determination and overlook the critical relationship between developments in market-determined exchange rates and the quality of domestic economic policies. Indeed, the evidence surveyed in this paper, covering between one and three years’ experience in the individual countries, does not appear to support these concerns. In one or two countries, the domestic currency has actually appreciated after floating, partly in response to the effect of higher domestic interest rates in generating net capital inflows. In other countries, the official exchange rate has depreciated sharply with the introduction of floating, but in these cases the black or unofficial market exchange rate prior to the action appears to have provided a floor for the initial movement. Subsequently, the rate has tended to move over time, generally in accordance with relative inflation rates. Further, the volatility of exchange rates to date in countries adopting floating rate systems has in most instances been less than that prior to floating, especially in terms of the real effective exchange rate.

As might have been expected, the correction of the exchange rate following floating has led to price level increases through cost-push effects. The rise in the inflation rate, however, has been considerably less than implied directly by the share of imports and the magnitude of the depreciation, for two reasons. First, the cost-push effect has not, to some degree, been accommodated by monetary policies, and second, many prices at the consumer level had already adjusted to reflect the shortage of exchange for imported goods reflected in the black or unofficial market exchange rate. To a large degree then, the price corrections were confined to the official sector and to goods subject to official price controls—notably, imported foodstuffs and oil. Following the initial surge corresponding to the exchange rate correction and the freeing of controlled domestic prices, in many of the countries both exchange rate movements and inflation have tended to recede rather than gain momentum.

Table 14.Comparison of Average Retail Prices for Selected Goods in Uganda, Zaïre, and Sierra Leone, First Quarter 1984In U.S. dollars per unit
UnitsUgandaZaïreLeone Sierra
Kerosene1 1.0.510.191
Cloth1 m.1.203.10
Memorandum: Exchange Rate (per US$1)292.0635.2882.50
Source: Data provided by national authorities.

    Other Resources Citing This Publication