1. Main types of regimes

Although virtually all industrial countries now maintain floating exchange rate systems, ^3/ most developing countries continue to have pegged or managed exchange rates and relatively few have floated their currencies freely (Table 1). The major forms of regime are distinguished by their flexibility--that is, the frequency with which the rate is permitted to adjust.

Single currency pegs: The country pegs to a major currency, with infrequent adjustment of the parity on the basis of discrete decisions by the authorities. However, although pegged, the currency fluctuates against most currencies because of the floating relationship between the currency to which it is pegged (mainly the U.S. dollar and French franc) and other currencies. About one half of all developing countries maintain this form of arrangement.

The most extreme form of single currency peg is that by which the currency of another country is used in circulation. Presently, Panama and Liberia are the only Fund members that use other currencies in this way--the U.S. dollar in both cases, (Liberia also circulates its own coins, but not banknotes). Several other smaller countries associated with Fund members also use an industrial country’s currency as their currency of circulation (Table 2). The drawbacks are a loss of seignorage and independence of monetary policies, but for a very small country this may be compensated for by reduced administrative costs and perhaps greater financial stability. One-to-one parities with the peg currency have also become relatively rare, because different inflation rates have necessitated adjustments over time. No Fund members have such pegs at present, although until the mid-1980s the currencies of the Bahamas, Dominican Republic, and Guatemala were at parity with the U.S. dollar.

In several member countries, owing in part to political sensitivities, the authorities have declared a peg to a currency unit which in practice they do not adhere to closely. Several oil-exporting countries are pegged to the SDR, but in practice limit the flexibility of their exchange rate on a short-run basis against the U.S. dollar (the “quasi-peg”).

Currency composite pegs: Such pegs represent an attempt to stablize the value of the currency against some average of major trading partner currencies. Most often, the weights applied to average the basket of currencies are trade (export and import) weights, although in some instances services trade or major capital flows may also influence the weights. The weights may also be constructed in such a way as to reflect the import and export elasticities through which the foreign currency fluctuations affect the external sector. Other main parameters in the operation of a composite peg are the base period with reference to which the peg is stabilized, and the period over which the weights are averaged. About a quarter of all developing countries have currency composite pegs.

In a small number of developing countries, rather than adopting a currency composite tailored to the trade composition of the individual country, the SDR has been adopted as the composite peg. The particular advantage of the SDR relative to tailored composite pegs is the ease of computation because it is available on a daily basis from the Fund. Seven developing countries presently use the SDR as a peg.

Cooperative arrangements: The cooperative arrangements maintained by certain European Community countries represent a cross between pegged and floating arrangements. The currencies of the European Monetary System (EMS) members are pegged to each other, but float against other currencies. The ability to so peg cross exchange rates is achieved by harmonization of monetary and fiscal policies within the region. The essential differences from a single currency peg are that policy harmonization arrangements exist within the region, and that there are active floating markets for determining the value of each EMS currency against those outside the EMS arrangements. With single currency pegging, the harmonization does not usually exist between the financial policies of the pegger and the country to whose currency it is pegged, and active markets do not exist for that currency vis-à-vis the other currencies against which it floats de facto. The cooperative arrangements have correspondingly less potential for overvaluation or undervaluation of exchange rates.

Indicator arrangements: In the case of a single currency peg, the exchange rate may be varied from time to time in response to various indicators, but in a managed indicator arrangement the basis on which the changes are made is formalised. The member identifies certain key indicators that it will use more or less automatically in making a decision to revalue the exchange rate. A common form of indicator arrangement is the inflation-adjusted (“real”) exchange rate peg, which has the aim of achieving continuous competitiveness against a composite of major partner countries. In setting up such an arrangement, the same consideration arises as for composite pegs--i.e., the weights applied to the partner-country currencies, and the base period and frequency of adjustment--and also the choice of cost or price indices. Another form of indicator arrangement is that of the preannounced exchange rate or “tablita”. Under such arrangements, the exchange rate crawls at a predetermined rate--for instance, at a given percentage per month.

A problem with both forms of indicator arrangement is the resulting predictability of exchange rate movement which may create obvious profit-making possibilities and adversely affect expectations of future price movements, even if it is used only for relatively short-run management of the exchange rate. In the specific case of real exchange rate rules, there may be lags in the availability of inflation data for readjustment of the exchange rate, and the market may in the meantime take a view of where the rate will be, leading to speculative capital flows in the meantime. With preannounced rates, this problem has been particularly serious, so that the arrangements are hardly used any more.

