Faculy Research............ of X............ C,.f 1 *ll 4)1IIJl

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK THE UNIVERSITY OF MICHIGAN BUSINESS SCHOOL WORKING PAPER FOREIGN EXCHANGE MARKETS AND CURRENCY RISK April 2002 Gunter Dufeya and Pheng-Lui Chngb Forthcoming in Handbook of Modern Finance. (Dennis Logue and James Seward, eds.) New York, RAI, 2002. aProfessor of CSIB and Finance, The University of Michigan Business School and Adjunct Prof., NTU/ Nanyang Business School. <gdufey@umich.edu> b Senior Lecturer, NUS Business School and Visiting Assistant Professor, TheUniversity of Michigan Business School (Fall 2002). <bizcpl@nus.edu.sg> 1

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK I INTRODUCTION Business firms have internationalized their activities considerably. This trend has manifested itself not only in increased involvement in international trade and foreign operations, but also in the fact that even firms without explicit international transactions have become subject to the direct and indirect effects of foreign competition to a much larger extent than in the past. Thus, the impact of exchange rate changes on business operations tends to be pervasive, and concern is not limited to specific financial functions such as corporate treasury.l For the purposes of analysis, it is useful to consider foreign exchange (forex) rate changes as a source of risk, distinct and separate from other risks, such as shifts in technology, consumer preferences, labor relations, and government policies. The impact of exchange rate changes on the firm must not, however, be viewed in a simplistic, mechanical fashion. It is necessary to take account of all effects on the firm for which a causal link with exchange rates can be identified. For example, exchange rate changes are often directly linked to exchange controls or drastic shifts in macroeconomic policy. By the same token, forex rate changes may cause adjustments in consumer behavior or may alter the future path of labor costs in a systematic fashion.2 II THE FOREIGN EXCHANGE MARKET [1] Basic Function of the Foreign Exchange Market The world economy is composed of nation-states. A fundamental economic characteristic of modem nation-states is the exercise of monetary sovereignty3. Technically, this involves two distinct aspects: 1. The government. Usually through its agency, the central bank, the government maintains a monopoly on supplying the means of payments. The issuing of coin and paper currency, together with the institutional arrangements regulating the volume of bank demand (current) liabilities used to transfer funds among private and public entities in the economy, represent the essential aspects of the supply of money in a national economy. 1 Waters (1979). 2 Oxelheim and Wihlborg (1987). 3 One glaring exception to the rule of one-country-one-currency is the European Union that has adopted the Euro (EUR) as its common currency. Other exceptions would include countries like Liberia, Panama, and more recently Ecuador, where the USD is used as legal tender. 2

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK 2. Legal tender laws. In order to assure the use of the national currency in good times and bad, these laws stipulate that debts to both private and public parties be legally discharged with the transfer of an appropriate amount of the national money. In other words, creditors cannot refuse payment in local currency. All of this means that money is very much a national phenomenon, useful only within the political boundaries of a country. By the same token, money never leaves the country, casual observation notwithstanding.4 What is usually referred to as an outflow of money in the context of international transactions is in reality merely the transfer of the ownership of means of payments, usually demand deposits in the national banking system, from domestic to foreign residents. Technically, demand deposits never leave the national payments circuit. Thus, in order to make payments abroad, it is necessary ultimately to obtain forex or, in other words, ownership of the means of payment (currency or demand deposits) in the country of the recipient of the funds. The recipient may accept foreign means of payment but usually sells the foreign money immediately in order to obtain the domestic money needed to pay expenses. Therefore, international payments involve exchanges in ownership of national means of payment. This exchange occurs in the forex market, where buyers and sellers of currencies interact (that is, one country's means of payment is bought in return for another country's means of payment, which is then sold). Parties to international transactions, such as exporters and importers, usually do not participate directly in the foreign payment systems; instead, their banks act for them. Also, central banks may and often do participate in the forex market, attempting either to influence the value of their respective currencies in pursuit of national economic goals or to make international payments related to "autonomous" purposes, such as servicing debt, importing merchandise on behalf of government agencies, increasing ownership of foreign liabilities (reserves), and giving foreign aid. Anyone whose consumption, saving, or investment behavior leads to international payments is directly or indirectly involved in the forex market. Cross4 Merchants may accept currency from another country as a service, but they will bring it quickly to the local bank, where it is counted, packed, and shipped back to its country of issue. Also, when people lose trust in the future value of local money, they may attempt to use another country's means of payment, but this is exactly what legal tender laws are designed to curtail. 3

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK border payments arise in conjunction with almost all international transactions, i.e., gifts, exports, imports, service transactions, and investments (lending or borrowing). The notable exceptions among international transactions not involving the forex market are private or public donations of real goods and certain intergovernmental transactions, including those with international agencies such as the World Bank or the International Monetary Fund. All transactions by the participants in the forex market affect the demand or the supply of foreign means of payment, which are called forex, or foreign currencies. The interaction of supply and demand in the market for such monetary assets determines the exchange rate at any given point in time. It is defined as the price of one unit of currency in terms of another currency. Therefore, transactions in the forex market can be expressed in terms of supply and demand of national currencies. Specifically, transactions classified as credits in the balance of payments - exports of merchandise, services, and claims (capital imports) - are sources of forex and result in an increased supply of forex relative to the supply of domestic currency. The increased supply of forex and the increased demand for the domestic means of payment caused by payments made by nonresidents for domestic exports will result in the domestic currency's appreciation (i.e., an increase in the domestic currency's value relative to foreign currencies). Transactions classified as debits in the balance of payments - imports of merchandise, services, and claims (capital exports) - are uses of forex and result in increased demand for forex relative to the demand for the domestic currency. The increased demand for forex and the concomitant additional supply of the domestic currency result in the domestic currency's depreciation (i.e., a decrease in value relative to foreign currencies). Because of the reciprocal nature of the forex market, in which the demand for one currency equals the supply of the other and vice versa, each transaction affects the exchange rate through its effect on supply and demand. [2] Foreign Exchange: Spot and Forward Rates So far, it has been assumed that there is no difference between the time buyers and sellers agree on a price (rate) and the time the "merchandise" (ownership of demand deposits) is actually delivered. However, as in many other markets, agreement on a rate (price) and actual delivery (payment, settlement, or clearing) usually happen at two different times. 4

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Conventions in the forex markets call for delivery on the second business day, after a rate has been agreed on, for so-called spot transactions.5 Trades that call for delivery on the third day or later are known as forward transactions. Typically, forward transactions are quoted for "even dates" (i.e., one week, a month (30 days), or multiples thereof.)6 However, banks accommodate corporate customers for "odd" dates at a small price in the form of a less advantageous rate. The general purpose of a forward transaction is to fix the exchange rate for the respective maturity, instead of waiting for the (uncertain) spot rate that will prevail just prior to that time. [3] Controlling Risk Why would anyone want to fix the exchange rate contractually for a future period? One reason is the anticipated receipt of foreign currency funds in the future or an obligation to make a payment at a future point in time. The classic illustration is a U.S. importer who orders merchandise from abroad, knowing that he must pay, e.g., CAD 100,000 (Canadian) 90 days later. By purchasing CAD 100,000 90 days forward for USD 65,000 (U.S.), he knows exactly the USD price of the merchandise, on the basis of which he has made his decision. In other words, through the forward market he has "locked in" the USD price; the explicit cost is unaffected by the vagaries of the spot rate between the time he commits to the purchase and the time he has to deliver the funds. To put it differently, the transaction is covered (hedged);7 should the future spot rate of the CAD rise above the contractually agreed-on forward rate of USD 0.65, e.g. to USD 0.66, the importer incurs a loss on his payable (liability), because CAD 100,000 now costs USD 66,000 (i.e., USD 1,000 more.) On the other hand, he has gained on his forward contract, which obliges the other party (usually a bank) to deliver CAD 100,000 for USD 65,000. Relative to the then-prevailing spot rate of USD 0.66, he has gained USD 1,000 on the forward contract (an asset) that guarantees he will receive 5A holiday in either country simply moves the settlement day for both currencies one day further. USD against CAD and MXP (Mexican pesos) are typically traded for next-day value. Modem communications technology sometimes permits dealing in European currencies for next-day value ("Tom next"), or even same-day settlement (cash), but the rate quoted to the party that requests these special terms will be slightly different from the spot rate. 6 In order to find the settlement date for, e.g., 3-month (90-day forward) rates, one finds the value date for the spot transaction and checks whether the same date 3 months hence is a business day, as long as the business day is within that same month. If the value date would fall into the next month, the last (common) business day that occurs in the settlement months is used. 7 Within the narrow context of the forex market, the terms "hedging" and "covering" are used synonymously; in the context of international investment, finer distinctions must be drawn. 5

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK CAD 100,000 for only USD 65,000. Thus, the loss on the liability was exactly offset by the gain on the asset. Of course, should the spot rate move in the other direction, e.g., to USD 0.64, his potential gain on his payable is offset by the loss on the forward contract. But such is the nature of hedging; one gives up the opportunity for both gains and losses, focusing instead on the profitability of the underlying transaction that will give rise to the future receipt or payment of foreign funds: exports, imports, investments, and borrowings. There is a second reason for fixing the exchange rate for some future date. If one feels strongly that the future spot rate, say 90 days hence, will be lower than the 90 -day forward rate available now, one might want to engage in a little speculation: entering into a forward contract to exchange, e.g., CAD 100,000 for USD 65,000 90 days from now will yield USD 1,000 if the expectation comes true and one is able to purchase CAD at the then-prevailing rate of USD 0.64. Indeed, money can be made at any future spot rate that is lower than the forward rate of USD 0.65, but one would lose money if the spot rate happens to be above USD 0.65. Because of the profit potential, many people are constantly attempting to predict the future spot rates and enter into forward commitments. All of these transactions will not be without effect on the forward rate relative to the current spot rate. To illustrate, say that today's spot rate for CAD is USD 0.64, but market participants as a group think the CAD is going to depreciate (weaken), that is, the spot rate is likely to drop within the next 3 months. In that case, the 90-day forward rate will be lower than today's spot rate, e.g., USD 0.63. Thus, the difference of USD 0.01 reflects market expectations; it represents the market's forecast of the future spot rate 90 days from now. But the difference between the present spot rate and the present forward rate has additional economic content. Note, in the case of the U.S. importer's problem with the CAD payable, the use of the forward market is not the only way to remove uncertainty stemming from the future spot rates. The importer could simply borrow an appropriate amount of USD when he enters into the commitment to pay CAD 100,000 90 days hence;8 he would then exchange them at the prevailing spot rate, invest them in Canada for 90 days, and use the funds at maturity of this CAD asset to pay off the CAD liability. This transaction is known as a money market hedge. 8 The appropriate amount of USD is equal to that required to purchase CAD 100,000 less the amount of interest earned in Canada for 90 days. 6

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK When comparing the two alternatives, the importer will look at: (1) the difference between the interest rate for the USD loan less the interest on his CAD investment, and (2) the difference between the spot rate and the forward rate. Thus, if the spot rate is USD 0.64 and the forward rate for 90 days is USD 0.63, the difference would amount to USD 1,000 on a purchase of CAD 100,000 forward. If the amount of interest received on the CAD investment exceeds the interest cost on the USD loan by more than USD 1,000, the importer will protect himself through the money market (borrowing and investing) transactions; if the interest difference is less than USD 1,000, he will find it advantageous to cover through the forward market, i.e., purchase CAD forward (sell USD forward). These illustrations show that the difference between the spot rate and the forward rate not only reflects market expectations but is analogous to interest. A well-known approximation formula allows conversion of an exchange rate differential into a per annum rate: Forward rate - spot rate ( 360 1 spot rate day in forwardcontract) = + premium or discount per annum For the above illustration: USD 0.63 - USD 0.64 360 ) ) (1 00 = - 6.25 per annum USD 0.64 90 Thus, the 90-day forward CAD trades at a 6.25 percent per annum discount to the spot CAD relative to the USD.9 9 From a Canadian perspective, one would say that the USD trades at 6.35 percent per annum premium, since 1 - 1.5873 0.63 1 - 1.5625 0.64 1.5873-1.5625 (360 \ 1.5873-1.5625 360(100)=6.35%per annum 1.5625 90 7

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Once the interest equivalence of the forward premium or discount is understood, a new way of making money opens up: international interest arbitrage. Arbitrage is defined as the simultaneous buying and selling in two or more markets to profit from price discrepancies for assets of the same riskiness. If the interest rate differential between, e.g., the United States and Canada is not equal to the forward differential, one can undertake international borrowing and lending (investing) and cover the exchange risk with a forward transaction. To illustrate, if interest rates on 90-day USD loans are 10 percent per year and Canadian money market rates are 16 percent, one can move funds to Canada by exchanging them at the spot rate of USD 0.64, "cover" the investment with a forward contract at USD 0.63, and earn 1.2 percent (16 - 10 - 4.8) without exchange risk'0. The simultaneous purchase or sale of one currency (in this case, CAD in the spot market) for one value date and the offsetting sale or purchase of the same currency for another (here CAD 90 days forward) is called a foreign currency swap. Hence, the per annum forward premium or discount is known as the swap rate or the swap basis. Because many people are attracted by the opportunity to earn returns without exchange risk, such opportunities disappear rapidly; the activities of international interest arbitragers (if unrestricted by governmental controls) affect spot and forward rates as well as interest rates (for instruments of equal riskiness) in the respective currencies until interest rate parity is established, as shown by FWD (1+iA) SPOT (+ iB) where iA and iB refer to the representative interest rates in countries A and B, respectively, and FWD and SPOT refer to the price of a unit of currency B expressed in terms of currency A. For many currencies, viable forward markets do not exist. Invariably, this is because of the existence of (or fear of impending) exchange and credit market controls. It is pertinent that the USD is the reference currency in all forex transactions. Hence, all forex calculations are done from a USD perspective. Exceptions to this are the U.K. pound (GBP), Australian dollar (AUD), and New Zealand dollar (NZD). Also, the percent per annum appreciation and depreciation are not exactly equal as the base used to calculate the premium/discount is different. This is referred to as the 'Change paradox'. 10 Settlement risk and credit risk still exist. Thus, it is not quite correct to call this a riskless transaction. 8

