Abstract

There is a small but growing number of forward exchange markets in developing countries in which forward cover is provided to the private sector under competitive conditions by commercial banks.15 In some cases, these markets have been introduced in association with floating spot exchange systems and are at an early stage of development (as they are in Jamaica, Nigeria, the Philippines, and Zaire). In other cases, forward markets have arisen in association with relatively advanced financial systems or relatively free exchange systems, or both (as in Brazil, Chile, Indonesia, Jordan, Korea, Malaysia, Singapore, Thailand, and the United Arab Emirates).

There is a small but growing number of forward exchange markets in developing countries in which forward cover is provided to the private sector under competitive conditions by commercial banks.15 In some cases, these markets have been introduced in association with floating spot exchange systems and are at an early stage of development (as they are in Jamaica, Nigeria, the Philippines, and Zaire). In other cases, forward markets have arisen in association with relatively advanced financial systems or relatively free exchange systems, or both (as in Brazil, Chile, Indonesia, Jordan, Korea, Malaysia, Singapore, Thailand, and the United Arab Emirates).

More numerous in developing countries are forward cover facilities, either provided by the commercial banking system on terms that are officially regulated (and supported by official forward 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. In such cases, access to forward cover is usually restricted either to traders or to rescheduled liabilities to foreign creditors. In some cases, official forward cover is provided at terms that are designed not to be loss making, or are intended to simulate the terms that a free market would offer. Examples are schemes for exchange cover of private sector debt service payments in Mexico and the Philippines, and cases where official cover is provided at forward premiums which deliberately approximate international interest differentials, so that covered interest parity holds as in free markets without exchange or credit controls, or political risk (although, as discussed below, such an approach will generally result in losses). In most instances, however, official forward premiums have been fixed for long periods without reference to market conditions, or have entailed deliberate subsidies from government budgets.

Even if forward cover is provided at estimated “commercial” terms, a central bank which sells forward foreign exchange to an importer or debtor will make losses if the domestic currency depreciates over the period of the forward contract by more than the forward discount implicit in the contract, unless the central bank closes its position by simultaneously buying spot or forward exchange. Since central banks (unlike commercial banks) have tended not to close their positions in this way—largely because of the absence of a sufficiently developed domestic market in which to cover the risks, or because of reserve constraints and the desire to avoid consequent pressure on the spot exchange rate which would result from such operations—forward exchange losses have been common. They have been much more common than profits because of the typical combination of a depreciating currency with a preponderance of official forward sales (over purchases) of foreign exchange.

Although the institution of forward arrangements may impose initial resource costs on the central bank, there are important macroeconomic benefits that flow from such a market. The first is the increase in efficiency and reduction of markups on imported goods that should result from lower exposure to exchange rate risk. The second is some improvement of the investment climate, as a more stable environment is provided for investors through the protection against short-term (but not longer-term) exchange risk. A third benefit is that forward markets reduce the needs of traders for working balances in foreign currency, and thus improve the overall availability of foreign exchange. A fourth benefit is that the arrangements will encourage importers to gain access to foreign sources of financing, thus providing further support to the balance of payments.

To show how investment is promoted by forward markets, take the example of agricultural investment. While the investment is needed over a season, receipts are realized only at the end; in this case, spreading the risk over time through the forward markets may help to promote the necessary annual investments. Another lag occurs when the produce is being marketed abroad—here again the forward exchange market would help eliminate this form of risk to the trader and spread it through the banking system. Lower needs for working balances may take the form of reducing balances in banks, releasing foreign exchange for use by the economy (in the sense that it augments central bank reserves and thus supports a country’s creditworthiness), or may lead to a repatriation of working balances held abroad by residents. In either way the forward market could contribute to the reversal of capital flight.

The exchange losses (or profits) resulting from official action in operating nonmarket cover schemes may have implications for the Fund’s jurisdiction under Article VIII, Section 3 which enjoins members to avoid multiple currency practices. Multiple currency practices have arisen in the operation of forward markets because the spot transaction that takes place to consummate the forward contract may involve an effective exchange rate that gives rise to a spread of more than 2 percent between buying and selling rates. To be considered a multiple currency practice, the Fund requires that the spread results from action by a member or its fiscal agencies in itself, and differs “unreasonably from those that arise from the normal commercial costs and risks of exchange transactions.“16 Since the 1981 Decision, several official exchange rate guarantee and cover schemes have been found to be multiple currency practices. The Fund has been cautious in exercising jurisdictional judgments as to whether or not the official losses and consequent effective exchange rate spreads are commensurate with normal commercial practice in this area. Practices that appear to involve jurisdiction have therefore often been subject to surveillance over several Article IV consultation cycles with the members, and the staff reports have in the meantime referred to the likelihood or possibility of a multiple currency practice. In this interim period the Fund has recommended steps that members might take to avoid subsidization—either in the context of Article IV consultations or of economic programs supported by the use of the Fund’s resources.

