LC H19/3.‘ 87-429 E A/ CRS REPORT FOR .1 '%9"':fiJ3Tlent p,,,,,.,c I. . . ,g . . a‘ ' \‘:§' ,.__';«"‘ ‘4 ‘ -... -~ ~-~7 7_, ...‘ V.“ .,\_l .3‘, _'_ ‘fl _ . ‘ * U” “"‘-':'5~ {‘*'«».»=~'::' V .fi:—..!-xi‘; fgaitut 6 I 7 :.‘”/\"+. 35 7 E P F 32 H ‘ A 4 2 ‘-A ‘-1- v: =- .. /«E :5 /, at if 2 3 . ‘Eu’ 3 5? I ' .; 5. '3 -\ ‘I‘ . .. ‘C»\_)Y 5. 1 E; r~_«, K ‘SJ Q . '1'; I L‘. a§?P§Q?‘°” Umver §T§‘.a,£F%' K :9’, :21‘. s’ ' - 2. Laws’ Moxtgallggrartes HOW MUCH HAS THE INTERNATIONAL EXCHANGE VALUE OF THE IXJLLAR DECLIHED? The movement in the international exchange value of the dollar is important to a number of public policy issues as well as understanding movements in GNP, employment, and inflation. Unfortunately, the various exchange indexes do not register the same movement in the value of the dollar over a given time period. — by Gail Makinen Specialist in Econanic Policy Economics Division May 13, 1987 vers ‘of issouri C um a ° mmiiili iflfllmii mlnu ‘Mm ol o1o- o39399"oL’ san bias, in man,y,.fo_r '»-.‘- “Ev-’ . The Congressional Research Service works exclusively for the Congress, conducting research, analyzing legislation, and providing information at the request of committees, Mem- bers, and their staffs. The Service makes such research available, without parti- g s including studies, reports, compila- ;_ ground briefings. Upon request, CRS t sists colmmiitfl es in analyzing legislative proposals and A ;~~¢c*v*-'~*'&”*'~ . . $3, T; gessmg the poss1ble effects of these proposals d _‘ i' ateflatiyes. The Service’s senior specialists and ' M i - also available for personal consultations ectlve fields of expertise. ,1 .,; 4-K. - . CRS-iii CONTENTS WHAT IS A FOREIGN EXCHANGE RATE INDEX? . . . . . . . . . WHY IS A FOREIGN EXCHANGE RATE INDEX IMPORTANT? . . . . . HOW SHOULD MOVEMENTS IN THE INDEX BE INTERPRETED? . . . . WHY DOLLAR INDEXES DO NOT RECORD THE SAME RATES OF CHANGES Does the Index Measure Real or Nominal Magnitude? . Are the Sample Sizes the Same? . . . . . . . . . . . What Type of Weighting Pattern is Used? . . . . . . O O O O O O I O O I O O O I O O O O I O O O O \J UV-§f\)t\> HOW MUCH HAS THE INTERNATIONAL EXCHANGE VALUE OF THE DOLLAR DECLINED? There is general agreement that the international exchange value of the dollar has fallen since early 1985. There is, however, a general diversity of opinion on the extent of the decline. Depending on which dollar exchange index is consulted, the fall in the value of the dollar through the fourth quarter of 1986 ranges from a low of about 10 percent to a high of nearly 32.0 percent. why such a diverse range? which figure is correct? What are the implications of the depreciation of the dollar? Answers to these questions are important for understanding a range of national issues and for formulating various policy options. The goal of this report is to describe the purposes served by a foreign exchange rate index and why existing indexes yield such a wide range of estimates on the extent of dollar depreciation over the past two years. WHAT IS A FOREIGN EXCHANGE RATE INDEX? Quite simply, a foreign exchange rate index measures a relative price. In the case of the United States, the index is the price of the dollar in terms of the currency of one or more countries. Thus, the index could measure the price of the dollar in terms of, or relative to, the British pound or the French franc or the German mark or to a combination of all three currencies. This report will consider the relationship of the dollar to a group of currencies for countries that are U.S. trading partners. WHY IS A FOREIGN EXCHANGE RATE INDEX IMPORTANT? There are a number of important economic variables whose movement can be explained by movements in the international value of the dollar. For ex- ample, changes in the international value of the dollar can influence GNP growth, the level of employment, the rate of inflation, the ability of ex- port and import competing firms to sell in various markets, and, hence, the composition of aggregate output and employment. In addition, the value of the dollar has implications for the level of well-being of the entire country for it governs how many resources embodied in goods and services the Nation receives from abroad for each unit of resources given up. There are also a number of important national policy issues that can be formulated, clarified, and understood, and a range of policy options pro- posed, by reference to movements in the exchange value of the dollar. For a country such as the United