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Even for the small number of major currencies for which relatively unregulated, active markets for medium-term and long-term credits at fixed interest rates exist, forward facilities beyond 12 to 18 months used to be quite rare. One explanation is that longerterm forward contracts would require bank traders to create offsetting asset or liability positions that are regarded as risky and impinge on established capital adequacy norms. With the growth of active offshore markets for fixed-rate notes and bonds in various currencies and refined hedging techniques, markets for long-dated forwards have improved considerably. Where markets are subject to exchange controls, but where there continues demand for hedging these controlled/restricted currencies, banks have devised new foreign exchange contracts, so called non-deliverable forward contracts (NDF's) to cater to these needs. NDF's are designed similar to f/x futures, except that they focus on currencies that have a history of being subject to controls. They are also often referred to as exotic currencies. Essentially, the customer contracts to buy or sell the foreign currency at an agreed rate (Contracted exchange rate) for settlement at a specific date in the future. The notional amount of the NDF will be stated and the reference ( or settlement) currency for these contracts is the USD. The contracted rate will be compared against the actual rate (fixing exchange rate) that prevails at the future specified period. Where rates have moved in the customer's favor, the bank from whom the NDF is bought from, or sold to, will reimburse the customer. Where rates end up being against the customer, the bank must be reimbursed instead. The settlement amount is determined as follows: Settlement Amount = ( F - C ) / F * N Where F = Fixing exchange rate, expressed in local currency per USD C = Contracted Exchange Rate at inception of the NDF, expressed in local currency per USD. And N = Notional amount of the NDF Settlement for the NDF is on a non-delivery basis; this specific feature allows the bank to take a bet with the customer about the likely expected price of the exotic currency, without having to worry about accessing the (controlled) cash market to procure the necessary relevant currency. Listed below is an example of quotes for the USD against the KRW (Korean Won). 9

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK j1o: X " 0..b "n ST ',:NZ Cii. FX: G:: tX;).. ' i.:1iv.1 I 4.4-3 ti.S' S Ni;,,.'.;C.., I:C}:FX,St:::,N:0:; 3 1M 4 141.2 It I 43,2 S: C A2T-: 1;CtN5 5;M; ~;~: 1 ~13B '1 40S:' "().3S:BC..:t.:':: S::, 00::2;' 3 4::,..8 I 4, S A.DC A FX...:.;0.:;2:-,9 9:: 43.3 4,:' T. E B:45,\;:1N:B:F.S. 02.5 0:8,0 ':142:,5 I:l,5,:C-II.X: I:06::, [4] Trading Practices and Rate Quotations Changes in net forex positions arise continuously as banks deal with their corporate clients. In order to accommodate those customers effectively without holding large inventories, banks rely on the interbank market. Indeed, interbank trading for most phenomenon can be explained by liquidity and information requirements. When a trader offers a two-way quote to counter-parties (including significant customers), it basically provides them with an option to increase or decrease his/her inventory of foreign exchange. While the dealer can influence the probabilities of being hit on the buy or sell side through pricing, it is the counterparty that will make the decision. More often than not, the dealer is 'hit' on the 'wrong' side: the customer decides to buy when the dealer wants to increase his/her inventory of foreign exchange and sells, when the dealer really wanted to decrease the inventory. Thus, to offset such undesired effects in terms of either size or direction, every 'original' transaction that is based on a sale or purchase of services, goods, or financial assets leads invariably to many subsequent transactions in the market when dealers want to rebalance their inventories. This is one of the major reasons why the volume of foreign transactions exceeds the volume of trade and investment transactions by a large multiple - an observation that is frequently misinterpreted by journalists or participants as proof for the market being a playpen for idle "speculators". Furthermore, profitable forex trading requires the acquisition of timely and relevant information. While traders are provided with expensive monitoring equipment and information services such as Bloomberg, Bridge, Dow Jones, Reuters, and others, relevant information about the state of the market and the actions of the various players can be obtained ultimately only from the continuous, active participation in the market. 10

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK It is only the access to detailed information and the traders' skill in interpreting it that allows them to take profitable positions over time. Obviously, banks attempt to control the inherent risks. They do so by limiting the size of positions that traders can take. The allocation of these limits is influenced mainly by the bank's strategic objectives, including the availability risk capital, with respect to forex trading, the currencies traded, the creditworthiness of counterparties, and each dealer's track record. The trader's intent to buy or sell is expressed through his or her quotes (although traders make sure that their intent is not too obvious in order to deprive the counterparties of useful information). For example, the senior traders in any dealing room get a feel for the market by reviewing the relevant news on their terminals. If markets are volatile, they may begin work earlier in order to contact colleagues in dealing rooms of the bank that are several time zones ahead to discuss market activity. After quick discussions with management within and contacts outside the bank, and after checking markets for gold and various other financial instruments, traders would arrive at their own quotes on the basis of the desired position in the currency (or currencies) for which they have primary responsibility. Traders usually give two numbers when quoting forex rates. For example, in a Swiss Franc (CHF) quote by a European bank of 1.506/1.507, the deal would indicate the commitment to buy one USD at CHF 1.5060 and sell one USD at CHF 1.5070, following the classic trader's principle to buy low and sell high. A quote of 1.5060/70 by a U.S. dealer signals his or her interest in buying CHF at USD 1.5070 and selling them for 1.5060. To compound the confusion, 1.5060 is still referred to as the bid rate and 1.5070 as the offer rate. Obviously, as the dealer must obtain more CHF per USD than he or she sells to avoid going broke. Unfortunately, this quotation practice causes difficulties for students at forex markets in the United States. Confusing matters further when there is an exception: GBP are quoted in "direct" terms, i.e., USD 1.42 per GBP, or USD 0.8843 per EUR. Modem telecommunications advances enable dealers to obtain rate information through the foreign currency page of one of the data transmittal services, such as Bloomberg, Reuters and Telerate. Typical screen information is as follows: 11

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK SPOT 1 Month 3 Months 6 Months 12 Months 1114 CHFa 152 1112 FRFb 607 1115 NLGC 206 40/50 39/37 107/105 202/198 290/282 10/20 113/121 346/365 675/700 1,265/1,340 10/30 38/37 105/102 199/195 281/271 1118 BEFd 38 18/38 1120 ITLe 1,295/1,297 27/25 85/82 172/168 280/270 1116 JPY 153 08/18 22/25 11/13 25/28 86/92 180/200 1110 CADf 133 1111 GBPg 154 1117 EUR 113 aSwiss franc; g U.K. pound.; hEuro. 57/56 162/160 293/289 589/590 15/25 22/32 b French francs; c Dutch guilder; d Belgian franc,.e Italian lira; f Canadian dollar; Because time is money, traders also use shorthand to quote rates. Typically, a trader will say: "Swissies are 40/50, 39/37, 107/105, 202/198, 290/282." This means that the spot rate is 1.5240/50 and the 1-month, 3-month, 6-month, and 12-month forward rates for the USD, from a Swiss bank's perspective, are CHF 1.5201 bid, 1.5213 offer; 1.5133 bid, 1.5145 offer; and these rates are obtained by subtracting the "points" 39/37, 107/105, and so forth from the bid-ask prices for spot USD. A descending order of the points means subtraction from the spot rate, or a discount on forward USD. If the USD had traded at a premium, traders would quote first the lower adjustment points, signaling that these must be added to spot bid-ask rates. In order to remember these rules, one might make use of the fact that bid-ask spreads tend to become wider for longer maturities. This fact will prove helpful in double-checking forward rate quotations. Traders do not quote the full rate on the telephone but simply the last two decimal points, e.g., "40 and 50," as their counterparts are aware of the preceding numbers at any moment in time. Suppose, in response to a CHF/USD quote of 1.5240/50, a bank sells the trader in Zurich USD 1 million. Before filling out the 12

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK purchase voucher,"l the trader says loudly to his or her colleagues, "We get USD 1 million at 40," upon which the chief trader might adjust the rates to, perhaps, 35/45 if he or she wants to avoid further purchases of USD. Quotes between traders of major banks are good for up to USD 5 million or the equivalent if it is a major currency. Otherwise, traders specify the amount. If a trader is unwilling to commit himself or herself, he or she add "for indication only." The difference between the buy and sell rates is the spread for the trader. If the trader is not interested in dealing, quotes with a very wide spread between bid and ask will be given. If the trader wishes to purchase a currency, the bid price will be raised, and likewise, if the trader is interested only in selling, the sell (or ask) price will be lowered. If the trader is interested in volume, dealing on both sides of the market, he or she would naturally quote in such a way as to show very narrow spreads between the bid and the ask price. Still, the extent to which traders will reduce their spreads is limited, because the spread is one source of trading profit. Other sources of profits are knowing the "deal flow", i.e. knowing what customers are likely to do, value dating, i.e. crediting customers' accounts later and very rarely, getting inside information about the monetary/economic policy of the central bank or the government. Unlike in share trading, obtaining such information on an exclusive basis is rare. However, in those cases where a government pursues policies that are inconsistent with economic reality, such as keeping an exchange at "unrealistic" levels, the whole trading community, banks and no-banks will try to exploit such trading opportunities, leading what is referred to in the press as "destabilizing speculation". Governments will frequently loose this battle and either let the currency find a new level, by devaluing or revaluing - or impose exchange controls. [5] Black Markets and Multiple Exchange Rates The essence of exchange controls is government proscribing the acquisition or sale of foreign money. In many countries restrictions on foreign payments are a fact of life. And while forex control systems vary a great deal, not only do they invariably affect the level of exchange rate, but they tend to bring forth parallel markets, with the result that there are two or more exchange rates for the same currency. 1 Also referred to as "ticket," this procedure is becoming more and more automated as the use of minicomputers spreads in trading rooms. 13

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Assume that a given country has a comprehensive system of exchange controls: residents are not permitted to purchase forex without explicit permission from the exchange control authorities12. Further, residents who receive foreign moneys, for whatever reason, must surrender these funds to the authorities in return for domestic money at an official rate. The law typically includes rules prohibiting unauthorized export of domestic currency and other negotiable securities.13 Without fail, black markets spring up, both inside and outside the country, where bank notes and other negotiable instruments are exchanged for foreign currencies. In addition, the rate at which domestic currency is bought and sold is higher than the rate at which the authorities make foreign currency available; more of the domestic currency must be delivered for each unit of the foreign currency. The premium of the black market rate over the official rate depends on a combination of three factors: (1) the allocation (exception) policy of the authorities; (2) the level of the official rate relative to demand; and (3) the penalties and efficiency of enforcement policies. Apart from this differential over the official rate, there is also a substantial spread between buy and sell rates, reflecting the risk of making illegal exchanges within the country and/or the danger involved in bringing domestic bank notes outside or inside the country in contravention to laws and regulations.14 Because of the high administrative cost of comprehensive exchange controls, where literally every international transaction has to be approved, less ambitious countries have resorted to an approach based on segregating international transactions, forcing each through a distinct segment of the forex market. A good example used to be the United Kingdom. For many years prior to 1978, U.K. investors were allowed to purchase securities outside of the sterling area (the group of countries pegged to sterling) only with forex funds that became available through the sale of foreign securities by other U.K. residents. The demand for such funds regularly exceeded the supply at the prevailing exchange rate, particularly since a 12 When classifying exchange control, one distinguishes rules that restrict resident, nonresidents, or both. Accordingly, currencies are fully convertible or restricted for residents, non-residents, or for all. 13 Such laws are often supplemented by rules that limit the importation of domestic currency to catch funds that have found their way abroad illegally. A good survey of exchange controls in various countries can be found in the annual report "Exchange Restrictions" published by the International Monetary Fund. 14 The extreme spreads for some countries' bank notes are caused not so much by the low volume but rather by the existence of exchange controls that makes arbitrage risky and expensive. 14