Managed floating: Under this form of arrangement, the central bank rather than the market sets the exchange rate, but varies it frequently. The difference between this and other forms of pegged arrangements or indicator arrangements is (a) the broad judgemental factors used in assessing indicators for adjusting the rate, and (b) the frequent but non-automatic nature of the adjustments. Because of the diverse nature of these arrangements, it is difficult to generalize about them. The rate may be set with regard to a judgemental mix of many factors--for example, the real effective exchange rate, developments in the balance of payments, in international reserves, and in the parallel or black markets for foreign exchange.

Independent floating: An increasing number of developing countries have decided in recent years to float their exchange rates using auction or interbank market arrangements (these arrangements are discussed in more detail in Chapter IV). The essential feature of the arrangements is that the exchange rate is market-determined. A major reason for adopting a float has been the developing country’s desire to shed political responsibility for devaluing the exchange rate, because discrete adjustments to managed or fixed exchange rates often have unpopular political results, particularly for groups of society favored by the previous exchange rate regime.

The form of intervention associated with these arrangements is purchases or sales of foreign exchange by the authorities. Generally speaking, the intervention is aimed either at stabilizing the market against periodic unsustainable movements in either direction, or at slowing down the rate of change by leaning against market pressures. Despite this intervention, the systems are differentiated from the fixed or managed arrangements above, in that the rate is not fixed over the short run but responds directly to exchange market pressures. Exchange controls have been increasingly unsuccessful in stemming hot money flows, and the size of these flows may be sufficiently large that official intervention through purchase and sales is often powerless to achieve an effect sustained beyond one or two days. Increasingly sophisticated and integrated international financial markets have tended to discount such official efforts. Exchange rate coordination between the developed countries has therefore focused increasingly on fundamental monetary, fiscal, incomes, and structural policies.

2. Considerations in the choice of regimes

In an unsticky and ideal world, the choice of regime would be unimportant because the authorities could choose both the magnitude and frequency of adjustment in such a way as to replicate the operation of any other regime, including free-floating. In practice, however, issues of political economy dominate the choice of fixed or semi-fixed exchange rates, and the exchange and trade controls that have supported the fixed rates have often been directed to favoring a particular political groups within the economy. Political considerations have therefore often delayed necessary adjustments of the exchange rate under the fixed or managed regimes.

An argument often cited for fixed or managed regimes is that of smoothing the path of the exchange rate in the face of exogenous shocks. It is argued that a floating exchange rate will reflect reversible events and therefore be more volatile. However, the practical difficulty for the authorities in managing the rates so as to smooth volatility is in deciding what shocks are wholly or partly reversible, or the degree to which they are reversible. For example, the present level of oil prices--is it permanent and therefore something that the oil exporter’s exchange rate should adjust to, or is it temporary? The mixed experience with buffer stock schemes for stabilizing international commodities prices suggests that predicition of such price movements is problematic.

Another argument that used to be given for the use of fixed or managed exchange rates was that the alternative of floating was not open to developing countries because of institutional weaknesses. However, the experience in this decade during which some 19 developing countries have adopted floating auction or interbank exchange markets, has shown clearly that this is not the case. Markets in these countries have proven themselves capable of smoothing out considerable seasonality in the balance of payments. They have also operated efficiently even in countries where there are only one or two commercial banks. The availability of trading among customers themselves often serves as a natural check against collusion or the exercise of monopoly power by the banking system.

At a more basic level, the case for or against the various exchange rate regimes can be argued only on the basis of performance. However, it is difficult to judge the performance of economic systems because their influence is diffused. Exchange rates are only one of the many policy variables that shape the course of economic developments. Since 1983, in particular, the group of non-oil developing countries has been subjected to severe and widespread balance of payments difficulties, with many unable to remain current on their external obligations. Developments in external payments arrears of the groups of countries maintaining different exchange rate regimes therefore provide an indication of the efficacy of the regimes. Have certain regimes been more conducive to avoiding or reducing arrears than others?

Chart 1 presents some data on the relationship between arrears and the main forms of exchange rate regime since the advent of systemic debt difficulties in the early 1980s. The correlation of the two must be viewed as no more than suggestive. However, on the basis of both the number of countries within each regime incurring arrears and those in that same regime that have not incurred arrears, the single currency peg arrangements appear to have fared worse. Those countries with single currency pegs have shown a much higher ratio of increasing arrears to decreasing arrears than both the composite peg and managed and independently floating groups. In addition, the proportion of countries incurring arrears has been higher in the single currency pegging group than in the latter two groups.

CHART 1

NON-OIL DEVELOPING COUNTRIES: EXCHANGE RATE REGIMES AND EXTERNAL PAYMENTS ARREARS

View Expanded

Source: IMF, Annual Report on Exchange Arrangements and Exchange Restrictions, 1984 through 1988 editions.