States, facing a large current international trade imbalance, reference to the foreign exchange value of the dollar en- ables policymakers to understand the situation and formulate effective reme- dial action. By reference to the index, they can see how the value of the CRS-2 dollar has moved over time, whether the movement is, in some sense, "enough" to accomplish goals and, if it is not, to formulate policies that will move the dollar toward the "proper" level. HOW SHOULD MOVEMENTS IN THE INDEX BE INTERPRETED? Among the many issues related to the exchange value of the dollar, two are especially important: the relationship of the dollar to GNP growth and to the economic well-being of the Nation. For the former, a rise in the value of the dollar (or an appreciation) should signal that the ability of U.S. exporting firms to sell abroad and U.S. import competing firms to meet foreign rivals in the U.S. market is diminished. Insofar as economic well-being is concerned, a dollar appreciation should indicate enhancement, for now each dollar should command more foreign currency or each unit of U.S. resources given up should command more units of foreign resources. This measure of well-being is often called the "terms of trade" or, more specifically, the net barter terms of trade. A fall in the value of the dollar should signal the opposite: an in- crease or enhanced ability of U.S. export and import-competing firms to sell in their respective markets and a deterioration in the overall wellbeing of the United States (as measured by the net barter terms of trade). _1_/ Having discussed the nature of a foreign exchange rate index, why they are important and how they should be interpreted, we now turn to a discus- sion of why they record different rates of movement in the exchange value of the dollar. While this involves a discussion of the more technical aspects of index number construction, it should not be forgotten that our interest centers on whether the particular index serves the two purposes mentioned above. WHY DOLLAR INDEXES DO NOT RECORD THE SAME RATES OF CHANGES Does the Index Measure Real or Nominal Magnitude? Few who have lived through a period of inflation, such as that in the United States during the late 1970s and early 1980s, forget that a dollar does not always conmand an unchanged amount of goods and services. They soon learn that what is important is not the number of dollars received, but what they buy. Thus, they think in terms of the real value of the dollar rather than its nominal value. y The net barter terms of trade is not the only way to measure a change in economic well-being. This measure has the disadvantage of neglecting changes in productivity, both domestic and foreign. When productivity changes are also considered, it is possible for a depreciation of the dollar to be consistent with an enhanced national well-being. For that reason, the net barter terms of trade is useful primarily as a short-run measure of well- being. Similarly, not all depreciations of the dollar enhance the ability of U.S. firms to sell abroad and compete in domestic markets. One must first inquire whether changes in the value of the dollar are the cause of GNP changes, for example, or are caused by changes in GNP. CR8-3 Similarly,iJ1an.international context, for the dollar exchange rate to perforn1such.roles as an index of general well-being, asauiindicator of the ability of U.S. firms to sell in foreign and home markets, and as a signal of what might be expected to happen to future GNP growth, what is important is its real exchange rate not the number of units of foreign currency it can fetch. 3 To understand the distinction between a nominal and a real index and the importance of the latter, consider the following simple example. Assume the following: that initially the U.S. dollar exchanges for one Mexican peso, that over the course of a year Mexican prices double while U.S. prices remain unchanged, and that at the end of the year one U.S. dollar exchanges for two Mexican pesos. Clearly, the dollar has risen in nominal terms or in terms of the peso for it buys twice as many at year's end as it did at the beginning. A nomi- nal index would record a dollar appreciation. But does the dollar buy twice as many Mexican goods and services as previously? The answer is no. In fact, it commands the same quantity as before. Moreover, the United States is not in any sense better off nor have U.S. firms suffered any impairment in their ability to sell in home and