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK portion of the foreign currency proceeds was required to be surrendered to the Bank of England by the seller in return for domestic sterling funds. The resulting difference between the official exchange rate for current and investment transactions and the rate applicable to foreign investments became known as the London investment USD premium, expressed as a percentage over the official spot rate. Over the years, this premium had fluctuated widely, largely in response to expectations regarding the relative merits of investments outside of the United Kingdom versus those in the U.K. securities market. Many other countries have used similar devices. Belgium, France, and Italy, for example, have at times split the forex market into two sections: one for current account transactions, and one for capital, or financial transactions. The central banks intervened primarily in the market for current account transactions, while the financial rates have typically fluctuated without intervention. At times, South American countries have pushed such schemes to extremes, with rates not only for current and financial transactions but also for various exports and imports, depending on the particular category. From a forex market perspective, trading in these currencies becomes very difficult, with the result that markets become thin and the spreads between bid and ask become very wide, which is only one of the costs associated with such policies. (Footnote: The other cost is that no derivatives and other hedging instruments - with the possible exception of NDF's - will be available for exporters, importers and investors) [6] Theory: Interest Rates, Exchange Rates, and Prices and Output The equivalence of exchange rate changes and interest rates was discussed previously in the context of hedging a CAD payable arising from an export transaction. Note that this principle applies to all financial assets and claims, not just to payables, which include all kinds of securities with contractually set returns (i.e., not equities). This interest rate parity is what is known as an after-the-fact or ex post relationship, where currently available interest rates for equivalent instruments denominated in two different currencies and the difference between spot and forward rates are examined. Since a sufficient difference leads to (riskless) opportunity, it is the basis for foreign exchange swaps and related arbitrage transactions. Another way of expressing this relationship between exchange rates and interest rates is to specify what is actually exchanged in forex and securities markets. In the forex markets, current and future units of foreign money are traded for domestic money; 15

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK similarly, on the domestic (foreign) securities market, current units of domestic (foreign) money are exchanged for cash in the future. The schematic presentation in Figure 1 shows the markets and the relevant rates.15 Knowing how these markets are interrelated however, does not help in forecasting what rates will be in the future, i.e., the beforethe-fact or ex ante relationship. Figure 1 Interrlationship Between Forex and International Credit Markets USD to (D 0 ra C/ (D cE 0 rvL USD credit market USD rate Interest Rate differential CAD rate CAD credit market 0 ACD USD tn 0 CAD to CAD tn Traditional market theory explains the determination of prices (rates) of financial assets by supply and demand. This can easily be reconciled with the modem view of markets by stressing the role of expectations. What really matter are the expectations of market participants about future demand and supply and all of their determinants. It follows that in markets where prices are established in such a manner, current prices incorporate the expectations of market participants about future prices, such as exchange rates and interest rates. What Figure 1 shows is that interest rates and 15 This view has been suggested by Deardorff (1979) 16

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK exchange rates are simultaneously determined by the same set of expectations. But, if one assumes that market participants act rationally, expectations must be formed on some basis. Which factors determine the relevant expectations? The determination of the level of interest rates in an economy is a complex process, and a number of competing explanations describe the processes by which interest rates are determined.16 In an international context, the primary concern is, of course, with relative changes in interest rates. The monetarist framework focuses on the excess supply of money. Differences in the rates of growth of excess money and changes in money demand among countries cause (relative) price-level changes. This relationship is shown, in a one-country context, in Link 1 of Figure 2. More Figure 2 International Equilibrium Framework 5 16 This section relies on a synthesis by Giddy (1976). Theories of interest rate determination are associated with the works of Wicksell, Keynes, Hicks, and Fisher. Any good introductory book on money and banking will bring the reader up to date. 17

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK profoundly, the rate of change in the general price level (inflation) in a country is determined by the difference between the rate of growth in real output and the growth in the effective supply of money. The underlying idea is that when the rate of change in the money supply differs from the need for additional liquidity resulting from changes in real output, the general level of prices in the economy will change. The demand for money changes because of changes in the level of output of goods; if more goods are produced; more money is required to support the greater volume of transactions at the existing price levels. If the money supply changes by an amount larger than the increase in output of goods, it can only be absorbed by an increase in prices such that the new output at new, higher price requires the new, greater amount of money supply. As a consequence, an excessive increase in the money supply causes prices to rise. Of course, a (temporary) increase in money balances by individuals, businesses, and government may delay this inflationary effect from occurring right away. The second link identifies the relationship between changes in interest rates and changes in price levels. In a theory generally attributed to Fisher,17 interest rates have two components: a real return plus an adjustment for price-level changes. If the holders of loanable funds are to be induced to give up their money for a period, they must receive compensation. This compensation is represented by the higher purchasing power of the funds in the period in which they are returned and is necessary because, without it, the holders of funds have no incentive to forgo current consumption and, in addition, to be exposed to the risk of default. Thus, the compensation is termed the real return to the lenders of money. If prices do not change, this return would represent the increase in the purchasing power of the loaned funds. But when prices are expected to rise are aware lenders will demand compensation for inflation in order to protect the real rate of return. This demand will be met by including a factor for the expected rate of inflation in the interest rate. Thus, the interest rate will include a real return that increases the purchasing power of the loaned funds plus a return that offsets the loss of purchasing power due to inflation. Link 2, between price-level changes and interest rate, implies that as the price level is expected to change (e.g., because of excessive money supply), the interest rate will change, so that changes in the nominal interest rate completely offset changes in prices, keeping the real rate of interest constant. The change in interest rates will thus 17 Fisher (1980) 18

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK be equal to the percentage change in expected price levels, which equals the rate by which the increase in the money supply exceeds the real needs of the economy. It should be clear by now that if by the end of the period, prices have changed more than was expected, the rate of interest would have been underestimated and borrowers would have gained at the expense of lenders. Similarly, lenders would have gained if the price changes had been overestimated. Given the process by which interest rate changes are determined in a national economy, how do differentials in interest rates between different countries come about? Two additional relationships, shown in Figure 2 as links 3 and 4, contribute to the explanation. Link 3 relates price-level changes in different countries to exchange rate changes; it is usually referred to as the purchasing power parity (PPP) theorem. The other link, which relates differences in interest rates, is sometimes referred to as the International Fisher Effect to distinguish it from the relationship of price-level changes and interest rates in a closed economy. The PPP theorem can be stated in different ways, but the most common representation links the changes in exchange rates to those in relative prices.'8 The relationship is derived from the basic idea that in the absence of trade restrictions, changes in the exchange rate mirror changes in the relative price levels in the two countries. At the same time, under conditions of free trade, prices of similar commodities cannot differ between two countries, because arbitragers will take advantage of such situations until price differences are eliminated. This is known as the law of one price. (That is, the rate of change in the exchange rate equals the difference in inflation rates.) Thus, link 3 in Figure 2 relates expected changes in exchange rates to the expected differences in inflation rates. Of course, it must be recognized that this statement, like all simple relationships, is based on some strong assumptions (e.g., that consumers in all countries have the same consumption basket and that all goods are equally tradable, so that relative price changes are not possible within one country). If these conditions are not met, the postulated relationship between the respective inflation rates and the exchange rate is approximate at best, even after accounting for problems of measurement. The relationship between the interest rates in two countries and the expected exchange rate changes has already been discussed in general terms. Now the linkage between these rates can be treated in a more formal manner. In the absence of effective controls on capital flows, risk-neutral investors will employ their funds wherever the 18 For a comprehensive review of PPP, see Officer (1982) 19

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK (expected) return is the highest. Thus, if the interest rates between two countries are unequal, investors transfer their funds to the country with the higher interest rate. Will the higher interest rate not come down under the influence of these flows? According to the International Fisher Effect, the interest rate differential will exist only if the exchange rate is expected to change in such a way that the advantage of the higher interest rate is offset by the loss on the forex transactions. An investor in the lowinterest country can convert its funds into the currency of the high-interest country and expect to earn a higher rate of interest, but its gain (in terms of higher interest rate) will be offset by the expected loss because of forex rate changes. This relationship is based on expectations only. However, it is possible to replace the expected future spot exchange rate with the forward rate, given certain specific assumptions. This forward rate, like interest rates, is a contractual rate; this leads to the interest rate parity theorem. According to this theorem, the observed differences in the interest rates will be equal to the premium or the discount of the forward exchange rate over the spot rate. This relationship is represented by link 5 in Figure 2, and its mechanics have been discussed in the context of the operation of the forex market. The PPP theorem is based upon the assumption of perfect arbitrage in markets for goods and services. Goods markets are, however, far from perfect; costless and instantaneous arbitrage is usually not possible. Therefore, it is often argued that because of imperfections in the goods markets, PPP is at best a long-run tendency, and any test of the theorem should examine the behavior of deviation from PPP over an extended period. If there is a direct link between national price levels and exchange rates, deviations from PPP, over many periods, should not be significantly different from zero. Finally, there could be an indirect link between inflation and the exchange rate, owing to free movement of capital. A test of this proposition reveals that deviations from PPP have been constant under both fixed and flexible exchange rate regimes.19 Thus, there is some support for the notion that higher inflation rates would eventually lead to a change in the exchange rate, whereby the cause may not be arbitrage in goods markets but rather arbitrage by way of national money markets, which is less costly. 19 Adler and Lehmann (1983). However, even when done carefully, the measurement problems underlying such tests are formidable. One vexing issue is the change in the composition of the "relevant" index as the basket of goods and services changes not only over time due to changes in technology and tastes, but changes in response to relative price changes of different goods which cause users to substitute goods and services that have become less expensive for those that have increased in price. 20

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK The implication is that deviations from PPP may not correct themselves over the long run, which substantially diminishes the usefulness of the PPP concept in exchange rate forecasting. In summary, interest rate differences and exchange rate changes are linked in two ways: (1) through the international Fisher effect, which is based on expectations and (2) through the interest rate parity theorem, which reflects an actual, arbitraged relationship in the market. The integration of the ideas represented by the five links of Figure F2-2 can be summarized as follows: 1. Monetary theory links different rates of excess money supply and changes in money demand with changes in price levels. 2. Changes in price levels and changes in interest rates are linked by the Fisher effect. 3. Changes in price levels and changes in forex rates are linked by the PPP theorem. 4. Changes in interest rates and forex rates are linked ex ante by the international Fisher effect. 5. Changes in interest rates and forex rates are linked ex post by the forward rate by means of the interest rate parity theorem. Anyone familiar with economic theory will recognize that the precise nature of each of these relationships continues to be the subject of considerable controversy. Questions can be raised about both the assumptions and the empirical verification of these relationships. If these relationships hold at all, they do so only under the assumption that the markets for goods, capital, and currencies reasonably meet the requirements for perfect markets, especially with respect to governmental regulations and restrictions. However, at this point the concern is not so much with empirical validation as with presenting a simple, consistent, and comprehensive set of relationships that can serve as a point of departure for further analysis. [7] Forecasting Exchange Rates and Managerial Implications Observers of markets with competing private, profit-seeking traders have long recognized that prices are determined by the anticipations of demand and supply. The pursuit of profits in speculative markets requires that market participants take future prices into account in their current buying and selling decisions. They form opinions about future prices by continually searching for information and interpreting it in order to make predictions. Those that perform this task well grow and dominate the market; losers are eliminated. This concept of market efficiency is based on the notion that current prices reflect all available information, including market participants' 21

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK expectations about future prices and, further, that all new information that is received by the market is analyzed and immediately incorporated into expectations about future prices. From these expectations, decisions to buy and sell are made, and thus the expected prices are converted into current prices. These considerations apply to all organized asset markets, which include markets for spot and forward forex. The current spot rate reflects anticipated supply and demand conditions, and the forward rate contains information about the future spot rate. This is because as a contractual price, the forward rate offers opportunities for speculative profits for those that correctly assess the future spot price relative to the current forward rate. Because expectations of future spot rates are formed on the basis of currently available information (historical data) and an interpretation of its implications for the future, they tend to be subject to frequent and rapid revision. As previously noted, the actual future spot rate may therefore deviate markedly from the expectation embodied in the present forward rate for that maturity. These deviations are in the nature of a forecast error that is known only after the fact, i.e., when the future has become the present. Formally, the forecast error is expressed as Ft,n - S t + n = et,n where Ftn represents the forward rates in period t for maturity n, St + n represents the actual spot rates at period t + n, and e reflects the difference. Another view considers these errors to be speculative profits or losses. Can they be consistently positive or negative? It is reasonable to assume that this is not the case. Otherwise, one would have to explain why consistent losers do not quit the market or why consistent winners are not imitated by others or do not increase their volume of activity, thus causing adjustment of the forward rate in the direction of their expectation. Barring such an explanation, one would expect that the forecast error is sometimes positive, sometimes negative, alternating in a random fashion, and driven by unexpected events in the economic and political environment. Over sufficiently long periods, allowing a large number of decisions, et, should average out to zero. This is the statistical basis for the view that the forward rate is an unbiased predictor of the future spot rate.20 20 If the series of errors has a mean value of zero, it represents a martingale process. If the series etn is purely random, i.e., not correlated with any past or future value, it is referred to as white noise, which is the formal requirement for an efficient market. Tests of market efficiency on this basis involve testing two propositions jointly: (1) They test the underlying model that generates forex rates and (2) they test the proposition that participants indeed set the forward rate equal to the expected future spot rate. Thus, conclusive empirical verification of the existence of market efficiency is virtually impossible to achieve. 22