Exchange Rate Guarantees

The predominant form of forward exchange arrangements in developing countries involves an exchange rate guarantee provided, directly or indirectly, with official resources. In most instances the guarantee is provided for a fee, but in some (such as Costa Rica before 1983) the guarantee is for the exchange rate prevailing at the time that the guarantee is provided, without charge for the service. The guarantees are in some instances obtainable directly from the central bank (as in Argentina, Malta before 1981, and Mauritius), but in others they are administered, at a spread, by the commercial banking system (Table 3).

Table 3.

Summary Features of Forward Exchange Systems in Selected Developing Countries, December 31,19861

article image
Sources: International Monetary Fund, Annual Report on Exchange Arrangements and Exchange Restrictions, 1987; and national authorities.

For a more detailed description of the systems, see Appendix II.

Parallel (not officially recognized) market.

Provided only to public sector agencies engaged in trade in special circumstances (not being utilized in Egypt).

Market was inoperative in early 1988.

Central Bank provides forward cover to banks only for corporations or projects of vital national interest.

Transactions eligible for official forward cover of this variety generally involve the repayment of suppliers’ credits for imports (Pakistan) and provision of financing to domestic exporters (India). On occasion, the guarantees have also been applied by the government to debt service payments (as in Argentina, Costa Rica, Mexico, the Philippines, and Venezuela), mainly at the behest of creditor groups who have wished to ensure both the commercial and sovereign viability of obligations being rescheduled.

Forward cover is not normally requested for export receipts. One reason is that in many countries the domestic currency has been expected to depreciate by more than the available managed forward premium, so that exporters have been content to take the exchange risk. In fact, in many instances the government has attempted to take advantage of the additional local currency equivalent of the depreciation through the institution of foreign exchange surrender requirements which limit the extension of suppliers’ credits. There is no known case, for example, of exporters obtaining at the time of surrender the exchange rate prevailing at the later time of customs clearance or expiration of the export credit period.

An implicit rationale for forward cover provided to importers at a noncommercial or zero fee is that it is often viewed as increasing the overall availability of foreign financing to the economy through the trade credits made available to eligible importers. However, there is subsidization involved, because the net gain to the importer in obtaining risk insurance at zero premium is ultimately at the expense of the budget. Where a small fee is charged, the fee has tended to be inadequate compared to reasonably expected, and actual, movements of the spot rate—thereby resulting in a cumulative loss to the central bank.17

Market-Approximating Forward Exchange Rates

In several developing countries attempts have been made to approximate the workings of a market system of determining forward exchange rates, while retaining official regulation of the forward premium or discount. Relatively sophisticated versions of such schemes were introduced by Mexico in 1983 (FICORCA) and by the Philippines in 1985. In both instances, eligibility for participation was restricted to servicing debt outstanding at the time of a rescheduling, and the schemes involved relatively long periods of cover for the obligations—in fact, longer forward maturities than are usually available even in the forward markets of the industrial countries.

Both schemes utilized the covered interest parity condition noted earlier—and linkages through projected inflation rates to interest rates—to determine the calculated forward premium. In forward markets, covered interest parity will be maintained by riskless arbitrage, apart from a margin of indeterminacy resulting from transactions costs. Empirical evidence shows that the condition holds fairly precisely in the Eurocurrency markets, where these assumptions generally apply. Some forms of political risk may be reflected in interest rates and therefore need not affect adherence to the parity condition. In the case of the Mexican scheme, because of an original three-year grace period on repayments, and recent further restructuring, substantial payments from the FICORCA funds have yet to be made, although inflation and spot exchange rate depreciation have been more rapid than expected, and thus the FICORCA scheme is unlikely to have been self-financing. Nevertheless, there are specific provisions of the scheme for an acceleration of payments should the key assumptions of the projections inherent in the scheme not be satisfied.