Another way of looking at the problem of performance is how the non-oil developing countries reacted to such severe balance of payments difficulties in their choice of regime. The two main developments of the period as shown in Table 3 have been the decline in the use of the SDR, with only as half as many countries maintaining this form of arrangement in 1987 as in 1983. The second trend has been in the use of independently floating arrangements, where the number has almost quadrupled from three countries as of the end of 1983 to eleven at the end of 1987. Another trend that emerged earlier in this period was the declining use of the U.S. dollar as a peg, from 37 countries in 1983 to 31 in 1985. However, with the depreciation of the dollar itself against other currencies, countries have returned increasingly to this form of single currency peg, in effect following the dollar value down. Single currency pegs were therefore not adhered to over the period, although on a point-to-point basis their use was unchanged.

1. Main forms of controls

Import Licencing and associated foreign exchange budgets are the major form of quantitative restriction, imposed extensively by developing countries for various reasons--including balance of payments support, industrial protection, health, security, and political and social reasons. Only a small minority of developing countries are free of some form of import licencing, and of those countries maintaining such systems, over one half require licences for all imports.

The most efficient way to operate a relatively free system is to permit all payments and transfers to proceed unless they are specifically prohibited or are subject to prior approval (the so-called “negative list” approach). The relative brevity of the list of restrictions makes it transparent for the authorities and users alike. Conversely, under the “positive List” approach, all imports are in principal prohibited and therefore require individual approval unless explicitly exempted from approval. Most developing countries have now moved to a negative list approach.

Although many countries have imposed licencing schemes in an attempt to conserve foreign exchange, only a relatively small number have systematically distinguished non-essential consumer goods and imports from other import categories in the regulations, although this has been pursued administratively in many countries. Import licences are frequently granted liberally if financed with “own” foreign exchange obtained by the importer outside of the official exchange market.

Systems of comprehensive licencing exist mainly in Africa and the Middle East, although there are some in other regions. A common distinguishing feature of the comprehensive systems is a set of complicated administrative procedures for authorising imports. Some countries may announce an annual import program or foreign exchange budget, or lists of prohibited or restrictive goods, together with other liberalized items under “open general licence” arrangements. (For further discussion of the latter, see Chapter IV.) However, despite the regulations, what is important is the way in which the system is actually administered. Some countries in the positive list group have in practice granted import licences rather liberally.

Although it is common to measure the degree of restrictiveness of an import licencing system by looking at the ratio of the value of imports not subject to licencing to total imports, this can be misleading. For example, the ratio in Morocco fell in 1981 to 30 percent from 34 percent the previous year despite substantial liberalization, owing to an increase in the prices of some imports subject to licencing. Another ratio which is often used to show the degree of restrictiveness or openness of an import regime is the number of items subject to licencing in relation to the total number of import items, but here again, a small ratio could be a result of reclassification and aggregation. The detailed statistical base to ensure consistent application of the international conventions (such as the Brussels nomenclature) is often lacking.

Control may be effected primarily through foreign exchange procedures, or customs documentation, or both. In some cases, import Licences are a prerequisite for the purchase of foreign exchange. However, in other countries, foreign exchange budgets are the operative constraints and import licences are granted only if there is foreign exchange available. The role of regional arrangements may be important for a number of countries in determining their import licencing, with exemptions for member countries of the customs union.

Taxes on imports are known by a number of names--for example, tariffs, import surcharges, and stamp duties--and are applied by virtually all countries. They achieve their restrictive effect through price-distorting incentives, in contrast to quantitative restrictions which act directly to limit freedom for imports. The systems of taxation may be complex, with different types of taxes for different purposes compounding one another, or they may be simple, transparent, and uniform.

In some cases, the taxes are introduced specifically for temporary purposes. A transitional “flexible” tariff is applied by some countries in the process of liberalization. Surcharges may also be levied for specific fiscal purposes, and emergency tariffs may be applied temporarily to discourage imports viewed as nonessential or to protect a domestic industry facing sudden injury due to imports. Tariffs may also be retaliatory against certain products deemed to be subsidised in the country of origin, and thus be aimed at eliminating the offending subsidy.