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK However, this definition of efficiency leaves out sources of bias that may be known to market participants and may be incorporated, quite rationally, into the forward rate. Thus, there may be systematic deviations of the forward rate from the expected future spot rate that are compatible with market efficiency. The latter concept refers only to the processing of information, yet the relevant information may well contain factors that cause systematic deviations of forward rates from expected future spot rates. Forecasting errors can therefore be classified as random or systematic, leading to unbiased or biased forecasts. The latter, in turn, may be based on inefficient pricing processes or may be perfectly consistent with an efficient market. Two significant causes for bias can be identified: political risk and exchange risk.21 [8] Political Risk The incidence of political interference in the market process by government is a different matter altogether. If, for example, the U.K. government was prone to interfere with the rights of nonresidents' GBP assets, one might expect that those that purchase GBP for forward delivery would demand a premium: more GBP for every USD, other things being equal. By the same token, one would expect U.K. interest rates to be systematically higher than warranted on the basis of expected exchange rate changes alone. To illustrate this point in terms of interest rates: Given a spot rate of USD 2 per GBP, the 12-month forward is USD 1.92, while the expected future spot rate is assumed to be USD 1.94 under these conditions. Of course, the latter cannot be observed directly; the only evidence would be an "undershooting" of the forward rate by USD 0.02, or, in terms of interest rates, an interest differential of 100 basis points per annum on average, provided that the political risk premium is constant over times and that participants do not make systematic errors in judging the future exchange rate of the pound. While this example might clarify conceptually the effect of political risk on forward rates and interest differentials, it is much more difficult to estimate the actual 21 A small source of bias or systematic differences is a technicality known in the literature as the Siegel paradox, which is based on a mathematical concept called Jensen's inequality. If residents of two countries consume a common basket of goods, the forex rate generally lies between the expectation of the future rate defined in terms of one currency and the expectation of the future rate defined in terms of the other currency. However, the practical significance of the Siegel paradox is negligible when the (expected) variance of the exchange rate is not excessive, and it disappears altogether when PPP holds. For a thorough analysis of this phenomenon, see Beenstock (1985). 23

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK premium. Even though this may be an intractable task, a modest contribution can be made by attempting to clarify conceptually the components that comprise that catchall concept political risk. To begin with, the term political risk is quite broad and cannot readily be made operational. The focus should probably be on current and potential effects of the political environment on investors' transactions. More precisely, one should be concerned with the probability that changes in the political environment will reduce returns to the point where investments would no longer be acceptable on the basis of predetermined criteria.22 Obviously, exchange controls (i.e., quantitative restrictions on international flows of financial assets) that affect existing positions of nonresidents would fall under this category. What is less often recognized is that quantitative restrictions in domestic credit markets have the same effect: They add an additional constraining element to arbitrage. Government actions that affect exchange rates or interest rates through markets, such as open-market operations by the central bank, are generally held to contribute to exchange risk, which can be taken fully into account by market participants in setting forward premiums and interest rate differentials, respectively. But the differentiation between quantitative restrictions on the one hand and price fluctuations on the other is not so clear. Further, exchange rate fluctuations may themselves give rise to exchange controls if the body politic considers them excessive. In any case, market efficiency presupposes that prices are arrived at by the actions of competitive and profit-seeking market participants. If a central bank is not motivated by profitability and if its actions are so massive that it succeeds in shifting prices from equilibrium points, the forward rate or interest differential will not be an unbiased predictor, nor will the market be efficient. A further point of distinction is necessary. Risk refers to uncertainty, here, the uncertainty of future capital controls. It is therefore necessary to distinguish between interest differentials or premiums in the forward market that are caused by existing controls, which may place a tax-like burden on assets in a certain jurisdiction, and differentials that are due to expected, hence future, capital controls.23 [9] Exchange Risk 22 Kobrin (1979). 24

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK The analytical framework developed earlier suggests that exchange rates should not be affected by anything other than interest rates and inflation rates; thus, there should not be exchange risk premia. However, that framework relies heavily on several very strong assumptions about economic behavior that may not hold in the real world. One of the simplifying assumptions is that real interest rates are stable, so an unexpected change in monetary policy, for example, will affect only nominal rates. This assumes that prices of goods and services will adjust instantaneously to changes in financial markets. This is obviously not the case. Prices for goods and services tend to be less flexible downward than prices of financial assets because of the existence of contractual relationships and, more prominently, the prevalence of market imperfections (e.g., unions, large corporations with market power, and technological advantages). As a result, there will be transitory changes in real interest rates that may persist for some time when the change in monetary policy is not expected to last. To illustrate, an unexpected increase (change from the expected trend) in the money supply leads to a drop in real interest rates. Prices (or price level) may increase or there may be a lag in the increase, depending on the elasticity of the prices. Output increases, but there is also a lag effect; output takes time to adjust to new levels. Exchange rates fall. Likewise, an unexpected decrease in the money supply leads to an increase in real interest rates. Prices may fall, but the magnitude and immediacy of the fall depends on the degree of price rigidity. Output also decreases, but there is a lag effect as well. Here, exchange rates rise. Another assumption refers to the perfect substitutability of goods within a country and across borders, in effect, that all goods are homogeneous and there are no non-tradable goods. This implies that both the law of one price and PPP hold. However, since PPP refers to the average price (i.e., price levels), it is possible, when arbitrage in goods' markets is imperfect, that PPP holds but the exchange rate changes. This would be the consequence of relative price changes in each country and the resultant impact on the commodity composition of trade. Conversely, it is possible for the exchange rate to remain unchanged while the price-level changes differ at home and abroad.24 A final assumption in the equilibrium framework is the perfect substitutability of domestic and foreign financial assets. If domestic and foreign bonds 23 This point has been clearly made in Dooley and Isard (1980). 24 The conditions for the law of one price are therefore more stringent than those for PPP. It is frequently interpreted that the latter presumes that changes in price levels lead to changes in exchange rates, while the former allegedly implies that causation proceeds in the opposite direction. Both are equilibrium conditions, i.e., prices and exchange rates are simultaneously determined. 25

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK are perfect substitutes, they must have the same expected real return; hence, differences in nominal interest rates can only mirror the expected exchange rate change. The results of an equilibrium framework must be viewed as approximate, long-run tendencies only, since PPP may not hold for extended periods, the substitutability of assets may not be perfect, and the market for financial assets and real goods may not clear instantaneously. When accounting for non-money assets, any investor in fixed-income investments faces a purchasing power risk because of the uncertainty of future inflation. As long as inflation rates are not perfectly positively correlated among countries, it pays to diversify portfolios internationally. In portfolio balance, therefore, it is theoretically possible that real return expectations differ between domestic and foreign securities because of residual purchasing power risk.25 Put somewhat differently, systematic deviations from the international Fisher relationship are possible when a country has issued an excessive amount of government liabilities26 relative to its share in the minimum variance, internationally diversified portfolio.27 Accordingly, a premium above the expected depreciation could be explained as compensation for investors to increase their holdings of foreign currency assets, of which excessive amounts have been issued.28 The empirical evidence on the presence of risk premia is voluminous and the results are mixed. Earlier work by Solnik (1974), Grauer et al (1976), Stulz (1981), and Adler and Dumas (1983) show that international systematic risk can be a significant factor in asset pricing. The literature on currency risk at the aggregate level, for example Ferson and Harvey (1993, 1994), suggest that (historical) currency risk seems to be priced by the markets. In addition, Harvey (1991), Dumas and Solnik (1995) and De Santis and Gerard (1998) document that currency risk premia are time-varying for developed countries such the United States, UK, Japan, and Germany. Further 25 There should be no residual exchange risk because there is a perfect negative correlation for deviation from PPP for any two currencies, guaranteeing, in principle, complete diversifiability. 26 Government rather than private debt is necessary for a risk premium, for with private debt gains to borrowers will be losses to lenders, and the perfect negative correlation of such returns permits the complete elimination of this risk through an appropriate diversification strategy. 27 See, e.g., Dornbusch (1980) 28 The foregoing considerations obscure the issue of diversifiable versus nondiversifiable risk in the broader context of an international capital asset pricing model. It would, for example, be risky to hold securities denominated in a currency whose value is positively correlated with that of other assets; the securities are afflicted with a risk that cannot be diversified away, and a risk premium would be required. But difficult questions arise as to the definition of "other assets" and the relevant consumption patterns of international investors. 26

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK supporting evidence that currency risk is priced at the aggregate market levels are provided by Doukas et. al. (1999), Tai (1999) and Vassalou(2000). These pieces of empirical evidence seem to point towards the puzzle of the forward rate premium, i.e. the spread between the forward rate and the currently observed spot exchange rate. This premium/discount is often inversely correlated with subsequent changes in the spot rate. However, as Roll and Yan (2000) observe in a recent paper, "This (existence of a positive/negative premium) defies economic intuition and possibly violates market efficiency." It is not that various explanation have not been attempted: the explanations range from statistical phenomenon to economic justifications such as a country's degree of indebtedness ( Fukao, 1981), but none are convincing. Many researchers indeed have accepted the puzzle as documenting inefficient forex market and even justify the existence of profitable opportunities. Roll and Yan (2000) suggest there is really no puzzle at all. A simple model adequately fits the data and they conclude that the forward exchange rates are unbiased predictor of subsequent spot rates, confirming previous findings (Dufey and Giddy, 1978) using much less sophisticated methods. Roll and Yan show convincingly that "the puzzle has arisen because (a) the forward rate, the spot rate, and the forward premium all follow non-stationary time processes, and (b) the forward rate is (not a biased but simply) a noisy predictor." The bottom line for management is that if such biases exist (in markets unaffected by exchange controls), they are not exploitable in a meaningful way. They are not recognizable before the fact, they are unstable, and they tend to be dominated by transaction costs and risks. This insight has important implications for corporate hedging and financing policies in an international context. However, before attempting to deal with risk one must know what is exposed to exchange risk, and, in the context of a non-financial firm, this is not a trivial matter. III ASSESSING CORPORATE FOREIGN EXCHANGE RISK Developing foreign exchange risk management policies at the level of a business firm must begin with three important considerations. First, not all foreign exchange rate change are relevant. Since forex markets price expected changes, in forward rates and interest rate differences, from a risk management perspective, it is the (unexpected) deviations from the expected rate that matter. Secondly, and more 27

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK important is the condition that there must be a causal effect of the unexpected exchange rate change on the profitability and therefore the value of the firm. For example, a firm that sells it output in international markets (on a cash basis) for foreign currencies (on a cash basis), will not be affected by unexpected exchange rate changes if it can quickly adjust prices if it has market power. Third, the question must be addressed whether foreign exchange risk management should be properly taken care of at the level of the individual firm, or better be left to the investors. The task of gauging the effect of forex rate changes on an enterprise begins with measuring its exposure to risk, i.e., the amount, or value at risk. When a firm has primarily assets and liabilities with "contractually fixed" returns, such as receivables, loans, bonds and similar instruments, exposure analysis is relatively straight forward, complicated only by explicit or embedded option features. With regard to non-financial business firms the problem is much mor complex, as the core assets of such firms, such as inventories, plant and equipment do not have contractually fixed returns. The expected returns may well be significantly affected by exchange rate changes and therefore affect the value of the enterprise. This issue is clouded further by the fact that financial results for an enterprise tend to be compiled by methods based on the principles of accrual accounting. This means that the data provided are not those that are really relevant for business decision making, namely future cash flows and their associated risk profiles. As a result, considerable efforts are expended, both by managers and students of exchange risk, to reconcile the differences between the effects of exchange rate changes on an enterprise in terms of accounting data, referred to as accounting or translation exposure, and the 29 cash flow effects, referred to as economic (or sometimes transaction-) exposure. [1] Accounting Exposure The concept of accounting exposure arises from the need to translate accounts denominated in foreign currencies into the home currency of the reporting entity. (See Chapter F5 for more details.) The problem is most common when an enterprise has foreign affiliates that keep their books in the local currencies. For purposes of consolidation, these accounts must somehow be translated into the reporting currency of the parent company. Here, a decision must be made as to the relevant exchange rate to be used for the translation. While income statements of foreign affiliates are typically translated at a periodic average rate, balance sheets pose a more serious challenge. 29 Many authors have referred to this conflict. For a well-focused report, see Showers (1988) 28

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK The underlying difficulties are revealed by the struggle of the accounting profession to agree on appropriate translation rules and the treatment of the resulting gains and losses. A comparative historical analysis of translation rules may best illustrate the issues at hand. Over time, U.S. companies have followed essentially four types of translation methods, summarized in Figure 3. These four methods differ with respect to the presumed impact of exchange rate changes on the value of individual categories of assets and liabilities. Accordingly, each method can be identified by its manner of separating assets and liabilities into those that are exposed and, therefore, translated at the current rate (i.e., the rate prevailing on the date of the balance sheet), and those whose value is deemed to remain unchanged and that are therefore translated at the historical rate. Figure 3 Methods of Translation for Balance Sheets Current/ Monetary/ T Current/ Noncurrent Nonmonetary Current Assets Cash C C C C Marketable securities (at market value) Accounts receivable C C C C Inventory (at cost) C H H C Fixed assets H H H C Liabilities Current Liabilities C C C C Long-term debt H C C C Residual Residual Residual Residual ~Equity 'adjustment adjustment adjustment adjustment Note: In the case of income statements, sales, revenues and interest are generally translated at the average historical exchange rate that prevailed during the period; depreciation is translated at the appropriate historical exchange rate. Some of the general and administrative expenses as well as the cost of goods sold are translated at historical exchange rates; others are translated at current rates. C: Assets and liabilities are translated at the current rate, or rate prevailing on the date of the balance sheet. H: Assets and liabilities are translated at the historical rate. 29