A basic difficulty with this “quasi-market” approach is that the covered interest parity condition applies only when both domestic and foreign financial markets are free from controls. The interpretation of interest rates in terms of the equilibrium condition will be invalid when these assumptions do not hold, as is clear from major deviations from covered interest parity at times in industrial countries with less than perfectly competitive exchange systems. The calculated premiums will also tend to be biased downward in many developing countries that have constrained interest rates and apply credit controls. Dynamically, the situation may be even worse, as the low real interest rates at the outset of the cover period tend to feed into higher inflation, resulting ultimately in even lower real interest rates, and larger deviations of actual spot exchange rate movements from initial expectations. There is the further difficulty of forecasting the independent variables in the calculation, and there may be a tendency to be over-optimistic if the scheme is embarked upon as part of a broad program of economic stabilization. Nevertheless, the schemes represent an improvement over the straight provision of forward cover at either a zero premium, or premiums pegged without regard for inflation and interest differentials, and hence expected spot depreciation. The initial experience from both the FICORCA and Philippine arrangements suggests that a review clause should be included in any such formulas for determining the premiums, to take effect even before scheduled payments from the funds so that continuous realism can be assured. (See Appendix II for a fuller explanation of the Mexican and Philippine arrangements.)

Other applications of the covered interest parity condition by developing countries have offered less protection to the budget. Where the premium is set simply as the difference between the local and foreign interest rates, artificially low domestic rates have implied substantial losses. In such circumstances, shadow interest rates should be used to calculate forward premiums via the interest parity condition. A shadow domestic rate may be calculated from the expected rate of inflation plus a premium for time preference, risk, and other transactions costs. Given that the latter costs may be roughly equal between countries, application of a covered interest parity condition in this form will come close to the use of inflation differentials or purchasing power parity (PPP) to determine the forward premiums. Given the available evidence that PPP is broadly applicable in the longer term, use of the covered parity condition should result in minimizing extended losses in the provision of forward cover either by official or by officially supported agencies. However, losses may not be avoided and may be considerable in the short run under such managed schemes. The best approach is not for the central bank to assume or manage risk itself, but to encourage the development of a forward market in the private sector.

There are considerations other than avoiding adverse budgetary effects for the application of realistic premiums. An important reason is that the expected losses may themselves become a reason for not adjusting the spot exchange rate, thus extending inadequate competitiveness and balance of payments weaknesses into the future. In the projection of the inflation rate for this purpose, it may be necessary to err on the cautious side, by setting the premiums higher than might be implied by the targets for macroeconomic performance. Given that market participants may have strong impressions of past performance and will accordingly be slow to modify their own expectations, this is unlikely to result in the cover being unattractive to importers. On the other hand, the slow adjustment of expectations will deter exporters from using the scheme should the projected inflation rate (as modified for confidence effects of the adjustment program) be substantially lower than the previous inflation rate.

Cross Hedging

Cross hedging may be used where there is no domestic forward exchange market. It involves the use of a foreign forward market in an asset whose price has a large covariance with the exchange rate of the domestic currency. The most obvious example would be the use of the forward market in the currency to which the domestic currency is pegged. Tonga has recently instituted such an arrangement for the pa’anga, which is pegged at parity to the Australian dollar, through an Australian commercial bank. Another possibility would be to hedge currency risk in a forward market abroad for a commodity that is important to the payments position of the country, with the price of the commodity (which could be, for example, a major export item) serving as a proxy for determinants of balance of payments developments. This would be distinct from any hedging of the international price of the commodity itself (in commodities futures markets), as it is the foreign currency transactions rather than the sales revenues of the commodity that are hedged.

Although cross hedging may be valuable for closely integrated smaller economies where a fully fledged market arrangement may not be feasible, it may be of less value for more isolated economies.18 Even in closely integrated economies, the risk of divergent monetary policies causing a need to adjust the peg may be significant. The timing of step devaluations is politically determined in most instances and may be difficult to predict. In addition, cross hedging may have no application at all for countries with relatively diversified relationships—for example, those with close regional ties as well as close ties to a major currency country. Nevertheless, as a technique it is of more value than the provision of exchange rate guarantees with low or zero fees, and, unlike market-approximating techniques, it need not involve the central bank in risk, providing that the latter does not assume responsibility for variations between the hedge currency or commodity and the domestic currency (the Bank of Tonga, which is partly government owned, does assume this risk at present). Should it do so, then the potentially adverse fiscal and related effects are shared to this extent with the pegged schemes discussed above.

