The objective of Fund-supported adjustment programs is to bring about a viable balance of payments in the medium term. The formulation of exchange rate policies in such programs takes into account the overall stance of domestic and foreign policies, as well as prospective internal and external conditions over the medium term. Such an analysis may bring out the implications of various exchange rate policies, or, conversely, may be used to indicate the exchange rate policy consistent with a given set of assumptions about exogenous developments and the stance of other policies.
A variety of exchange rate indicators are available for use in such a framework to indicate both the need for an exchange rate action and its general magnitude. These indicators also can provide a basis for determining the amount of the action. If the rate is to be managed, indicators are needed as a guide to what the exchange rate should be, or how it should change over time. Even where the rate itself is to be determined by the market—that is, where the rate is floated— indicators play an important role in assessing the sustainability of the rate in light of the medium-term objectives of the program, thus helping to guide the formulation of other program policies.
While no one indicator is wholly reliable for assessing the exchange rate, if a number of them are used in combination, and if proper account is taken of their statistical and methodological limitations, an informed judgment can be made. Because indicators are interpreted in the context of a medium-term adjustment program, conclusions drawn from them need to be modified in the light of other developments, the stance of other policies, and, of course, the program objectives. In particular, because all indicators are in some sense based on the historical structure of the economy, conclusions drawn from them need to take into account any necessary structural changes in the economy.
Elasticities
Assessment of the degree of responsiveness of the elements of the balance of payments to the exchange rate—elasticities analysis—is integral to the formulation of exchange rate policies. Often the use of elasticities is implicit. A medium-term scenario, for example, may be expressed in terms of there being general consistency among the various values projected. The test of consistency, in effect, is that the implied elasticities be reasonable.
Elasticities may also be employed more formally, through econometric estimation followed by simulation, to arrive at a recommendation on the exchange rate, given a set of objectives and policies in other areas. Such estimation can be performed with varying degrees of sophistication. A simple and commonly employed approach relies on reduced form export-supply and import-demand functions to derive the effects of specified exchange rate adjustments on the current account. By influencing domestic relative prices, an exchange rate adjustment exerts real effects on import demand and export supply. The elasticity of demand for tradable goods, which partly determines the import-demand and export-supply responses, will depend largely on the income effects of the devaluation and the accompanying policies, because the elasticity of substitution between tradable and nontradable goods is likely to be small. In addition, the elasticities of import demand and export supply will depend on the supply response in the tradable goods sector.
Econometric estimates of supply elasticities vary widely, depending on the choice of country and commodity, the period covered, and the methodology. Hence, an evaluation of the expected supply response needs to take into account the country’s particular circumstances. Specifically, the size and speed of the supply response depend on the extent to which it results from (1) putting to use previously idle resources; (2) increased productivity through more intensive use of resources; (3) movement of resources from the nontradables to the tradables sector; and (4) movement of resources, within the tradable goods sector, to more productive employment (e.g., from previously protected import substitutes to exports). The supply response will be quicker and more substantial where the first two conditions prevail than in cases where a transfer of resources and an associated contraction in other sectors is required. The supply response also will be more favorable in countries where resource mobility is high. Among other things, resource mobility depends on the flexibility of markets and the adequacy of infrastructure. The supply response also depends on the gestation period of investment, which may vary widely according to the particular circumstances of individual countries.
In contrast to the indicators of competitiveness discussed below, elasticity analysis has the advantage that it is not limited to comparison with a particular base period. It also lends itself to formal modeling of the way factors, other than the exchange rate, affect the balance of payments. Like indicators of competitiveness, however, it is based on historical experience, and accordingly needs to be qualified in the presence of structural change. Data limitations, moreover, often make reliable econometric estimates of elasticities difficult to obtain.
Indicators of Competitiveness
Real Effective Exchange Rate Indices16
The indicator of competitiveness most commonly employed in determining the appropriate level of the exchange rate is the real effective exchange rate index, which is formulated as the nominal effective exchange rate index adjusted for movements in prices or costs at home relative to those abroad. An increase17 in a country’s real effective exchange rate index over its level in a base period when the external position was considered adequate thus provides an indication that external competitiveness has deteriorated. Competitiveness can be restored to its base period level through a change in the nominal exchange rate, a change in the level of domestic prices or costs relative to those abroad, or some combination of the two.