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK [2] Alternative Accounting Methods The current/noncurrent method of translation divides assets and liabilities into current and noncurrent categories, using maturity as the distinguishing criterion. Only current items are presumed to change in value when the local currency appreciates or depreciates vis-a-vis the home currency. Supporting this method is the economic rationale that forex rates are essentially fixed but subject to occasional adjustments upward or downward that tend to reverse in time. This assumption reflected reality to some extent, in the industrialized countries during the fixed-rate system. However, with subsequent changes in the international financial environment, it has become outmoded, although this translation method is still being used in some European countries. Under the monetary/non-monetary method, all items explicitly defined in terms of monetary units are translated at the current exchange rate, regardless of their maturity. Non-monetary items in the balance sheet, such as tangible assets, are translated at the historical exchange rate. The underlying assumption is that the value of such assets increases or decreases in local currency terms immediately after a devaluation or revaluation to a degree that compensates fully for the exchange rate change. A similar but more sophisticated translation approach supports the so-called temporal method. Here, the exchange rate used to translate balance sheet items depends on the valuation method used for a particular item on the balance sheet. Thus, if an item is carried on the balance sheet of the affiliate at its current value, it is translated using the current exchange rate. Alternatively, items carried at historical cost are translated at the historical exchange rate. This method synchronizes the time dimension of valuation with the method of translation. As long as foreign affiliates compile balance sheets under traditional historical cost principles, the temporal method gives essentially the same results as the monetary/non-monetary method. However, when current value accounting is used, i.e., when accounts are adjusted for inflation, the temporal method calls for the use of the current exchange rate throughout the balance sheet. The temporal method provided the conceptual base for the Financial Accounting Standard Board's (FASB) Statement of Financial Accounting Standards (SFAS) No. 8, which came into effect in 1976 for all U.S.-based companies as well as for non-U.S. companies that had to adopt U.S. accounting principles if they intended to raise funds in the public markets of the United States. The temporal method raises a more general issue: the relationship between translation and valuation methods for accounting purposes. When methods of valuation provide results that do not reflect economic reality, translation fails to remedy that 30

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK deficiency but tends to make the distortion very apparent. To illustrate this point, companies with real estate holdings abroad financed by local currency mortgages found that under SFAS No. 8, their earning were subject to considerable translation losses and gains. This came about because the value of their assets remained constant, as they were carried on the books at historical cost and translated at historical exchange rates, while the value of their local currency liabilities increased or decreased with every twitch of the exchange rate between reporting dates. In contrast, U.S. companies whose foreign affiliates produced internationally traded goods (e.g., minerals) felt comfortable valuing all of their tangible assets on a dollar basis. These companies were the ones that did not like the transition of the current/current method where all assets and liabilities are translated at the exchange rate prevailing on the reporting date. They felt (rightly) that the value of all assets denominated in the local currency of the given foreign affiliate changes in direct proportion to the exchange rate change was an assumption that did not reflect the economic realities of their businesses. [3] Statement of Financial Accounting Standards No. 52 In order to accommodate the conflicting requirements of companies in different situations while maintaining a semblance of conformity and comparability, at the end of 1981 the FASB issued SFAS No. 52, replacing SFAS No. 8. SFAS No. 52 uses the current/current method as the basic translation rule. At the same time, it mitigates the consequences by allowing companies to move translation losses directly to a special subaccount in the net worth section of the balance sheet, instead of moving them to current income. This provision may be viewed as a mere gimmick without much substance, providing at best a signaling function indicating to users of accounting information that translation gains and losses are of a different nature than that of items normally found in income statements. A more significant innovation of SFAS No. 52 is the functional currency concept, which gives a company the opportunity to identify the primary economic environment of its foreign affiliate and to select the appropriate (functional) currency for each of the corporation's foreign entities. This approach reflects the recognition by the accounting profession that the location of an entity does not necessarily determine the relevant currency for a particular business. Thus, SFAS No. 52 represents an attempt to take into account the fact that exchange rate changes affect different companies in different ways and that rigid, general rules treating different circumstances in the same manner provides misleading information. 31

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Figure 4 Application of SFAS No. 52 In applying SFAS No. 52, a company and its accountants must make two decisions in sequence. First, they must determine the functional currency of the entity whose accounts are to be consolidated. For all practical purposes, the choice is between local currency and the USD. In essence, there are a number of specific criteria that provides guidelines for this determination. As usual, extreme cases are relatively easily classified: A foreign affiliate engaged in retailing local goods and services has the local currency as its functional currency, while a "border plant" that receives the majority of its inputs from abroad and ships the bulk of the output outside of the host country has the USD as its functional currency. If the functional currency is the USD, foreign currency items on its balance sheet have to be restated into dollars and any gains and losses are moved through the income statement, just as under SFAS No. 8. If, on the other hand, the functional currency is determined to be the local currency, a second issue arises: whether the entity operates in a high-inflation environment. "High-inflation countries" is defined as those whose cumulative three-year inflation rate exceeds 100 percent. In that case, essentially th same principles as in SFAS No. 8 are followed. In the case where the cumulative inflation rate falls short of 100 percent, the foreign affiliate's books are to be translated using the current exchange rate for all items and any gains or losses are to go directly as a charge or credit to the equity accounts. SFAS No. 52 has a number of other fairly complex provisions regarding the treatment of hedge contracts, the definition of transactional gains and losses, and the accounting for intercompany transactions. In essence, SFAS No. 52 allows management much more flexibility to present the impact of exchange rate variations in accordance with perceived economic reality; by the same token, it provides greater scope for manipulation of reported earnings and reduces the comparability of financial data for different firms. In order to adjust to the diversity of real life, SFAS No. 52 became quite complex. Figure 4 provides a brief guide to the inherent logic of the standard.30 Even with the increased flexibility of SFAS No. 52, users of accounting information must be aware that there are three systematic sources of error that tend to misled those responsible for exchange risk management: 1. Accounting data do not capture all commitments of the firm that give rise to exchange risk. 30 Those interested in accounting aspects will find a number of useful publications by the large accounting firms. See, e.g., Price Waterhouse (1981). 32

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK 2. Because of the historical cost principle, accounting values of assets and liabilities do not reflect the contribution to total expected net cash flow of the firm. 3. Translation rules do not distinguish between expected and unexpected exchange rate changes. Regarding the first point, it must be recognized that commitments entered into by the firm in terms of foreign currency, e.g., a contract involving the purchase or a sale of goods, are not normally booked until the merchandise has been shipped. At best, such obligations are shown as contingent liabilities if they are judged to be material. More important, accounting data reveal very little about the ability of the firm to change costs, prices, and markets quickly. Alternatively, the firm may be limited in its actions by strategic commitments, such as investment in plant and facilities. Such commitments are important criteria in determining the existence and magnitude of exchange risk as well as its time dimension. The second point surfaced in the previous discussion of the temporal method: If asset values differ from market values, translation, however sophisticated, will not redress this shortcoming. Thus, many of the perceived problems of SFAS No. 8 had their roots not so much in translation but in the fact that in an environment of inflation and exchange rate changes, the historical cost accounting frustrates the best translation efforts. Finally, translation rules do not take into account the fact that exchange rate changes have two components: (1) expected changes that are already reflected in the prices of assets and the costs of liabilities (relative interest rates) and (2) the unexpected deviations from the expected change that constitute the true sources of currency risk. The significance of this distinction is clear: Managers have already taken account of expected changes; risk management efforts must focus on the effect of unexpected changes, as expected changes are reflected in prevailing prices and costs. The accounting profession has attempted to catch up with financial technology and one of the results is FAS 133: Accounting for Derivative Instruments and Hedging Activities. This standard requires that an entity recognize all derivatives as either assets or liabilities in the statement of financial position and measure those instruments at fair value. Also, the firm is required to establish at the inception of the hedge, the method it will use for assessing the effectiveness of the hedging derivative and the measurement approach for determining the ineffective aspect of the hedge. However, the new accounting standards do not solve the underlying issues: values of the assets to be hedged are based on accrual accounting and not market-based. 33

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK While there are legitimate reasons for these standards, for example auditing, the basic problem remains, business decisions are made on the basis of market prices that in turn are based on expected future cash flow. These requirements significantly impair the value of accounting data for hedging decisions. [4] Economic or Cash Flow Exposure The concept of economic exposure derives directly from the nature of the (non-financial) firm. Management in its entrepreneurial function perceives opportunities to generate profits, i.e., expected positive net cash flows. Those net cash flows, however are frequently subject to unexpected exchange rate changes. Since the firm claims its special expertise in the uncovering of market opportunities for real goods and services, the basic objective of risk management is to shield its net cash flows, and thus the value of the enterprise, from unanticipated exchange rate changes.31 This sketch of economic exchange risk has a number of significant implications, some of which seem to be at variance with ideas frequently found in the literature and apparent business practices. Specifically, there are implications regarding (1) the question of whether exchange risk originates from monetary or nonmonetary transactions; (2) a reevaluation of traditional perspectives such as transaction risk; and (3) the role of forecasting exchange rates in the context of corporate forex risk management. [5] Contractual Versus Noncontractual Returns Returns on contractual assets and liabilities of a firm are fixed in nominal, monetary terms. Such returns (e.g., earnings from fixed-income securities and receivables and the negative returns on various liabilities) are relatively easy to analyze with respect to exchange rate changes; when they are denominated in terms of foreign currency, their terminal value changes directly in proportion to the exchange rate change. Thus, with respect to financial items, the firm is concerned only about net assets or liabilities denominated in foreign currency, to the extent that maturities (or duration, to be precise) are matched. For assets with noncontractual returns, it is much more difficult to gauge the impact of an exchange rate change. While conventional discussions of exchange risk 31 For specific suggestions on distinguishing between the effects of unanticipated exchange rate changes on the firm and other related events, see Hekman (1986). 34

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK focus almost exclusively on financial assets, for trading and manufacturing firms at least, it is equipment, real estate, buildings, and inventories that make the decisive contribution to the total cash flow of those firms. And returns on such assets are affected in quite complex ways by changes in exchange rates. The most essential consideration is the reaction of the firm's prices and costs to an unexpected exchange rate change. For example, if prices and costs react immediately and fully to offset exchange rate changes, the firm's cash flows are not exposed to exchange risk, since they will not be affected in terms of the base currency. Thus, under those specific circumstances the value of noncontractual assets is not affected. This point can be illustrated in the context of inventories. The value of an inventory (in a foreign subsidiary or at the parent) is determined not only by changes in the exchange rate but also by a subsequent price change of the product, to the extent that the cause of this price change is the exchange rate change. Thus, the dollar value of an inventory destined for export may increase when the currency of the destination country appreciates, provided that its local currency prices do not decrease by the full percentage of the appreciation. The effect on the local currency price depends, in part, on competition in the market. The behavior of foreign and local competitors, in turn, depends on capacity use, market share objectives, the likelihood of cost adjustments, and a host of other factors. Of course, firms are interested not only in the value change or the behavior of cash flows of a single asset but also in the behavior of all cash flows. Again, price and cost adjustments caused by an unexpected exchange rate change need to be analyzed. For example, a firm that requires components from abroad for production usually finds its stream of cash outlays rising whenever its local currency depreciates against foreign currencies. Yet, the depreciation may cause foreign suppliers to lower prices in terms of foreign currencies for the purpose of maintaining volume, thus mitigating or even neutralizing the effects of the currency change. [a] Currency Denomination. In this context, it is worthwhile to distinguish between the currency in which cash flows are denominated and the currency that determines the size of the cash flows. In the example just mentioned, it does not matter whether, as a matter of business practice, the firm may contract, be invoiced in, and pay for each individual shipment in its own local currency. If foreign exporters do not provide price concessions, the cash outflow of the importer behaves just like a foreign currency cash flow; even though payments are made in local currency: they occur in 35

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK greater amounts. As a result, the cash flow, even while denominated in local currency, is determined by the relative value of the foreign currency.32 The choice of currency in which the accounting records are kept may complicate things further. For example, any debt contracted by the firm in foreign currency is always recorded in the currency of the country where the corporate entity is located. However, the value of its legal obligation is established in the currency in which the contract is denominated. It is possible, therefore, that a firm selling in export markets may record assets and liabilities in its local currency and invoice periodic shipments in a foreign currency, and yet, if prices in the market are dominated by transactions in a third country, the cash flows received may behave as it they were in that third currency. To illustrate, a Brazilian firm selling coffee to Europe may keep its records in BRL, invoice in Euro, have Euro-denominated receivables, and physically collect Euro cash, only to find that its revenue stream behaves as if it were in USD. This occurs because Euro prices for each consecutive shipment are immediately adjusted to reflect world market prices, which, in turn, tend to be determined in USD.33 [b] Time Dimension. An additional dimension of exchange risk involves the element of time. In the very short run, virtually all local currency prices for real goods and services (although not necessarily for financial assets) remain unchanged after an unexpected exchange rate change. However, over a longer period, prices and costs move inversely to spot rate changes; the tendency is for PPP and the law of one price to hold. In reality, this price adjustment process takes place over a great variety of time patterns. These patterns depend not only on the products involved but also on market structure, the nature of competition, general business conditions, government policies such as price controls, and a number of other factors. Considerable work has been done on the phenomenon of pass-through of price changes caused by unexpected exchange 32 The functional currency concept introduced in SFAS No. 52 is similar to the currency of determination, but not exactly. "Currency of determination" refers to revenue and operating expense flows, respectively; the functional currency concept pertains to an entity as a whole, and is, therefore, less precise. 33 The significance of this distinction is that the currency of denomination is relatively readily subject to management discretion, through the choice of invoicing currency. Prices and cash flows, however, are determined by competitive conditions, which are typically beyond the immediate control of the firm. 36