Market-Determined Systems

Outside the industrial countries, a small but significant number of Fund members maintain market-determined forward exchange systems. Although in recent years some seventeen developing countries have adopted market arrangements for determining their spot exchange rates, in only a few have relatively developed forward markets emerged. However, market-determined forward rates have evolved in several other countries maintaining fixed or managed exchange arrangements—mainly from official cover arrangements.

Forward markets in which rates are free to be determined by supply and demand and the range of external transactions is unregulated exist in some countries (such as Chile, Indonesia, Malaysia, Singapore, and the United Arab Emirates), but in others the availability of forward cover is limited to certain underlying commercial or financial transactions, usually the former. The forward market is thus limited to commercial transactions in Brazil, Jamaica, Jordan, Korea, Nigeria, Sri Lanka, and Thailand. As noted in Section I, the basic reason for such a restriction is the desire on the part of the authorities to curb speculative influences on the market (although the regulations may themselves lead to other forms of “speculation”). Another reason for the limitation to underlying “real” transactions is that such transactions tend to attract commercial lending. If domestic interest rates are relatively low and the exchange rate is depreciating, importers may not wish to take up otherwise attractive official and non-official credits abroad. The availability of forward cover without its costs being borne by the central bank may be sufficient to facilitate such borrowing from opportune sources when financing is tight.

There are several existing and potential variants of market-determined forward market arrangements in developing countries.

Auction Markets

It is, in principle, possible to devise mechanisms for forward auction markets similar to those used in auction markets for spot exchange (Bolivia, Ghana, Guinea, Jamaica, Nigeria, Somalia, Uganda, and Zambia have such spot auction systems).19 However, forward auction markets are likely to be more difficult to establish in practice because it may not be possible to predict the timing and supply of forward exchange to the auction accurately enough ahead of the actual transaction (foreign exchange is surrendered ahead of the spot auctions). For this reason, it would probably be necessary to have an iterative auction in which representatives of both the buyers and sellers are present at one time, so that bids could converge on an equilibrium price. No country at present operates such a forward system.

The only country that has a forward market as well as a functioning spot auction is Nigeria. However, this market, which is presently in an early stage of growth, operates in the autonomous (interbank) exchange market for non-oil foreign exchange and secondary sales of oil receipts, and not in the official auction market for channeling oil receipts to commercial banks.

Brokered Markets at the Central Bank

Brokering of forward transactions by the central bank involves in essence a barter operation. The central bank does not make the market on its own books, so that it does not take an exposed position itself, but matches up transactors at the various maturities on the basis of mutually agreeable rates. Owing to the coincidence of wants, a risk premium need not be charged for such a transaction—instead, a small charge may be made by the central bank for the cost of undertaking the brokering role (say of the order of one quarter of 1 percent of the transaction value).

One difficulty with such arrangements is the differing risk aversions of the two sides of the market. On the supply side, for example, exporters may be unfamiliar with the opportunity cost of not taking advantage of a better exchange rate and be content to take the depreciation that occurs. The result may be lack of supply of funds to the market and difficulty in matching the exporters and importers at mutually agreeable rates. Where the risk element is large—for example, where there is already a substantial differential between the current official and parallel market exchange rates or where the official rate is volatile—difficulties in evaluating future risk will be a more important consideration.

Funded Markets at the Central Bank

One role for central banks would involve setting up a small fund for the initial provision of foreign exchange at covered or qualified covered interest differential determined premiums, which would function in a similar way to the system of exchange guarantees provided by central banks discussed in Section III. The difference would be that, under a market-oriented approach, the aim would be to ensure eventual independence from the fund by adjusting the forward premium to clear the market. A major drawback could be that, in the meantime, the central bank would be exposed to losses if the premium for a depreciating currency was set too low at the outset and a supply-demand imbalance in favor of importers resulted in a run down of the fund. On the other hand, if the premium were set too high, there would be no demand for the facility. (This approach is explored further in the next section.)