Real effective exchange rate indices can be constructed with various patterns of partner country weights and with various indicators of costs or prices. The country weights used in indices developed in the Fund typically reflect bilateral non-oil trade flows, although in some instances they reflect the importance of export competitors in foreign markets. In developing countries, the consumer price index is commonly used as an indicator of cost developments domestically and abroad, because better indicators, such as unit labor costs, are usually not available. The consumer price index is generally preferable to the wholesale price index, because the former gives a larger weight to nontraded goods, thereby reflecting more fully developments in costs and prices of domestically produced goods and services. The consumer price index is preferable also because it is an important determinant of labor costs.
Certain features of consumer price-based real effective exchange rate indices need to be taken into account in interpreting the indices. First, in the presence of differing productivity trends across countries and across sectors, such indices are meaningful proxies for changes in relative costs only in the short run. Second, where exchange rate adjustments and other policy changes are being undertaken to correct an inappropriate exchange rate, the desired adjustment of relative prices is likely to involve a divergence between unit labor costs and consumer prices. This is particularly so if traded goods are prominent in the consumer price index. Under these circumstances, an increase in consumer prices might be associated with successful adjustment of an overvalued currency, and would not necessarily indicate a need for further adjustment. On the other hand, where the exchange rate is rigid or adjusted sluggishly in the face of relatively high domestic inflation, the same argument suggests that the index might understate the extent of real appreciation. Third, wage or price controls also give rise to divergences between movements in unit labor costs and consumer prices, and need to be taken into account in interpreting the indices.
While the limitations of real effective exchange rate indices, like those of other indicators, need to be kept in mind in using them, such indices have an important advantage in that competitiveness—relative profitability—as a concept is relatively easy to interpret and discuss. Other indicators, such as elasticities, may have less intuitive appeal.
In using real effective exchange rate indices, however, it will often be necessary to go beyond restoring competitiveness to the level prevailing in some base period when the country’s external position was regarded as sustainable. If domestic and foreign policies and conditions are more adverse than in the base period, there will normally be a need for both decreased real domestic absorption and an expansion in the tradable goods sector in relation to its previous size. This would apply, for example, when a country’s terms of trade have deteriorated, resulting in a decline in real income and an external imbalance. This would require improved competitiveness beyond that of the base period. In many cases, moreover, long-standing disequilibrium, during which the balance of payments deficit has been held to financeable levels only through severe restrictions, means that competitiveness needs to be enhanced beyond the level recorded in any recent period.
Internal Terms of Trade
The internal terms of trade index (ratio of prices of traded goods to those of nontraded goods) is an indicator of the internal competitiveness of the traded goods sector, i.e., its ability to compete with the nontraded goods sector for scarce factors of production. Such a relative price index has the advantage that it reflects adverse developments in external prices, which are likely to be particularly important in the case of primary exporting countries.
While traded goods price indices are relatively common (e.g., import and export unit values), nontraded goods price indices are more difficult to calculate because commonly available price indices, such as consumer price indices, wholesale price indices, and gross domestic product deflators, generally do not separately identify nontraded goods. Consequently, movements in the ratio of an index of traded goods prices to any of these indices would tend to understate changes in the internal terms of trade.
Internal terms of trade indices share many of the limitations of real effective exchange rate indices. For instance, changes in the internal terms of trade that result from differing productivity trends among sectors do not indicate changes in the profitability of the traded goods sector. As with real effective exchange rate indices, moreover, internal terms of trade indices focus on changes that have occurred in competitiveness after some base period and do not allow for other developments that also affect a country’s external position.
Commodity-Specific Analysis
In most developing countries, production of tradables is concentrated in a few key commodities. Analyses of developments for particular commodities cap help avoid the aggregation problems inherent in broad indicators of competitiveness. Such analyses can be used retrospectively to determine the change in the exchange rate needed to restore the profitability of production of particular commodities to the level observed in some base period, but can also be helpful in considering what the response to different exchange rates is likely to be.
Commodity analyses can obviously be carried out with varying degrees of sophistication. The domestic resource cost approach18 is particularly appropriate where distortions in costs and prices are widespread in the context of publicly administered prices and resource allocation. Under these conditions, an exchange rate action, even if accompanied by proper macroeconomic policies, would not direct resources via price to the most productive uses. Likewise, competitiveness indicators yield biased results to the extent that they reflect distortions in the economy.