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK rate changes.34 And yet, because all of the factors that determine the extent and speed of pass-through are firm-specific and can be analyzed only on a case-by-case basis at the level of the operating entity of the firm, the strategic business unit. The firm's foreign currency exposure is greatly influenced by the time frame within which management cannot react to unexpected rate changes by (1) raising prices; (2) changing markets for inputs and outputs; and/or (3) adjusting production and sales volumes. Sometimes, at least one of these reactions is possible within a relatively short time; at other times, the firm is locked-in through contractual or strategic commitments extending considerably into the future. Again, firms that are free to react instantaneously and fully to adverse unexpected rate changes are not subject to exchange risk - as long we abstract from foreign currency denominated receivables. A further important implication is that exchange risk stems from the firm's position in the market for inputs and outputs, rather than from any specific international involvement. Thus, companies engaged only in domestic transactions that have dominant foreign competitors may notice the effect of exchange rate changes in their cash flows as much as or even more than some firms that are actively engaged in exports, imports, or foreign direct investment35. [6] Implementation of Economic Exposure Analysis From this analytical framework, some practical implications emerge for the assessment of economic exposure. First of all, the firm must project its cost and revenue streams over a planning horizon that represents the period during which the firm is locked in, or constrained from reacting to unexpected exchange rate changes. It must then assess the impact of a deviation of the actual exchange rate from the rate used in the projection of costs and revenues. Thus, sensitivity analysis is called for with respect to the impact of unexpected exchange rate changes. Subsequently, the effects on the various cash flow of the firm must be netted over product lines and markets to account for diversification effects where gains and losses could cancel out, wholly or in part. The remaining net loss or gain is the focus of economic exposure management. For a multiunit, multiproduct, multinational corporation, the net exposure may not be very large at all because of the many offsetting 34 For a review of the literature, see Levi (1983), and more recently, Flood and Lessard (1986). For other illustrations, see "Dollar Drop Helps Those Who Help Themselves," Federal Reserve Bank of Chicago, Chicago Fed Letter, No. 7 (Mar. 1988). 35 Refer to Hodder who introduced this point for the first time. 37

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK effects. In contrast, enterprises that have invested in the development of one or two major foreign markets are typically subject to considerable fluctuations of their net cash flows, regardless of whether they invoice in their own or in the foreign currency. The same holds for those that are exposed to competition from dominant foreign rivals. Figure 5 shows how exposure might be estimated using statistical techniques. For practical purposes, three questions capture the extent of a company's foreign exchange exposure: 1. How quickly can the firm adjust prices in order to offset the effect of an unexpected exchange rate change on profit margins? 2. How quickly can the firm change sources for inputs and markets for outputs? Or how diversified are a company's factor and product markets? 3. To what extent do volume changes associated with unexpected exchange rate changes have an effect on the value of assets? Normally, the executives within business firms who can supply the best estimates regarding these issues tend to be those directly involved with purchasing, marketing, and production. Forex specialists who focus exclusively on credit and foreign exchange markets may easily miss the essence of corporate forex risk. [7] Critique of the Traditional View The concept of exposure developed so far, while consistent with classical cash flow analysis, differs in significant respects not only from the accounting concept of exposure but also from the traditional view of transaction exposure. Transaction exposure typically involves an export or import transaction giving rise to a foreign currency receivable or payable. On the surface, when the exchange rate changes, the value of this export or import transaction is affected in terms of the domestic currency. However, after careful analysis, it becomes apparent that the exchange risk results exclusively from a financial investment (the foreign currency receivable) or a foreign currency liability (the loan from a supplier) that is purely incidental to the underlying export or import transaction; it could have arisen in and of itself through independent foreign borrowing and lending. Thus, what is involved here are simply foreign currency assets and liabilities, whose value is contractually fixed in nominal currency terms. Figure 5 A Statistical Perspective on exposure 38

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Since the essence of economic exposure of a business firm is the movement of its cash flows together with exchange rate changes, it can be represented using regression techniques. a A multiple linear regression of the future domestic currency market price on a set of forex rates for the corresponding periods will produce coefficients whose dimension is units of foreign currency. A coefficient can be positive, negative, or zero, depending on the nature of the relationship. The regression separates the net cash flow changes and the change in the market price of the firm into two components; one part is correlated to the movements in exchange rates and one part is independent of them. Hedging techniques can reduce the first component, which reflects economic forex exposure; hedging techniques cannot reduce the other part, which is randomly determined. While the value of the firm is represented by the present value of its future cash flows after tax, the price of the firm's stock can be used as a proxy for this value. When changes in stock market price are regressed on a set of changes in forex rates for the same periods, coefficients associated with each currency indicate the amount of exposure in that currency. The t-statistic indicates whether the coefficient is statistically significant. The coefficient of determination (R2) indicates the percentage of the total variability that is explained by the regression coefficients. These are indicators of the amount of total risk that can and cannot be hedged. The regression equation is usually stated in the following form: CF, = a + bEXCH, + ut Where CFt is the dollar value of cash flows during period t, EXCHt is the nominal exchange rate (dollar This model is designed to develop a hedge ratio appropriate to the firm's economic exposure. A crucial assumption in the model is stability of the underlying relationships, because it is the persistence of the co-movement over time that makes the hedge effective. The model can be refined as well to decompose the spot rate in expected and unexpected changes. This can be accomplished by measuring the change in exchange rate as the difference in the future rate (at time t-1) at the spot rate (at time t). While theoretically correct, this regression may be of little practical assistance if the t-values of the coefficients are small, i.e., when the standard deviation of the coefficient is large relative to the point estimate. This situation occurs when the firm has hedged its exposure by means of the currency denomination of its debt and/or the forward exchange market or both or when there are other causes of sampling error. c a See M. Adler and B. Dumas, "Exposure to Currency Risk: Definition and Measurement," Financial Management (Summer 1984), pp. 41-50. b See C.K. Gamrner and A.C. Shapiro, "A Practical Method of Assessing Foreign Exchange Risk," Midland Corporate Finance Journal (Fall 1984), pp. 6-17. c See Maurice D. Levi, "Measuring Foreign Exchange Exposure From Regression Coefficients: A Practical Appraisal," Working Paper, University of British Columbia, July 1987. While this traditional analysis of transactions exposure is correct in a narrow, formal sense, it is relevant only for financial institutions. Returns from financial assets and liabilities fixed in nominal terms can be shielded from losses with relative ease through cash payments in advance (with appropriate discounts), through the factoring of receivables, or by way of the use of forward exchange contracts, unless unexpected 39

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK exchange rate changes have a systematic effect on credit risk.36 However, the essential assets of nonfinancial firms have noncontractual returns (i.e., revenue and cost streams from the production and sale of their goods and services that can respond to exchange rate changes in very different ways). Consequently, these assets are characterized by forex exposure that differs significantly from that of firms with contractual returns.37 [8] Corporate Exchange Risk in a Theoretical Perspective It may be useful to summarize the effect of unexpected exchange rate changes on the internationally involved firm by drawing on some well-known parity conditions. Given sufficient time, competitive forces and arbitrage neutralize the impact of exchange rate changes on the returns to assets; the relationship between rates of devaluation and inflation differentials leads these factors to neutralize the impact of the changes on the value of the firm as well. This simply reflects the principles of PPP and the law of one price operating at the level of the firm. On the liability side, the cost of debt tends to adjust as debt is repriced at the end of the contractual period in order to reflect (revised) expected exchange rate changes. And returns on equity also reflect required rates of return; in a competitive market, these are influenced by expected exchange rate changes. In the long run, a firm operating in this setting will not experience increasing exchange risk. However, because of a variety of contractual or, more important, strategic commitments, these equilibrium conditions rarely hold in the short and medium term. Therefore, forex exposure and, significantly, its management are made relevant by these temporary deviations. IV ECONOMIC EXPOSURE: MANAGEMENT AND HEDGING Corporate forex exposure originates with the cash flows associated with the operations of the firm; that is, the analysis must begin with the asset side. For the trading or manufacturing firm, liabilities are secondary in the sense that such companies have liabilities only because they have assets, with all of the concomitant risks and headaches involved in ownership and operation of assets. However, before the problem of appropriate asset and liability strategies can be tackled, the issue of forex rate forecasting must be addressed. 36 Conventional forex management is concerned only with nominal gains and losses in the numeraire currency of the firm. Protection of returns in real terms involves more complex considerations. 37 It is therefore good practice to segregate for analytical purposes the "finance company" business composed of purely financial assets, such as cash and receivables, and the associated financial liabilities from the remainder of the firm in order to focus on the "real" side of the business. 40

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK [1] Forecasting for Risk Control Few firms commit real assets in order to take currency positions. Rather, they get involved with foreign currencies in the course of market competition. Rather than being based on currency expectations, competitive advantage is based on expertise in such areas as production, marketing, the organization of people, or technical resources. Any special expertise in forecasting forex rates can usually be put to use without incurring the risks and cost of committing funds to real assets. Thus, most managers of nonfinancial enterprises are in decision-making roles only because their employers produce and sell goods; they do not act principally as speculative forex traders. Forecasting exchange rate changes, however, is important for planning purposes. To the extent that all significant managerial tasks are concerned with the future, anticipated exchange rate changes are a major input into virtually all decisions of enterprises involved in and affected by international transactions. However, the task of forecasting forex rates for planning and decision-making purposes, with regard to determining the most likely exchange rate, is quite different from attempting to beat the market in order to derive speculative profits. Expected exchange rate changes are revealed by market prices when rates are free to reach their competitive levels. Organized futures or forward markets provide inexpensive information regarding future exchange rates, using the best available data and judgment. Thus, whenever profit-seeking, well-informed traders can take positions, forward rates, prices of future contracts, and interest differentials for instruments with similar risk levels (but denominated in different currencies) provide good indicators of expected exchange rates. In this fashion, an input for corporate planning and decision making is readily available in all currencies where there are no effective exchange controls. The advantage of such market-based rates over in-house forecasts is that they are both less expensive and likely to be more accurate. Market rates are determined by those that tend to have the best information and track record; incompetent market participants lose money and tend to be removed as market participants over time. The nature of the market-based expected exchange rate should not lead to confusing notions about the accuracy of prediction. In speculative markets, all decisions are based on the interpretation of past data; however, new information surfaces constantly. Therefore, market-based forecasts rarely come true. The actual price of a currency will either be below or above the rate expected by the market. Any predictable, economically meaningful bias would be corrected by the transactions of profit-seeking transactors.38 38 See supra section F2.03. 41

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK The importance of market-based forecasts in determining the forex exposure of the firm is that of a benchmark against which the economic consequences of deviations must be measured. This can be put in the form of a concrete question: How will the expected net cash flow of the firm behave if the future spot exchange rate is not equal to the rate predicted by the market when commitments are made? The nature of this kind of forecast is completely different from an attempt to outguess the forex markets. [2] Financial Versus Operating Strategies When operating cash inflows and contractual outflows from liabilities are affected by exchange rate changes, a general principle suggests itself: Any effect on cash inflows and outflows should cancel out as much as possible. This can be achieved by maneuvering assets, liabilities, or both. Among the operating policies are the shifting of markets for output, sources of supply, product lines, and production facilities as a defensive reaction to adverse exchange rate changes. It is obvious that such measures will be very costly, especially if undertaken over a short span of time. Therefore, operating policies designed to reduce or eliminate exposure are undertaken only as a last resort, when less expensive options have been exhausted.39 It is not surprising, therefore, that exposure management focuses not on the asset side, but primarily on the liability side of the firm's balance sheet. The asset side is largely determined by broad strategic decisions, for instance, regarding in which market the company should sell its products; changes in the asset side can only be made slowly. Changes in the liability side are comparatively easy to make. To be precise, corporate forex management becomes a matter of choosing the appropriate currency denomination of debt (and financial assets) and a suitable maturity structure in terms of the interest period.40 [3] Debt Denomination: Strategy Versus Tactics Each unit of corporate debt must be denominated in one currency or another. When international debt markets are not distorted in a systematic fashion by exchange or credit market controls, the interest differentials closely reflect the expected appreciation 39 See Srinivasulu (1981). 40 It is necessary to distinguish here between maturity of the interest period and maturity with respect to the availability of funds. On floating rate loans, the former is considerably shorter than the latter. 42