Mexican markets. A real or price-ad- justed exchange rate index would adjust the movement in nominal magnitudes to reflect the change in the price levels in the two countries and, thus, would show no movement (or appreciation) in the value of the dollar. Thus, one of the reasons various dollar exchange rate indexes show dif- fering rates of change is that some are nominal while others are real . One of the interesting examples of a nominal index is the X-131 index compiled by the Federal Reserve Bank of Dallas. It is an index that covers all 131 nations with whom the United States trades. when it was first announced in December 1986, it showed that the dollar had hardly fallen in value at all since early 1985. This, in turn, was used by some analysts to explain why the United States had failed to show improvement in its trade performance. The reason X-131 showed such a slight depreciation in the value of the dollar is that it was dominated by the sharp _1i§_e_ in the value of the dollar relative to the currencies of Mexico and Brazil; a rise designed to offset the much higher rates of inflation in those two countries. In fact, in real or inflation adjusted terms, the dollar scarcely rose against those two cur- rencies. In a purely nominal index, such as X-131, the dollar appeared to be stronger than it was. This example serves to highlight a dangerous aspect of nominal indexes. If they include countries whose rates of inflation are either markedly higher or lower than the United States, and the dollar adjuststx>reflect this differential, a nominal index will either understate or overstate the true movement in the dollar's value. Of course, if all of the countries in the index have similar rates of inflation, real and nominal indexes will tend to show the same rates of appreciation and depreciation. As an example, both the Federal Reserve Bank of Chicago and the Board of Governors compile a nominal and real dollar exchange index. From the first quarter of 1985 through the third quarter of 1986, the rate of depreciation of the dollar was 23.7 percent and 23.5 percent, respec- tively, for the real and nominal indexes of the Chicago Fed and 36.7 CRS-4 percent and 36.8 percent, respectively, for the real and nominal indexes of the Board of Governors. Having made the case that real indexes are preferable to nominal in- dexes for providing guidance on a number of national issues , various techni- cal problem related to computing real indexes should to be noted. The first actually applies to both types of indexes and involves the selection of exchange rates for inclusion in the index. It is comnon prac- tice in many countries to have multiple exchange rates, some of which may be unrelated to market conditions (e.g., an official rate fora host of trans- actions and a free or market rate for others). In others, rates may be con- trolled but, nevertheless, a black market rate is available. which rates should be used in making the computations? What type of adjustments should be made when countries switch from controlledtx>uncontrolled rates or vice versa? Second, and especially important for real indexes, is the choice of a price index to deflate the nominal magnitudes. In the United States, for example, a consumer price index, a producer price index, and a GNP deflator are available, as well as many subindexes, including one for tradable goods only. For other countries the choice of a price index does not arise because only one is available. Where multiple price indexes exist, however, a case can be made for the use of almost every one. Supposedly, the index selected should be comparable in coverage for each country. Special problems are present when countries impose wage and price controls while others do not. Controls can suppress reported inflation and distort the real exchange rate. An example of a special use of a price index is provided by the real dollar index computed by Morgan Guaranty Trust. In its index of 15 countries, weighed heavily by those in Europe and Japan, the nominal exchange rates are deflated by the prices of manufactured goods on the assumption that these are the principal components of trade between the countries in the index and the United States.. The prices of agricultural goods are excluded because such trade is thought to be primarily on an intra-European basis. In summary, even though real indexes are preferabletfinraddressing a rangecfifpolicy issues, for understanding some changes in GNP growth, and for forecasting purposes, they