Tariffs are generally comprehensive and therefore are analogous to positive import licencing. The exceptions nevertheless abound. Duty drawback schemes for imported raw materials are common and may be based on judgment as to whether local materials are competitive in price and quality. If they are, the drawback may not be allowed. Other forms of industrial promotion using the import tax system include the earmarking of certain import duties for export promotion purposes, taxes on capital or consumer goods to finance an export subsidy fund, and tariffs geared to protecting domestic infant industry. The support to local industry may also be more qualified and indirect. Raw materials may be subject to exemption from customs duties and commercial taxes, providing that products in which they are used are exported within a specified period, or they may be used as an incentive to direct investment by exemptions for joint ventures and special economic zones. Machinery imports are a common exception, and duties may even be modified by exemptions for specific importing persons or firms.

Structures of effective taxation rates on imports may vary from uniform to highly complex. Some countries have subjected imports to a uniform tariff rate, but these are relatively rare. In others, the range of nominal duty rates may range very widely, although in practice the bulk of transactions may take place within a relatively narrow range.

Advance import deposits constitute another form of implicit taxation of import payments. Under such arrangements, the authorities require each importer to place a (usually non-interest bearing) deposit with the central bank or commercial bank prior to the opening of the letter of credit. The deposit may have the aim of ensuring that the importer has sufficient liquidity for the transaction, or it may simply be there to act as a disincentive because of the foregone interest on the deposit. The tax equivalence of the deposit can be calculated by assessing the foregone interest against the value of the import.

Multiple exchange rates may also be used as a device for taxing or subsidising imports. Such practices have grown less widespread in recent years as countries have unified their exchange rate systems. In some countries, imports of “essentials” may be imported at a relatively appreciated exchange rate. Imports viewed as essential normally include foodstuffs and medical items, and imported raw materials for domestic industries. The rate of effective subsidization is the difference between the (usually appreciated) official exchange rate and the exchange rate that would clear the exchange market in the absence of the restrictions.

Restrictions on international service transactions are diverse and include restrictions on foreign exchange for payments such as travel, medical expenses, subscriptions, advertising and study abroad, restrictions on transactions insurance, transport and freight, banking, and restrictions on payments for services rendered by nonresidents. The latter includes measures affecting remittances, investment income and workers’ wages. Investment income flows are affected by restrictions on current remittances (profits and dividends) as well as by restrictions on investments described under the heading of “capital” as these influence the decision to invest and hence future current outflows. In some service sectors, where trade can be a substitute for foreign direct investment (computer services, some forms of insurance, construction, communications, research, consulting, accounting, architectural, engineering, legal and financial services) restrictions on capital transactions may affect trade to the extent that they make foreign direct investment unattractive, or provide promotional incentives to encourage it.

Virtually all developing countries have restrictions on foreign exchange for travel by residents abroad, usually in the form of maximum amounts for specific purposes. However, in a growing group of countries, foreign exchange in excess of the limits may be granted on a case-by-case basis, provided that the request for the larger allowance can be shown not to involve a capital transfer. Business or official travel is typically entitled to higher limits on foreign exchange than tourist travel. Maximum limits also normally apply to foreign exchange made available for students studying abroad and for medical expenses that are generally administered on a case-by-case basis with a requirement that similar levels of education and medical treatment are not available domestically.

Imports of banking services require licences or may be permitted without restrictions of any kind. Often, firms and banking insurance need to be located at the site of the transaction to facilitate operations, which raises the issue of the foreign firms right of establishment, or right to locate in the host country to provide services to residents. Specific restrictions are common in the banking sector. Often, foreign banks will be required to carry higher reserve requirements than domestic banks. Dealings with domestic customers may be limited to exclude nationals.

Repatriation of profits and dividends is usually guaranteed for registered investments, but is otherwise widely restricted. Reinvestment in the local economy is limited in some cases in the same sector or may even be limited to the same firm. Permits are required to effect capital transfers in the majority of developing countries.

Restrictions on payments for services have hindered foreign direct investment in many developing countries. Special arrangements for relatively free treatment have been made to facilitate debt-equity swaps in a number of countries--as a means of reducing an unsustainable debt load.

The supply of foreign exchange to the domestic economy is influenced by so-called surrender requirements and retention allowances in many developing countries. Arrangements for surrender of export receipts reflect the general restrictiveness of the payments system. Some developing countries, mainly those with liberal capital systems, require neither repatriation to the home country nor surrender to the central bank or commercial banking system. In other countries, the foreign exchange earnings are required to be brought back to the country by residents, but after repatriation the foreign exchange may remain in the ownership of the exporter. In addition, there are arrangements in some countries for repatriation followed by “domiciliation” of the exchange. The latter involves processing of both the export financing and customs documents within a single bank which then acts as the agent for the government in recording and regulating the transaction.