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK or depreciation of the currency over the period during which the interest rate is contractually fixed. Thus, the expected cost of debt denominated in all currencies for which competitive international markets exist is the same for all debts. However, when future currency values differ from the expected value, the realized cost of debt in different currencies varies. The principle of exposure management is to cause this variance to offset any impact of this currency on the asset side. If simulations for expected forex values (deviations from the expected rate) over a reasonable range of outcomes show that the firm may not have sufficient capacity for debt to hedge fully against a large swing in cash flows, the firm could use phantom debt (i.e., forward contracts or currency swaps). This raises the issue of the appropriate role of forward contracts in exposure management. It must be recognized that forward contracts in exposure management are simply tools to change the currency of denomination of financial assets or liabilities. For this reason, some of the literature on foreign currency risk management carefully distinguishes between covering and hedging. Covering a transaction means simply to change its effective currency denomination. Whether such a change will also hedge a position depends entirely on the nature of the exposure. For example, an importer with a foreign currency payable may cover it with a forward contract, which makes it, in effect, a domestic currency liability. If, however, the pricing of the imported products follows the foreign currency (currency of determination), the cover operation exposes its expected net cash flow to exchange risk. Debt denomination therefore is the overriding strategic consideration. It can be achieved either through outright borrowing in the desired currency or indirectly by the use of forward contracts and swaps. For example, foreign currency debt can be turned into USD debt by entering into a commitment to purchase the foreign currency at a future point in time, taking interest and principal into account. Also, USD debt can be turned into foreign currency debt by contracting to deliver foreign currency in the future in return for dollars. Once management has decided, based on exposure analysis, on the appropriate currency denomination of its debt, it must choose the method of implementing that decision. It can choose its denomination directly, through borrowing, or indirectly, through forward contracts; in choosing, it must assess such factors as government restrictions that may limit access to markets, transaction costs, tax factors, accounting practices, and financial public relations (the maintenance of certain financial ratios). 43

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Further differences in liquidity and flexibility must be considered.41 But these are tactical decisions; the strategic decision involves the determination of the appropriate currency denomination of debt. Of course, when markets are controlled, the decision to borrow may give rise to arbitrage opportunities that can be exploited by means of swaps and parallel loans, resulting in special gains or the avoidance of excess interest costs. If the achievement of such gains does not conflict with the desired debt denomination from an exposure point of view, the decision is clear. If, however, exposure analysis calls for debt denomination in, e.g., USD, but because of capital market controls or subsidization, particularly inexpensive local currency loans are available while forward markets are nonexistent, the firm is faced with a painful trade-off between the expected (but uncertain) gains from borrowing the "cheap" currency, and a net cash flow exposure to the vagaries of the exchange rate. Only after all avenues of debt policy have been explored does the firm resort to these types of operating, or asset, strategies if the remaining exposure is unacceptably large. The reason is simply that changes in debt denomination are relatively inexpensive to implement. The following is a summary of the steps involved in managing economic exposure: 1. Estimation of the planning horizon, as determined by the reaction period. 2. Determination of the expected future spot rate. 3. Estimation of the expected revenue and cost streams, given the expected spot rate. 4. Estimation of the effect on revenue and expense streams of unexpected exchange rate changes. 5. Choice of the appropriate currency for debt denomination. 6. Estimation of the necessary amount of foreign currency debt. 7. Determination of the average interest period of debt. 8. Selection between direct or indirect debt denomination. 9. Decision on the trade-off between arbitrage gains and exchange risk stemming from exposure in markets where rates are distorted by controls. 10. Decision about "residual" risk; considering adjustments to business operations and strategy. Realistically, a nonfinancial firm should not expect that exchange risk can be hedged completely with debt denomination, either direct or indirect. This is because of the nature of its assets; while a firm with a portfolio consisting only of assets with contractual returns can hedge ("immunize") its value in nominal terms, an enterprise 41 For an extensive analysis of debt denomination under various conditions, including the incidence of taxes, see Jacque and Lang (1987). 44

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK whose assets yield, for the most part, noncontractual returns cannot hedge its exchange risk perfectly by way of appropriate debt denomination. This is because a given exchange rate change has a different effect on the return of (noncontractual) assets than on (contractual) liabilities. Still, this residual exchange risk should be reasnonably small in most circumstances, and it does not detract from the importance of debt denomination as the lowest-cost hedging policy. [4] Foreign Currency Options and Exchange Risk Management Contractual cash flows denominated in a foreign currency can be hedged against unexpected changes of the exchange rate through the use of forward contracts and swaps. However, a company may often be confronted with contingent cash inflows or outflows denominated in foreign currencies. The most common case is when a company submits a bid for a foreign project; should the firm win the bid and complete the project, it would receive a certain amount of foreign currency. In this case, the forward contract is a perfect hedge instrument only if the firm is certain of winning the bid for the project. If the firm loses the bid, it will still be obliged to perform under the forward contract. Suppose that a U.S. company bids for a project in the United Kingdom, with payments to be received in GBP. If the company does not hedge this contingent open position and the bid is awarded, the actual USD cash inflow will be less if the future GBP-USD exchange rate is lower than expected at the time the bid was submitted. Thus, the firm must hedge this contingent open position against unexpected downward changes in the exchange rate. If certain instruments would give the firm the right to sell GBP for USD at a predetermined rate, subject to the condition of winning the contract, this contingent open position could be perfectly hedged. In fact, such instruments do exist in part, in the form of put and call foreign exchange options. In the preceding case, the firm would buy a call option on USD, that is, the equivalent to a put option on GBP, which would give the firm the right to buy USD at a predetermined rate (exercise price) on a predetermined date (maturity date) subject to the firm's winning the contract. Figure 6 summarizes the hedging policies that should be undertaken when a firm is exposed to contingent and fixed open positions. Unfortunately, options are usually written only on the value of the currency, not on the award of the contract. Thus, the firm still has exposure with respect to the value of the option, determined largely by the change in the expected volatility of the currency over the remaining life of the option. Figure 6 Foreign Currency Cash Flows 45

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Source: I.H. Giddy, "Foreign Exchange Option as a Hedging Tool, "Midland Corporate Finance Journal, Vol. 1 (Fall 1983) Outflow Inflow Amount known Buy forward Sell forward Amount unknown Buy call option Buy put option Amount partially known Hedge known portion with forward contracts and remainder with options V CORPORATE PRACTICE AND THE RELEVANCE OF CURRENCY HEDGING Two major issues remain to be addressed: (1) the variance of corporate practice with the normative model just developed, and (2) whether forex risk management should be a legitimate concern of management in light of recent developments in corporate finance and capital market theory. The first issue is considered from the perspective of the practitioner; the second issue is from the perspective of the theory of finance. [1] Corporate Exchange Risk Management Practice Corporate forex decisions tend to be highly political. Governments generally take offense at corporate expression of a lack of confidence in the purchasing power of a local currency by the taking of short positions, and speculation has a strong negative connotation everywhere. Most important, because forex decisions are easy to criticize with hindsight, such decisions tend to be highly sensitive in a corporate political environment. This is accentuated by the natural tension that exists between managers responsible for operating decisions and those responsible for financing decisions. Last but not least, the decision-making process may be obscured by motives other than the pure optimization of shareholder wealth.42 A glance at corporate policies reveals a disproportionate concern with accounting exposure, although most of the corporate financial staff have been well educated on the difference between accounting and economic measures of exposure. This preoccupation with accounting exposure would be understandable if the evidence 42 For an incisive analysis of the managerial issues in corporate forex risk management, see Lessard and Nohria (1989). 46

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK on stock market performance indicated that stock prices are influenced by reported translation losses and gains. However, the evidence suggests that stock market activity is based not on the accounting results reported but on the cases where translation losses and gains coincide with real losses and gains and situations where the two are at variance.43 There are a number of explanations that make corporate practice quite rational. First of all, economic exchange risk is reasonably complex. All relevant data must be estimated, and are therefore difficult to audit and to justify to a critical audience. Accounting losses, on the other hand, are very visible and can be determined precisely. Along the same lines, accounting-based foreign exchange management is straightforward and almost mechanical; translation rules indicate which assets and liabilities are exposed; the firm must then project its balance sheet to the end of the reporting period and ensure that exposed assets are funded with exposed liabilities or that net-exposed positions are covered with a forward contract. If these policies are not feasible because of capital market controls, various asset strategies can be used to reduce exposed assets: factoring of local currency receivables, conversion of monetary assets into real ones (and vice versa), and similar machinations. While the tools are very much the same as in economic exposure management, accounting exposure will often be different in amount and often in direction; a given exchange rate change may have a negative effect on accounting exposure but a positive effect on expected future cash flows.44 However, for a multi-product, multi-market company, economic exposure and, therefore, cash flow variability may be relatively small. In contrast, the necessity for an explanation of excessive earnings variability owing to translation gains and losses may be of concern to the manager selected for decision-making skills rather than analytical, expository capabilities. Furthermore, the reporting system in many companies seems to highlight forex gains and losses much more than the interest expense typically incurred in balance sheet hedging. Last but not least, incentive compensation systems that are tied to reported earnings explain much of observed behavior. Also, a considerable amount of speculation seems to be occurring without due regard for the nature of markets. Rarely do financial executives appear to inquire as to 43 Yeater, "The Impact of Statement of Financial Accounting Standard No. 8 on Corporate Value," Ph.D. Thesis, Cornell University, 1978: J. Williams, "Capital Market Reaction to Financial Accounting Standards Board Statement No. 8," Ph.D. Thesis, Pennsylvania State University, 1978; R. Dukes (1978). 44 See Dufey (1972). 47

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK the markets in which positive payoffs over the long run can result from forecasting efforts. Judging by the criterion of currencies covered by commercial forecasting services, the demand seems to be greatest for the major currencies where well-informed traders are most active because governments interfere least with transactions. This is difficult to explain, and the observation that speculation at the corporate level is usually quite modest relative to the size of the firms suggests that perhaps the true function of these activities is to keep the minds of treasury staff sharply focused on markets so they can better identify any market imperfections that can be exploited. [2] Value Creation though Managing Corporate Exchange Exposure45 The theory of finance suggests that the management of corporate forex exposure may be neither an important nor a legitimate concern. It has been argued, in the tradition of the Modigliani-Miller theorem, that investors themselves can hedge corporate exchange exposure by taking out forward contracts and similar derivatives in accordance with their share in the firm. But apart from transaction costs, which are typically larger for individuals than for firms (but not institutional investors), there are more serious reasons why fore risk should be managed at the firm level. 46 As pointed out above, the assessment of economic exposure requires detailed estimates of the susceptibility of net cash flows to unexpected exchange rate changes. Operating managers can make such estimates with much more precision than shareholders, which typically lack the detailed knowledge of markets and technology. Furthermore, the firm has considerable advantages in obtaining relatively inexpensive debt at home and abroad by taking maximum advantage of interest subsidies and minimizing the effect of taxes and political risk. While this reasoning may not hold up in a world where shareholdings are concentrated among institutional investors such as pension and mutual finds, sometimes it is argued that corporate forex risk management does not matter because of the functioning of PPP and the international Fisher effect: over time, prices of inputs and outputs adjust and so does the cost of funds, both debt and equity. However, as pointed out previously, deviations from PPP can persist for considerable periods. And such deviations can be especially pronounced at the level of the individual firm. The resulting variability of net cash flows is of significance, as it can cause the firm to be 45 For an extensive analysis of the underlying issues and the literature see Bartram (2000) 46 Dufey and Srinivasulu (1983). 48

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK unable to realize profitable investments if there are any constraints of access to capital markets. 47 However, such "costs" are simply part of a larger category of disadvantages related to the costs offinancial distress: the firm that shows financial weakness will find it increasingly difficult to attract suppliers who are willing to invest in the relationship by adding capacity, designing components and similar commitments. By the same token, customers do not like to purchase good from companies that may go out of business esp. when explicit or implicit warranty claims are in jeopardy. Last but not least, management personnel, present and potential, will worry about their future and will react by demanding a risk premia in their compensation packages. The same argument supports the importance of corporate exchange risk management against the claim that only systematic risk matters. To the extent that forex risk represents unsystematic risk it can, of course, be diversified (hedged), provided again that investors have the same quality of information about the firm as management. However, there is one task that the firm cannot perform for shareholders: to the extent that individuals face unique exchange risk as a result of their different expenditure patterns, they must themselves devise appropriate hedging strategies. After all, corporate financial management is only able to protect expected nominal returns in the reference currency.48 Suggested Reading Adam, T.R. "Risk Management in the Gold Mining Industry." Hong Kong University Working Paper (1997). Adler, M., and B. Lehman. "Deviations From Purchasing Power Parity in the Long Run." Journal of Finance, Vol. 38 (Dec. 1983), pp. 1471-1488. Adler, M. and Dumas B. "International portfolio choice and corporation finance: A synthesis." Journal of Finance (1983) 38, pp. 925-984. Aliber, Robert Z., ed. The Handbook of International Financial Management. Homewood, Ill.: Dow Jones-Irwin, 1989. Allayannis, G, and E. Ofek. "Exchange Rate Exposure, Hedging, and the Use of Foreign Currency Derivatives." New York University Working paper (1996). "Banking Gets 'in the CHIPS.' " Morgan Guaranty Survey (May 1975), pp. 12-14. 47 Froot et. al (1993) 48 Eaker(1981) 49