The domestic resource cost approach derives the domestic cost of producing exports and import substitutes per unit of foreign exchange earned or saved, after correcting for all price distortions and netting out taxes and subsidies. This provides for each commodity an implicit exchange rate (local currency per unit of foreign currency), allowing a ranking of activities according to comparative advantage and an assessment of the prevailing exchange rate. A structural reform based on such information would ensure that government decisions on prices and allocation of resources are consistent with considerations of comparative advantage. Furthermore, the information on implicit exchange rates, together with estimates of the potential output of individual traded goods, would allow the construction of a supply curve of exportable and importable commodities as a function of the exchange rate. The appropriate exchange rate could then be determined on the basis of a more comprehensive analysis that takes into account demand conditions and balance of payments objectives.
An important advantage of the domestic resource cost approach is that it focuses directly on profitability, the key consideration in production decisions. Because of its data requirements, however, it seldom can be rigorously applied. Moreover, by focusing on major commodities, it may tend to divert attention from other aspects of the balance of payments and from the way in which the exchange rate interacts with other policies. It is particularly important to guard against the temptation to resort to piecemeal adaptations of policies with respect to specific commodities rather than to aim at comprehensive adjustment of price and exchange rate policies. Even more than other indicators, therefore, the results of an analysis of domestic resource costs need to be checked against other relevant information.
Parallel Markets
None of the indicators discussed so far reflects market forces directly, and thus there is no guarantee that they accurately reflect the opportunity cost of foreign exchange. Even where the exchange rate continues to be set administratively, indicators based on market developments can be helpful in exchange rate management.
The existence of a parallel or secondary market, where a portion of current transactions takes place at floating exchange rates that are more depreciated than the rate in the official market, is prima facie evidence that the official exchange rate is inappropriate. A distinction may be made between legal and illegal parallel markets in this regard. A legal parallel market typically is introduced when the authorities, instead of proceeding with a uniform devaluation to correct an external imbalance, decide to establish, generally as a transitional measure, a dual (or multiple) exchange rate system. In such systems, some transactions, such as certain essential imports and specified exports, are carried out at the relatively appreciated official rate, while other transactions, which often include capital flows, are assigned to a legal secondary market. If the exchange rate in the secondary market is floating, it can help determine the appropriate level for the unified rate, although its significance depends on the importance of the secondary market and the extent to which it is free to reflect market forces.
The smaller the secondary market (as measured by the value of foreign exchange transactions) relative to the official market, and the higher the demand and supply elasticities of foreign exchange in the official market relative to the secondary market, the closer the equilibrium unified rate is likely to be to the official rate. However, if there is a large unsatisfied demand in the official market which cannot be diverted to the parallel market because of administrative restrictions, the equilibrium rate in a unified market would tend to be closer to the parallel market rate, or could even be beyond it.
The rate prevailing in an illegal parallel market is a less reliable indicator of the restriction-free equilibrium exchange rate. That market’s supply of foreign exchange can come from a variety of sources. These could include proceeds from smuggled goods, over-invoicing of imports, underinvoicing of exports, and unrecorded workers’ remittances. With a strict enforcement of the foreign exchange controls, the penalty for dealing in such a market can be substantial, reducing the total (i.e., official and parallel market) demand for and supply of foreign exchange relative to their restriction-free levels, and limiting the size of the parallel market. In general, the exchange rate in an illegal parallel market will be more depreciated than the restriction-free exchange rate.19 Due to the probable thinness of an illegal market, the rate is also likely to be subject to significant fluctuations.
An intermediate situation is that of an illegal, but tolerated, parallel market. Since participants in such a market typically operate outside normal channels, transaction costs are likely to be higher than in a legal parallel market, though lower than in a strictly illegal one. A strictly illegal market often develops into a tolerated one as its scope of operations expands and the authorities recognize its relative benefits and inevitable character.
Other Considerations
This section has emphasized the fact that exchange rate indicators cannot be considered in isolation. The recommendations that result from them need to be developed in the context of the objectives of the program and its other policies. Likewise, the recommendations also need to take account of developments that particular indicators may not fully capture. For example, as noted above, a deterioration of the terms of trade from some base period implies a need to enhance competitiveness beyond the level recorded in the base period. Similarly, indicators of exchange market pressures, such as persistent intervention in one direction, rising foreign borrowing, growth of arrears, or stricter rationing of foreign exchange, can provide qualitative guidance. The following section discusses some of the ways in which conclusions derived from main indicators are modified in the light of such circumstances.