In most developing countries, the requirements for surrender are partial and the exchange retained by the exporter may be used for either general or specified purposes. There are two major parameters in these arrangements: (1) the coverage of transactions, the proceeds of which may be retained, and (2) the percentage of receipts that may be retained within these groups of transactions. The coverage of transactions may vary for a number of reasons. In some instances, the proceeds of an original export are subject to 100 percent retention in order to give effect to the customs union. In other instances, the aim of the retention may be to promote direct investments re joint ventures. However, most often the retention rights are aimed at promoting certain export industries. These industries are usually distinguished as either traditional or nontraditional--although the export promotion bias may be towards either the former or the latter.

In categorising the export retention schemes, it is important to focus both on the restrictions that may be placed on the uses of retained foreign exchange and, if saleable, the exchange rate which is applicable. Retention allowances are a recognition that the exchange rate is unrealistic and that surrender at the official rather than the parallel exchange market rate cannot be enforced. In several countries, the foreign exchange may be sold through the medium of marketable certificates which enables the central bank to retain ownership of the exchange, but at the same time allows market determination of the incentive for surrender. In other countries certain transactions are permitted at negotiated rates that are either at the free market exchange rate, or close to it.

Exchange controls on capital transactions are subject to comprehensive restrictions in most developing countries, although a few (some 20) do not have controls on either the receipts or payments side of the market. A similarly sized group of developing countries have relatively liberal systems. Major categories of capital controls affect commercial banks’ international transactions and portfolio, direct, and real estate investments of nonbank residents. A characteristic of the capital controls is that they tend to blanket these various forms of asset transfers; if a country has controls on any one category of receipts or payments, then in general it will have controls on all. Capital receipts, however, are less controlled than payments—as might be expected in view of the present widespread foreign exchange scarcities in this group of countries. The receipts, however, are generally subject to repatriation and surrender requirements.

Most foreign direct investment controls include a review board or a government agency that can authorize investments, usually on a case-by-case basis. In some cases, the controls apply only to investments exceeding a certain value limit or in a certain sector. The controls may also entail incentives to investments in some sectors when development is planned, because the investment involves technology not available domestically. Foreign investors may be required to work jointly with nationals who are required to finance some proportion (often 51 percent) of the initial investment.

2. Considerations in the choice of restrictive system

In recent years, developing countries have increasingly opted for greater freedom of their exchange and trade systems. The general pattern of the main changes in restrictions in 1986 and 1987 is indicated in Chart 2, which shows the numbers of liberalizing and restrictive measures of different types (unweighted by the value of transactions affected by the measure). It shows clearly a preponderance of liberalizing measures, both overall and in each of the main categories of measures. Particularly striking is the broad movement towards liberalization of exchange and trade controls by developing countries--liberalizing changes exceed restrictive changes by a ratio of two to one. Also striking is the further progress towards dismantling exchange controls on capital movements. The liberalization by developing countries in these two years has also been at a more rapid rate than in industrial countries, albeit from a considerably more restrictive base.

CHART 2

MAIN LIBERALIZING AND RESTRICTING CHANGES IN EXCHANGE AND TRADE SYSTEMS, 1986-87

View Expanded

Source: IMF, Annual Report on Exchange Arrangements and Exchange Restriction, 1987 and 1988 editions.

Why has there been such a clear movement towards liberalization of restrictions in recent years? Perhaps the basic reason can be found in the increasing internationalization of information flows, with all Fund members being increasingly integrated into a global market place. This has made it more difficult for individual countries to isolate their systems from those of other countries, because the financial disadvantages created by the controls have become more evident. Even in the industrial countries, which have relatively sophisticated methods of administering control systems, capital controls have been abandoned in part because they are no longer effective. Goods are easier to trace than monetary flows, but nevertheless in a number of countries they have been shown to be permeable across borders, as smuggling has responded on a large scale to the financial incentives created by controls.

The ineffectiveness of import controls in meeting the social objectives for which they were designed is clear in a number of respects. To the extent that the controls are designed to limit the overall import bill without the need to depreciate the exchange rate, the question of efficacy is in fact two questions: (1) Were imports limited to the level sought, or did smuggling result? (2) If the authorities were able to make the controls effective, were they more efficient than raising the price of imports through a depreciation of the exchange rate, or has a depreciated exchange rate in fact been reflected in the price of the goods? In a number of countries, it is clear from statistical data that the final price to the consumer of imports that were restricted at an overvalued exchange rate has not reflected that exchange rate. When goods are correctly surveyed, i.e., at the final point of consumption or input, it is often found that they are considerably higher in price than would be calculated by converting the international dollar import price at the official exchange rate in the country concerned. The difference reflects the parallel or black market exchange rate, and is an economic rent attaching to the controlled scarcity of the goods. This rent accrues to importers, or in some countries to officials administering the import control system. This outcome is at odds with the stated social objectives where essential goods are imported with the aim of ensuring cheap foodstuffs for the lower income groups, or for supporting a productive industry through inexpensive raw materials.