must be interpreted with care. As with any price index, there are problems associated with comparability of coverage across countries and within countries over time, with measurement error and with distortions introduced by government actions, especially wage and price controls. Are the Sample Sizes the Same? One of the noticeable differences across the various dollar exchange ratetindexes is the number of countries whose currency is included in the index. All of the indexes include the principal trading partners of the United States. The coverage, however, varies. The index compiled by the Board of Governors includes only ten U.S. trading partners. At the other lend oftjmzscale, the index compiled by the Federal Reserve Bank of Dallas is composed of all 131 nations that trade with the United States. Within this range there is considerable variation: the Federal Reserve Bank of CRS-5 Chicago has a 16-country index, the Federal Reserve Bank of Atlanta an 18- country index, and Morgan Guaranty Trust both a 15- and 40-country index. Sample size can be asource of difference in the reported change in the value of the dollar if the dollar does not move uniformly against all our- rencies and some are excluded from the sample. In such instances the re- ported depreciation or appreciation will either under or overstate its ac- tual change. As an example, the index compiled by the Board of Governors records a very high rate of depreciation of the dollar subsequent to the first quarter of 1985. This has been attributed'to the fact that the Board index contains mainly the currencies of European countries and Japan -- cur- rencies that have recorded the largest appreciation relative to the dollar. What Type of Weighting Pattern is Used? Computing a dollar exchange rate index involves computing an average value for the dollar relative to the currencies of other countries. A sim- ple average would accord equal weight to each country in the sample. This would be most unsatisfactory since a purpose of an index is to permit infer- ences about the probable consequences for the U.S. trade balance of a given appreciation or depreciation in the value of the dollar. For that reason, among others, exchange indexes are usually weighted averages, with the weights measuring the importance of each country's trade to the United States. The most common type of weighting system is one that weights by the fraction of U.S. trade accounted for by each country. As an example, assume that the U.S. trades with only three countries who account, respectively, for 75 percent, 20 percent, and 5 percent of U.S. trade (both imports and exports). In computing the average exchange value of the dollar the respec- tive exchange rates receive weights of .75, .20 and .05. This type of weighting pattern is known as bilateral. That is, the in- dividual exchange rates are weighted by the fraction of total U.S. trade ac- counted for by an individual country. There are a number of economists who either question the general usefulness of bilateral weights or who claim that at most they are useful primarily in measuring short run changes in the value of the dollar. Their criticism is focused on the fact that bilateral weights do not capture "third country" effects of exchange rate relation- ships. For example, both the United States and Japan might sell to Mexico and the yen-dollar relationship and changes therein might be important in determining some of these sales irrespective of the dollar-peso relation- ship. However, a bilateral weighting scheme will not pick up this indirect effect. Alternatively, a multilateral weighting scheme can be used. This dif- fers from the bilateral scheme in that the weight for each exchange rate is the sum of that country's exports to and imports from the other countries in the sample divided by total trade among all the countries in the sample. Thus, both the numerator and denominator of the weight is different than would be a corresponding bilateral weight. The major criticism of multilateral weights is that extensive trade be- tween the countries in the sample who do not simultaneously trade heavily CRS-6 with the United States will distort the weights of those countries who do trade extensively with the United States - because this non-U.S. trade in- creases the value of the denominator of the weight. Thus, important bilat- eral trade patterns are undervalued. It is also claimed that multilateral weights as they are typically constructed may not be especially efficient in picking up the "third country" effects allegedly omitted by bilateral weights. which weighting scheme should be used is not an unimportant matter. In table 1 the bilateral weights of the Atlanta Federal Reserve index are corn- pared to the multilateral weights used in the Board of Governors index. Note in particular the wide disparities given the weights of Canada, Japan, West Germany, France, Italy and the United Kingdom. Note also the almost exclusive reliance placed by the Board of Governors' index on U.S.