Parallel Forward Markets

In some countries, parallel markets that are officially unrecognized may exist both for spot and forward exchange (this occurs in Argentina). However, in such instances the parallel forward market exists mainly to cover risks in the parallel spot market. Risks inherent in the availability of, and the exchange rate for, official foreign exchange may be different for all of the reasons that the official and parallel rates diverge. These include the timing of political action, controls on domestic and foreign money markets and capital markets, and the fact that disequilibria may not be evenly spread between the spot and forward markets. Nevertheless, the premium in the parallel market may provide an indication of the forward premium applicable to the official spot market. Evaluation of the usefulness of this information will require a separate analysis of the collinearity of the spot and forward markets, which will depend on the special factors involved in the individual country.

Forward Exchange Markets in the Private Sector

The development of private forward exchange markets in the developing countries that have adopted floating spot exchange rates is at a relatively early stage. There is no such country at present that could be considered to have an organized and satisfactorily operating forward exchange market. A limited volume of forward transactions has been observed in Jamaica, Nigeria, the Philippines, Uruguay, and Zaïre. In Jamaica, a plan had been drawn up for a forward market and the system was put in place in 1984, but there were few transactions under the floating rate system, partly because of inflexible interest rates which made trade financing in domestic currency more attractive. The suspension of the floating arrangements in the spot market since November 1985 is an additional complication. Nigeria instituted arrangements in October 1986 in the private sector for forward trade cover of up to six months, and some transactions have since taken place in the autonomous market. In the Philippines, developments in the organized forward market have been subject to general uncertainties and few transactions initially took place after the floating of the currency. In the past year or so, however, the market has stabilized and grown somewhat in depth. The relative thinness of interbank transactions in the spot market in Zaïre has precluded the development of the interbank forward market, although the authorities are considering policies to improve the operation of the spot market.

The major thrust of efforts to develop forward markets in countries adopting floating spot systems has been to give the commercial banks free rein to operate such markets rather than to have the central bank participate. The requirement of underlying commercial transactions in Jamaica, Nigeria, and Zaïre has been set mainly to focus the commercial banks’ attention on what has been considered to be the most important part of the market; it has also been motivated in part by the exchange controls on outflows of capital retained by these countries. In Nigeria and Zaire, forward transactions have in principle been included in the limits on foreign currency exposure (spot plus forward) introduced by the authorities to preclude market cornering by the banks. There is no indication that these limits have served to curb the development of the markets to date, although, should the markets expand, this could be a consideration against continuing them.

One market in this group that appears to have developed some depth is that of the Philippines. A reason for this may be that the Philippines has had in place for some time extensive official arrangements for forward cover which have increased the private sector’s awareness of the benefits of cover. The pre-existing arrangements may therefore have served in practice as a transitional device for the development of the private forward market.

Forward markets have evolved (mainly through official cover schemes) in several countries (such as Brazil, Chile, Indonesia, Jordan, Korea, Malaysia, Singapore, Sri Lanka, Thailand, and the United Arab Emirates) that do not have free floating arrangements in their spot exchange markets. The markets in Indonesia, Malaysia, Singapore, and the United Arab Emirates are not regulated. The market in Thailand is subject to some regulation, although it is minimal. Forward foreign exchange premiums or discounts in these markets have closely followed international interest rate differentials.

Requirements that the forward cover be for underlying commercial purposes, and that maturities correspond to trade financing terms, exist in a number of these markets. In Brazil, Nigeria, and Sri Lanka the forward market is open for commercial purposes. Only importers in Jamaica eligible for the spot auctions may participate in the forward market, and in Jordan the market is confined to specified commercial operations. Korea does not have restrictions on the coverage of forward transactions between banks, but transactions with nonbanks are restricted to underlying commercial transactions. The most common restriction on maturity limits is for up to six months (this restriction exists in Brazil, Chile, Jamaica, Nigeria, Sri Lanka, and Thailand). In Jordan, cover is permitted for up to one year, and the same rule applies in Korea for nonbanks (interbank transactions have unrestricted maturities).

Generally speaking, the central bank does not intervene in these markets directly—although the Central Bank of Jamaica has a provision to deal in the forward markets subject to overall limits on its own covered and uncovered commitments of US$2 million. Sri Lanka ceased intervening in the forward exchange market in February 1987. Limits on the open positions of individual banks exist in Brazil, Chile, Jordan, Thailand, and Nigeria, while in Jamaica there is an aggregate limit on the uncovered forward commitments of commercial banks (US$10 million).