The key point is that the same limiting effect could be achieved without distorting the structure of relative prices and thus creating artificial rents, by valuing the exchange rate at a realistic level. If social aims such as inexpensive food for the needy are sought, this can be achieved through more targeted measures such as a food stamps program, or a more progressive income tax structure. To subsidise imported foodstuffs for all income classes creates a number of distortions and is a very expensive method of obtaining the social effect. In a number of countries it has destroyed the domestic agricultural base.

Likewise, the social aim of supporting productive industry and employment in an infant industry phase is also inefficiently served by import controls. What often happens is that capital goods, mainly imported, are effectively subsidised by the correspondingly overvalued exchange rate, shifting the production function away from domestic Labor towards the imported capital goods, and adding to underemployment. The outcome is the opposite of that intended.

If there are concerns that the general price level will rise as a result of exchange rate adjustment and liberalization of the import controls, these are often allayed by looking at the true retail or wholesale prices at point of consumption or input into production. The impact effect can be made explicit through calculations of cost-push that examine the expected depreciation of the exchange rate, say to the parallel rate level, in terms of the ratio of imports to GDP adjusted for prices that already reflect the parallel exchange rate. Generally speaking, the impact effect of the exchange rate adjustment is considerably mitigated because the free market exchange rate is already embodied in the final prices. Because the free or black market exchange rate is the point to which a floating exchange rate tends to move following liberalization of the system, this gives some indication of the overall inflation effect, and is often much less than feared by politicians and officials. Finally, the impact effect should be offset by adjustments in macroeconomic policies that accompany the exchange rate adjustment. These may bring the inflation impact down to zero, or be aimed at sharply reducing the rate of inflation in effect before the reforms.

Nevertheless, liberalizing restrictions and allowing the exchange rate to adjust to an equilibrium level may be relatively slow in producing productive effects on exports. The lags in re-establishing export industry may be a matter of two to three years depending on the extent of the restructuring required. Import effects may be more immediate, but less politically tolerable.

In these circumstances, a major consideration in the short-run is the impact of the reform package on capital flows. While most developing countries have capital controls, very few of them have succeeded in stemming large capital outflows in recent years. In some cases these outflows would have been sufficient to finance the debt incurred by the country as a whole, and to avoid rescheduling. Clearly, setting in place incentives that can quickly stem the capital outflow or produce a reflow of capital has strong implications for the desirability of one-step liberalization packages. Evidence is mounting that a combination of freeing interest rates, and allowing the exchange rate to find an equilibrium level (both in the spot and forward foreign exchange markets) can serve as a strong incentive for the repatriation of capital. This can effectively smooth the transition towards the new set of relative prices resulting from the liberalization package by providing early support to the balance of payments.

1. Auction and interbank exchange markets

There are two main types of market arrangements for an independently floating exchange rate system, the interbank and the auction market. ^4/ The participants in an interbank market are commercial banks, and in some instances licenced foreign exchange dealers. Individuals and firms are permitted to bid through the commercial banks or dealers acting as their agents. Of the developing countries that have adopted an interbank system, most already had a sufficient number of commercial banks and foreign exchange dealers operating in the country to ensure a competitive environment. Experience has shown that it takes relatively few banks--even one or two may be sufficient--in order to ensure sufficient competitiveness, particularly if the government allows private individuals to transact freely outside the market. The key feature of this system is that the exchange rate is determined in negotiations between banks and their clients, and in transactions between the banks. The exchange rate is therefore free to vary from hour to hour and from day to day.

The role of the authorities in an auction system is a more central one than in an interbank market. Receipts from specified exports and services are surrendered to the central bank at the prevailing exchange rate and are auctioned by the authorities on a regular (say, weekly) basis. The central bank decides the amount of exchange to be auctioned and the minimum reserve price below which it will not accept bids. The basic difference between the auction system and the interbank arrangements is in the treatment of foreign exchange supplied to the market and the frequency or continuity of adjustment of the rate. The auction market requires the surrender of foreign exchange to a centralised point, which to date has been the central bank of the country organising the market. This has meant that the government has become more closely associated with the operation of the auction market, and the important objective of shedding political responsibility for the rate has not been realized. Another drawback of the auction market is that, owing to the surrender of foreign exchange to the market, the information available to bidders on the overall supply of exchange tends to be more limited than for the interbank market. Where the authorities have borrowed heavily short-term to sustain an appreciated rate, and then have had to go to the market to repay those loans, volatility has risen and the consequent sharp corrections in the exchange rate under an auction system have sometimes led to its abandonment. Interbank markets have been less prone to such destabilizing actions by government.