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Bartram, S. "Corporate Risk Management as a lever for Shareholder Value Creation, Financial Markets, Institutions & Instruments, Vol. 9, No. 5, (December 2000) Beenstock, M. "Forward Exchange Rates and Siegel's Paradox." Oxford Economic Papers, Vol. 37 (1985), pp. 298-303. Biger, N., and J. Hull. "The Valuation of Currency Options." Financial Management, Vol. 12 (Spring 1983), pp. 24-28. Bodnar, G.M., and G. Gebhardt. "Derivatives Usage in Risk Management by U.S. and German NonFinancial Firms: A Comparative Survey." Wharton Working Paper (1999)., G.S. Hayt, and R.C. Marston. "1998 Wharton Survey of Financial Risk Management by U.S. Nonfinancial Firms." Financial Management (1998) 27:4, pp. 70-91. Business International Money Report (Oct. 28, 1977), pp. 339-340. Coninx, Raymond G.F. Foreign Exchange Dealer's Handbook. New York: Pick Publishing, 1982. -. Foreign Exchange Today. New York: John Wiley & Sons, Inc., 1978. Cornell, B. "Inflation, Relative Price Changes, and Exchange Risk." Financial Management, Vol. 9 (Autumn 1980), pp. 30-34.. "Spot Rates, Forward Rates, and Exchange Market Efficiency." Journal of Financial Economics, Vol. 5 (1977), pp. 55-65. De Santis, G., and Gerard, B. "How big is the premium for currency risk?" Journal of Financial Economics (1998) 49, pp. 375-412. Deardorff, Alan V. "One-Way Arbitrage and Its Implications for the Foreign Exchange Markets." Journal of Political Economy, Vol. 87 (1979), pp. 351-364. Dooley, M.P., and P. Isard. "Capital Controls, Political Risk and Deviations From Interest Parity." Journal of Political Economy, Vol. 88 (Mar./Apr. 1980), pp. 370-384. Dornbusch, R. "Exchange Rate Economics: Where Do We Stand?" Brookings Papers on Economic Activity, Vol. 1 (1980), pp. 143-185.. "Expectations and Exchange Rate Dynamics." Journal of Political Economy, Vol. 84 (Dec. 1976), pp. 1161-1176. Doukas, J., Hall, P.H., and Lang, L.H.P. "The pricing of currency risk in Japan." Journal of Banking and Finance (1999) 23, pp. 1-20. Dufey, G. "Corporate Finance and Exchange Rate Variations." Financial Management, Vol. 1 (Summer 1972), pp. 51-57. Dufey, G. and I. Giddy. "International Financial Planning: The Use of Market-Based Forecasts." California Management Review, Vol. 12 (Fall 1978), pp. 69-81. 50

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Dufey, G., and S.L. Srinivasulu. "The Case for Corporate Management of Foreign Exchange Risk." Financial Management (Winter 1983), pp. 54-62. Dumas, B., and Solnik, B. "The world price of foreign exchange risk." Journal of Finance (1995) 50, pp. 445-479. Eaker, M.R. "The Numeraire Problem and Foreign Exchange Risk." Journal of Finance, Vol. 36 (May 1981), pp. 419-427. "Covering Foreign Exchange Risks: Comment." Financial Management, Vol. 9 (Winter 1980), pp. 64-65. Eiteman, D.K., Stonehill, A. I., and M. H. Moffett. Multinational Business Finance, 9th ed. Boston: Addison Wesley Longman 2001. Fehle, F. "A Cross-Sectional Study of Corporate Hedging." University of Texas Working Paper (1998). Financial Accounting Standard Board, "Summary of Statement 133: Accounting for derivative instruments and hedging activities." Issued 6/98. Fisher, Irving. The Theory ofInterest. New York: Macmillan, 1980. Flood, E., Jr., and D.R. Lessard. "On the Measurement of Operating Exposure to Exchange Rates: A Conceptual Approach." Financial Management, Vol. 15 (Spring 1986), pp. 25-36. "Focus: The Foreign Exchange Markets." Columbia Journal of World Business (Winter 1979). Foreign Exchange and Money Market Operations. Swiss Bank Corporation. Zurich: June 1987. "The Forex Market Goes West." Euromoney (Aug. 1983), pp. 59. Franke, G. "Uncertain Perception of Economic Exchange Risk and Financial Hedging." Managerial Finance (1992) 18:3/4, pp. 53-70. Frankel, J. "On the Mark: A Theory of Floating Exchange Rates Based on Real Interest Differentials." American Economic Review, Vol. 69 (Sept. 1979), pp. 610-622.. A Test of the Existence of the Risk Premium in the Foreign Exchange Market vs. the Hypothesis of Perfect Substitutability. Federal Reserve System, International Finance Discussion Papers No. 149. Aug 1979. Frenkel, J. "A Monetary Approach to the Exchange Rate: Doctrinal Aspects and Empirical Evidence." Scandinavian Journal of Economics, Vol. 78 (June 1978), pp. 255-276. Frenkel, J.A., and R.M. Levich. "Covered Interest Arbitrage: Unexploited Profits?" Journal of Political Economy, Vol. 83 (Apr. 1975), pp. 325-338. 51

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Froot, K.A., D.S. Scharfstein, and J.C. Stein. "Risk Management: Coordinating Corporate Investment and Financing Policies." Journal ofFinance (1993) 48:5, pp. 1629-1658. Fukao, Mitsuhiro. "The Risk Premium in the Foreign Exchange Market." Ph.D. Dissertation, University of Michigan, 1981. Giddy, I.H. "Exchange Rate Volatility, Hedging, and the cost of Capital." in Exchange Rates and Corporate Performance. Amihud, Y., and R.M. Levich (eds). Irwin (New York, NY), (1994) pp. 155-162. "Exchange Risk: Whose View?" Financial Management, Vol. 7 (Summer 1977), pp. 23-33.. "Foreign Exchange Options," The Journal of Futures Markets, Vol. 3 (1983), pp. 143-166.. "An Integrated Theory of Exchange Rate Equilibrium." Journal of Financial and Quantitative Analysis, Vol. 11 (Dec. 1976), pp. 883-892.. "Measuring the World Foreign Exchange Market." Columbia Journal of World Business (Winter 1979), pp. 36-48. Goedhuys, Diederik, ed. The Foreign Exchange Market in the 1980s: The Views of Market Participants. New York: Group of Thirty, 1985. Grauer, F.L.A., Litzenberger, R.H., and Stehle, R.E. "Sharing rules and equilibrium in an international capital market under uncertainty. " Journal of Financial Economics (1976) 3, pp. 233-256. Hacche, G., and J.C. Townsend. "A Broad Look at Exchange Rate Movements for Eight Currencies, 1972-80." Bank of England Quarterly Bulletin (Dec. 1981), pp. 489-509. Harvey, C.R. "Time-varying conditional co-variances in tests of asset pricing models." Journal of Finance (1991) 46, pp. 289-317. Hekman, C.R. "Don't Blame Currency Values for Strategic Errors: Protecting Competitive Position by Correctly Assessing Foreign Exchange Exposure." Midland Corporate Finance Journal, Vol. 4 (Fall 1986), pp. 45-55.. "Foreign Exchange Exposure: Accounting Measures and Economic Reality." Journal of Cash Management (Feb./Mar. 1983), pp. 34-45. Henderson, Dale W. The Dynamic Effects of Exchange Market Intervention Policy: Two Extreme Views and a Synthesis. Federal Reserve Board, International Finance Discussion Papers No. 142, June 1979. Hilley, J.L., C.R. Beidleman, and J.A. Greenleaf. "Why There Is No Long Forward Market in Foreign Exchange." Euromoney (Jan. 1981), pp. 94-103. Hodder, James E., "Exposure to Foreign Exchange-Rate Movements", Journal of International Economics, Vol. 13 (11). pp. 375-386. 52

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Isard, Peter. Exchange Rate Determination: A Survey of Popular Views and Recent Models. Princeton University, International Finance Section, Studies in International Finance No. 42. Princeton, N.J.: May 1978. Jacque, L.L. "Management of Foreign Exchange Risk: A Review Article." Journal of International Business Studies, Vol. 12 (Spring/Summer 1981), pp. 81-101. Jacque, L.L., and P. Lang. "Currency Denomination in Long-Term Debt Financing and Refinancing: A Cross-Hedging Paradigm." Journal of the Operational Research Society, Vol. 38 (1987), pp. 173-182. Kobrin, S.J. "Political Risk: A Review and Reconsideration." Journal of International Business Studies, Vol. 10 (Spring-Summer 1979), pp. 67-80. Kohlhagen, Steven W. The Behaviour of Foreign Exchange Markets - A Critical Survey of the Empirical Literature. New York University, Monograph Series in Finance and Economics, Monograph No. 4. New York, 1983. Kolb, Robert W., ed. The International Finance Reader, 2nd ed. Miami: Kolb Publishing, 1993. Kubarych, Roger M. Foreign Exchange Markets in the U.S., rev. ed. New York: Federal Reserve Bank of New York, 1983. Lessard, Donald R., ed. International Financial Management: Theory and Application, 2nd ed. New York: John Wiley & Sons, Inc., 1985. Lessard, Donald R., and Nitin Nohria. "Rediscovering Functions in the MNC: The Role of Expertise and the Emergent Matrix in Firms' Responses to Shifting Exchange Rates," Managing the Global Corporation, Bartlett, Doz, and Hedlund, eds. London: Routledge, 1989. Levi, M.D. "Exchange Rates and the Valuation of Firms." in Exchange Rates and Corporate Performance. Amihud, Y., and R.M. Levich (eds). Irwin (New York, NY), (1994) pp. 37-48. Financial Management and the International Economy. New York: McGraw-Hill, 1983. Lewent, J.C., and A.J. Keamey. "Identifying, Measuring, and Hedging Currency Risk at Merck." Journal of Applied Corporate Finance (1990) 2:4, pp. 19-28. Logue, D.E., and G.S. Oldfield. "Managing Foreign Assets When Foreign Exchange Markets Are Efficient." Financial Management, Vol. 6 (Summer 1977), pp. 16-22. Mandich, D.R., ed. Foreign Exchange Trading Techniques and Controls. Washington, D.C.: American Bankers Association, 1976. Mayers, D. and C.W. Smith, Jr. "Corporate Insurance and the Under-investment Problem." Journal of Risk and Insurance (1987) 54:1, pp. 45-54. McCormick, F. "Covered-Interest Arbitrage: Unexploited Profits: Comments." Journal of Political Economy, Vol. 87 (Apr. 1979), pp. 411-417. 53

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Officer, Lawrence. Purchasing-Power-Parity and Exchange Rates: Theory, Evidence and Relevance. Greenwich, Conn.: JAI Press, 1982. - "The Purchasing-Power-Parity Theorem of Exchange Rates: A Review Article." IMF Staff Papers (Mar. 1976), pp. 1-60. Price Waterhouse. Foreign Currency Translation: Understanding and Applying FASB 52. New York: 1981. Reier, S. "Wall Street Muscles In On Foreign Exchange." Institutional Investor, international ed. (Jan. 1984), pp. 133-134. Report on Exchange Restrictions. Washington, D.C.: International Monetary Fund (annual). Revey, P.A. "Evolution and Growth of the United States Foreign Exchange Market." Quarterly Review. Federal Reserve Bank of New York (Autumn 1981), pp. 32-33. Riehl, Heinz, and Rita M. Rodriguez. Foreign Exchange and Money Markets. New York: McGraw-Hill, 1983. Rodriguez, R.M. "Corporate Exchange Risk Management: Theme and Aberrations." Journal of Finance, Vol. 36 (May 1981), pp. 427-439.. Foreign Exchange Management in U.S. Multinationals. Lexington, Mass.: D.C. Heath, 1980. Roll, R. and S. Yan. "An explanation of the forward premium puzzle." European Financial Management (2000) 6, pp. 121-148. Santomero, A.M. "Financial Risk Management: The Whys and Hows." Financial Markets, Institutions & Instruments (1995) 4:5, pp. 1-14. Shapiro, A.C. Multinational Financial Management, 6th ed. New York: John Wiley 1999. Shapiro, A.C., and D.P. Rutenberg. "Managing Exchange Risks in a Floating World." Financial Management, Vol. 5 (Summer 1976), pp. 48-58.. "What Does Purchasing Power Parity Mean?" Journal of International Money and Finance (Summer 1983), pp. 295-318. Showers, Janet. Hedging the Currency Exposure of Foreign Subsidiaries: Strategic Issues. Salomon Bros. Bond Portfolio Analysis Group. New York: Oct.1988. Smithson, C.W. Managing Financial Risk, 3rd ed. New York: McGraw Hill 1998. Solnik, B., "An equilibrium model of the international capital market." Journal of Economic Theory (1974) 8, pp. 500-524. Stein, J.L., M. Rzepczynski, and R. Selvaggio. "A Theoretical Explanation of the Empirical Studies of 54

FOREIGN EXCHANGE MARKETS AND CURRENCY RISK Futures Markets in Foreign Exchange and Financial Instruments." The Financial Review, Vol. 18 (Feb. 1983), pp. 1-32. Stulz, R.M. Derivatives, Risk Management, and Financial Engineering. Southwestern Publishing Co. (Cincinnati) (2000). "Rethinking Risk Management." Bank of America Journal of Applied Corporate Finance (1996) 9:3, pp. 8-24. "Optimal Hedging Policies." Journal of Financial and Quantitative Analysis, Vol. 19 (1984), pp. 127-140. "A model of international asset pricing." Journal of Financial Economics (1981) 9, pp. 383 -406. Tai, C.S. "Time-varying risk premia in foreign exchange and equity markets: Evidence from AsiaPacific countries, Journal of Multinational Financial Management (1999) 9, pp. 291-316. Vassalou, M. "Exchange rate and foreign inflation risk premiums in global equity returns." Journal of International Money and Finance (2000) 19, pp. 433-470. 55