-European trade even though during the past decade U.S. trade has shifted toward the Pacific Rim countries (countries included in the broader Atlanta index). TABLE 1. Weights of Nation's Currencies in Two Dollar Indexes (Percent) Atlanta Board of Country Fed Governors (Weighting Year) (1984) (1972-76) Canada 28.8 9.1 Australia 2.0 -- Japan 21.3 13.6 Belgium 2.2 6.4 France 3.7 13.1 West Germany 6.8 20.8 Italy 3.3 9.0 Netherlands 3.0 8.3 Spain 1.3 -- Sweden 1.3 4.2 Switzerland 1.5 3.6 United Kingdom 6.9 11 9 Saudi Arabia 2.4 -- Taiwan 5.0 -- Hong Kong I 3.0 -- South Korea 4.1 -- Singapore 2.0 -- China 1.6 -- Source: Federal Reserve Bank of Atlanta Table 1 also contains the years used for calculating the weights. Each index uses a base year weighting scheme. An additional choice, available to both bilateral and multilateral schemes, is to use current year weights. A sort of combination of the two is to use some type of moving average (fre- quently a 12-quarter moving average is calculated). Whether to use base year weights or current year weights is an important issue in the theory of index number construction and need not concern us. CRS-7 In summary, table 2 contains a comparison of the major dollar exchange rate indexes in use at the current time. ‘And in table 3, the appreciation of the dollar from l980:l through 198531, the period of its major rise, as well as its subsequent fall, is shown for each index. How much has the dollar fallen is, indeed, a question to which a range of answers can be given. ‘ CONCLUSION Dollar exchange rate indexes are in corrmon use in the United States and are helpful in deciding a number of policy related issues as well as under- standing movements in GNP, employment and inflation. Unfortunately, the in- dexes in use register markedly different rates of dollar appreciation and depreciation over the 1980-86 period. This report has sought to clarify why these differences exist. As such it distinguished real from nominal indexes, noted that alllindexes are not based on a uniform sample size and that the indexes used different weighting schemes (bilateral or multilateral), base year or current year trade patterns for calculating the weights, and different price indexes for deflating the nominal exchange rates. CRS-8 TABLE 2. Major Dollar Exchange Rate Indexes Number of Trade Weights Countries Base Real or Index in Sample Type Period Years Nominal Board of Gover- 10 Multilateral, March 1972- Both nominal nors of the Exports plus 1973 1976 and real are Federal Reserve imports I computed Chicago Federal 16 Bilateral, 1973: 12 Q Both nominal Reserve Bank exports plus 1Q Moving and real are imports Average calculated Atlanta Federal 18 Bilateral, 1980 1984 Nominal Reserve Bank exports plus imports Dallas Federal Reserve Bank: X-131 131 Bilateral, 1973: Moving exports plus IQ Annually Nominal imports X-101 101 Bilateral, 1973: Moving Real exports plus 1Q Annually imports Morgan Guaranty Trust: A 15 Bilateral, 1980- 1980 Both nominal trade in 82 and real are manufactures calculated B 40 Bilateral, 1980- 1980 Real trade in 82 manufacturers International 17 Multilateral, 1980 1977 Real Monetary Fund derived from econometric study Sources: the International Monetary Fund. Assorted publications of the Federal Reserve System and CRS-9 TABLE 3. Changes in the Value of the Dollar, 1980-1986 Percent Appreciation Depreciation of the Dollar of the Dollar Dollar Index 1980:1-198S:1 l98S:l-l986:4 Board of Governors 4 Federal Reserve System Nominal 79.1 31.6 Real 69.9 31.4 Federal Reserve Bank of Dallas X-131 (Nominal) 87.0 1.3 Federal Reserve Bank of Atlanta Nominal 37.0 19.1 Federal Reserve Bank of Chicago Nominal 40.7 21.4 Real 34.8 20.8 Morgan Guaranty Trust 15 countries (Real) 44.4 23.9 International Monetary Fund Real 58.7 25.7 Source: Assorted Publication of the Federal Reserve System, the International Monetary Fund and the Morgan Guaranty Company. irb/jw/rb/mls/rb LIBRARY OF WASHINGTON UNIVERSITY _ s1‘. Lows - MO.