There is an important question whether floating exchange rates in developing countries have led to “free-fall” or over-shooting of the exchange rate. However, following introduction of the floating arrangements, the experience has been that the immediate movement of the rate has been either to the black market exchange rate, or to a rate somewhat appreciated with respect to the black market. Subsequently, there has been relative stability in the real effective exchange rate determined by the market systems. However, if domestic fiscal and monetary policies have been insufficiently moderated, then the real stability of the exchange rate has been achieved at a higher level of domestic inflation and, consequently, a depreciating nominal exchange rate. Fiscal and monetary policies have therefore been important in determining the extent to which the float has been feasible--because the public and the political groups benefitting from the previous system have tended to focus on the movement in the nominal rate as the gauge of the stability of the market.

2. Forward foreign exchange markets

There is a small but growing number of forward exchange markets in developing countries in which forward cover is provided by commercial banks to the private sector under competitive conditions. ^5/ However, more numerous in developing countries are forward cover facilities either provided by the commercial banking systems on terms that are officially regulated (and supported by official cover facilities provided to the banks), or provided directly to the private sector and public sector enterprises by the central bank or some other official institution. The central bank losses from risk exposure under nonmarket forward facilities have been extremely large in some countries--in some cases representing multiples of the monetary base. These experiences suggest strongly that governments and central banks need to be wary of participating in nonmarket cover arrangements.

There are several variants of market determined forward systems. In developing the market, it is clearly preferable to have commercial banks handle transactions as much as possible, and to have the central bank withdraw both its support for or regulation of the rate as early as possible. In the initial stages, training requirements are considerable because of the technology required. Experience in both industrial and developing countries suggests that outright forward contracts for commercial cover would be the most desirable point at which to commence operations. In any event, options often involve forward contracts and both options and futures markets require a large volume of transactions to be efficient. Futures and options markets may well emerge later, as in some of the more advanced developing country markets.

For developing the forward markets, the environment provided by the exchange restrictive system has been important in promoting, although not in determining, the flexibility of forward rate arrangements. Among the countries with market determined forward exchange rates, most maintain no restrictions on outward capital transfers or at least permit residents to hold foreign currency deposits. In all but one developing country with a market determined forward exchange rates, there has been no substantial ongoing parallel market for a spot exchange.

Forward exchange markets reduce the risk associated with foreign trade to the extent that the importers’ demand for, and exporters’ supply of, foreign currency are matched in the market at a given exchange rate, or the risk is shifted to agents (speculators) who are willing to assume it. Forward markets combined with realistic domestic interest rates make borrowing abroad attractive to importers. They are an important development where trade lines are open to a government experiencing temporary balance of payments difficulties, but importers are unwilling to assume the exchange risk in the absence of forward cover. Forward foreign exchange markets also facilitate some of the more sophisticated financial transactions that are necessary if the domestic banking sector is to develop an appropriate share of the international services market.

3. Import licence auctions

In the event that import licencing is retained for a period following an exchange reform, an auction system for import licences embodies the basic elements of the restrictive system, while at the same time improving efficiency within that system to the maximum possible extent. ^6/ Under these arrangements, importers bid for licences up to the total value allowable, but with no limitation on the type of goods that may be imported within that limit. Efficiency results because the licences go to those importers who place the lowest bids, thus squeezing out large margins that have often accrued to importers and have been borne by the public in general. The second element of improved efficiency in the auction arrangements is that licences flow to goods for which there is the most demand. This means that there are not surpluses or shortages of individual goods relative to the demand for them within each auction in which separate value limits are set for the various categories of imports. The more categories which have pre-assigned limits, however, the less efficient are the arrangements and the greater is the possibility of surpluses or shortages of particular categories of goods.

The auctions provide a clear indication of the extent of restrictiveness of the import controls that have been retained. This indication is provided by the size of the government revenue from the sale of licences. The revenue reflects the scarcity of imports created by the licencing system and may be an important element in narrowing the fiscal deficit.

4. Open general licensing

Selective decontrol of categories of imports in the transition to full liberalization of imports may be defined in terms of the range of commodities involved or on the basis of the origin of the foreign exchange for financing the imports. Open general license (OGL) or similar arrangements are maintained by a number of Fund members. The arrangements apply in most instances to product categories of imports, in contrast to specific licenses that are required for imports that are not under the OGL and are therefore subject to approval. The documentation required for open general licence is to permit checking for overinvoicing as part of the capital control and income tax regulatory structure.