Foreign Exchange Deposit Accounts

In many developing countries commercial banks or the central bank offer facilities for external accounts, with special exemptions from the exchange control regime. These accounts may involve either convertible foreign exchange per se, or foreign-exchange-denominated deposits, without a freely transferable claim on a foreign asset. Because the foreign exchange or the foreign-exchange-denominated deposit may be held by residents against future liabilities abroad, they constitute one means of hedging against exchange risk.

Relative to forward market cover of exchange risk, the difficulty with foreign exchange deposits that are held on account with the commercial bank or the central bank is that in the meantime they cannot be used by the depositor. They therefore represent liquidity in excess of efficient working balances. If the deposits are held with the central bank, their counterpart may be counted as official reserves, and may therefore contribute to the overall credit-worthiness of the country concerned. Where they are held with commercial banks, they may increase the supply of credit in foreign currency. However, the foreign exchange deposits are sometimes subject to special reserve requirements that may make them less attractive to banks as a source of liquidity, and their ready availability to the depositor for unforeseen uses may require the commercial bank to maintain foreign exchange reserves higher than otherwise. By comparison, provision of forward hedging facilities minimizes the need to tie up liquidity, often at less than remunerative deposit interest rates.

Foreign-currency-denominated accounts in a sense constitute forward hedging arrangements. Claims per se on foreign entities are not involved, and the major distinguishing feature of the deposits is the provision of cover against exchange risk (although there may be ancillary features of the accounts, such as treatment under exchange controls). The same questions, therefore, arise as those noted in the section on exchange rate guarantees—namely, the extent of subsidization by government, or the exposure of commercial banks to exchange risk beyond their ability to be involved prudently, or at unrealistic regulated rates. In some instances (as in Brazil) the deposits have constituted cover for debt-servicing obligations, with the implicit cost of cover arising from exchange rate risk being borne by the Government. (The provision of these accounts by Brazil has been marked by large losses, which have been monetized.) In other instances, foreign-currency-denominated accounts have served as a vehicle for resident foreigners to engage in transactions with their country of origin. In the event that the exchange rate has changed abruptly, there have been impediments placed on exchange transactions through these accounts because of the inability of the banking system to meet the guarantees (this happened with the MEXDOLLAR deposit scheme in 1982). Further, although liquid foreign exchange is not tied up in these accounts, liquidity requirements for domestic currency deposits are higher than those for forward contracts in foreign currency, resulting in the same inefficiencies of excess cash balances for individual depositors.

Another vehicle for implicit hedging of exchange rate risk is the swap transaction, as discussed earlier. On-lending of foreign credits by governments to the domestic private sector (by which the government undertakes the debt servicing in foreign currency and the private sector in domestic currency) is a common form of such implicit hedging arrangements in developing countries. Frequently, the interest rate will be swapped at the original foreign rate (say, at close to LIBOR). If domestic interest rates are controlled at relatively low levels, the implicit subsidization may appear to be relatively low. However, the counterpart of the low domestic interest rate is likely to be a rate of domestic inflation that substantially exceeds that in the country in whose currency the foreign debt has been arranged or denominated. As the servicing proceeds, the exchange rate must be adjusted to restore competitiveness, raising the local currency cost of repayments to the government, but not to the ultimate private sector user of the credit.

In arranging such swaps, the authorities must therefore evaluate the role of exchange rate risk in determining the interest rate at which the on-lending is to take place. For this, the considerations are the same as for forward transactions—where no domestic forward exchange market exists, the cost of the implicit forward cover may be established by the various criteria discussed earlier. In the case of the swap transaction, the question of who should pay this cost—the government or private sector user of the credit—may depend on the use to which the foreign currency proceeds of the loan are put. If the intention of the foreign credit is simply to provide the government or a domestic third party with foreign exchange, based upon the creditworthiness of the ultimate recipient of the credit who uses the foreign currency proceeds either in part or not at all, then the cost of hedging the exchange rate risk would need to be allocated to the government and the domestic third parties. On the other hand, if it is the creditworthiness of the government (or a third party) that is being used to obtain foreign exchange for the ultimate borrower, then the cost of hedging would need to be factored into the on-lending charges. In short, the exchange rate risk should be realistically established, and corresponding forward premiums should be charged to the ultimate end-user of the foreign exchange proceeds, who may or may not be the borrower—in addition to the interest charged to the borrower.