The advantage of an OGL approach is that it may be calibrated to the degree of openness of the economy, with the list of OGL commodities being progressively broadened as liberalization proceeds. At some point an economy with an essential “positive list” approach will progress to the “negative list” stage. However, in terms of the documentation, the transition will not require an adaptation of the system itself, simply the master list against which the payments and customs clearance documentation is checked.

In OGL systems, the major issues relate to the level of the exchange rate for those goods on the free OGL list. An overvalued exchange rate could lead to serious overimporting of these goods and consequent deterioration of the balance of payments. Exchange rate or equivalent pricing action is therefore necessary to accompany adoption of such a system.

5. Reform of import tariffs

Tariff reforms have been initiated in conjunction with liberalization of imports in a number of countries during the late 1970s and early 1980s. However, most early reforms were suspended for a period under pressure from severe balance of payments problems in the wake of the second oil crisis. A second spurt of liberalization occurred in 1986-88. The underlying objectives of the tariff reforms have generally been to increase the efficiency of domestic production by increasing competition from abroad and improving resource allocation within the economy by rectifying an anti-export bias. In addition to the objective of improving efficiency, some reforms were explicitly aimed at dampening domestic inflationary pressures. In other cases, more specific targets were pursued: the pre-emption of retaliatory protection measures by trading partners, reduction of government interference, improvement of government revenue, and compensation for exchange rate devaluations.

There are several common structural elements observable in the tariff reform programs. Such reform is most often combined with, or preceded by, a reduction of quantitative restrictions. In some cases, tariffs were raised temporarily on items that were free from quota restrictions in order to give effective domestic industries time to adjust. Another feature of most programs was the simplification of the tariff structure. Approaches to simplication included unifying all import charges into one, and terminating exemptions or abolishing import reference prices. Another important aspect of the reform programs was the reduction in the dispersion of tariff rates. This was usually accomplished in tandem with the lowering of the average tariff rate--by the “concertina” approach of allowing tariffs up to a certain rate only, and leaving the rates below that unchanged, reducing high tariffs more than other tariffs, or increasing low tariffs while reducing all other tariffs. The reform programs were usually phased over four to six years, although shorter programs were initiated in some countries. The sequence of liberalization tended to start with a relaxation of quantitative restrictions, followed by simplification of the tax structure, and the reduction of the average level and distribution of tariff rates. The measures were at times undertaken concurrently.

6. Capital liberalization

To some extent it has been assumed in the literature that capital liberalization should take last priority in the sequence of second best liberalizing measures. However, it can be argued that freeing capital controls can play an important role in reversing capital flight in the initial stages of reform.

As noted above, about one in four developing countries maintain free or relatively liberal capital control systems. ^7/ In the other countries, the controls are normally severe and range across portfolio and real investments. Nevertheless, these controls have not been able to stem widespread capital flight in recent years. Indeed, a case can be made that the presence of controls has contributed to the capital outflows, because even where cover against exchange risk and realistic domestic interest rates are offered, problems of transfer resulting from the controls have induced residents to hold currency abroad (“mouse trap effect”). Residents will not repatriate capital to the domestic economy if, by doing so, they lose flexibility in its use thereafter. This suggests that simultaneous liberalization of the exchange rate, interest rates and capital controls is likely to provide the strongest incentive for reversing capital flight and supporting the balance of payments in the short term. Given that lags in the response of the real sector to exchange rate adjustment are likely to be long, the short run effects may be critically important for the success of the program.

The benefits of direct investment for growth and adjustment are readily apparent. The investment serves as a direct support to the balance of payments, and its impact on growth will be considerable even during the instalment of plant and equipment. Tax holidays granted in some countries may mean that the direct support from the increased fiscal revenues is delayed although the yield from taxation of incomes of the local population employed in the new industries will be immediate.

It is important to distinguish the motives for capital flight. In some countries, a relatively permanent pressure for flight may arise because of the presence of an expatriate population, some of whose ultimate plans are for resettlement abroad. There is probably little that can be done in such circumstances to solve the basic political problems by economic means, but attempts should be made to confine the economic damage as much as possible. Relatively generous arrangements for repatriation of emigrants’ capital, besides having a basis in equity, are more likely to be observed than highly stringent ones. In the extreme case, it may be that the “mouse trap effect” is strong enough that freeing up possibilities for depositing money abroad would actually lead to increased domestic deposits, especially if appropriate interest and exchange rate policies are put in place at an early stage.