As is customary each June, the estimates of U.S. international transactions have been revised to incorporate methodological and statistical revisions. This year, like last year, a number of improvements have been implemented as part of continuing efforts by the Bureau of Economic Analysis (BEA) to address gaps in coverage of transactions. The gaps and the plans to fill them were outlined by BEA in its Strategic Plan for improving BEA's economic accounts (see the February and April 1995 and June 1996 issues of the SURVEY OF CURRENT BUSINESS). The improvements also address various gaps noted by the International Monetary Fund and the National Academy of Sciences./1/ In large part, the gaps have arisen because of the dynamic nature of international markets. The major improvements this year respond to rapid changes in both the capital markets and the services markets.
The newly available benchmark data, improved methodologies, and several reclassifications are discussed in detail in the remaining sections of this article. In addition to these major revisions, revisions result from the incorporation of regularly available data from BEA's annual and quarterly surveys, from the U.S. Treasury Department's quarterly and monthly surveys, and supplemental data from other U.S. Government agencies and private sources. Among the estimates most affected by the regular updating of data are travel and passenger fares, which incorporate updated data from the U.S. Department of Commerce's Office of Tourism Industry's In-Flight Survey; transfers under U.S. military sales contracts, which include more complete coverage; and private remittances, which also include more complete coverage. Revisions are also made to the outward and inward direct investment estimates to incorporate revised or new BEA quarterly survey results and to the services estimates to incorporate new BEA annual survey results. In most cases, quarterly patterns did not change much. However, the quarterly pattern for travel and passenger fares exports in the last half of 1996 was changed significantly. The quarterly pattern for goods exports was changed as a result of the introduction of a new approach to seasonal adjustment.
Tables 1 and 2 in this article present a summary of revisions from all sources. Table 3 presents detail on the revisions due to new benchmark data, methodologies, and reclassifications. For 1997, $1.0 billion is added to goods exports, reducing the goods deficit by the same amount. For services, $5.0 billion is added to services receipts and $2.6 billion is added to services payments, resulting in an upward revision of $2.5 billion to the services surplus. With the exception of the addition to transfers under U.S. military sales contracts, other revisions to the services accounts were about offsetting. In the income accounts, $5.7 billion is added to income receipts and $3.2 billion is subtracted from income payments, resulting in a $9.0 billion reduction in the deficit on investment income. Net unilateral transfers were revised to reflect $1.2 billion more outflows. As a result of these revisions, the current-account deficit for 1997 is revised down $11.2 billion.
In tables 1 and 10 in the standard presentation of the U.S. international accounts, net U.S. currency flows are now shown separately in line 59; previously, they had been combined with private transactions in U.S. Treasury securities in line 58. In table 3, the number of components of the "other" transportation estimates has been reduced; this change is described in the section on transportation. In table 7, three lines have been added to accommodate the reclassification of transactions of financial intermediaries. In table 9, net U.S. currency flows are removed, since they are now shown separately in table 1.
Dividend payments are revised to incorporate results of the U.S. Treasury Department's benchmark survey of foreign portfolio investment in the United States for 1994. Dividend payments on U.S. stocks are revised up $752 million for 1994 because of a higher position reported in the survey ($398 billion) than was previously estimated ($369 billion). Dividend payments are recomputed to reflect the higher outstanding positions. Differences in country and area totals are also adjusted to the new position estimates. An examination of the industry distribution of foreign holdings confirmed the continued use of the Standard and Poor's composite index of 500 stocks for estimating both the pricing of foreign stock holdings and the dividend yields applied to those holdings. Revisions are carried forward through 1997 and back to 1989, the last benchmark year.
Interest payments are revised to incorporate results of the U.S. Treasury Department's benchmark survey of foreign portfolio investment in the United States for 1994. Interest payments on U.S. corporate bonds are revised down $4.7 billion for 1994 because of a lower position reported in the survey ($276 billion) than was previously estimated ($304 billion) and because of the introduction of lower yields into the interest computation. Revisions are carried forward through 1997 and back to 1989, the last benchmark year.
The benchmark survey data include the amount outstanding, maturity, and yield on each Eurobond issue. Eurobonds, which are bearer instruments for which it is difficult to determine the nationality of ownership, account for over half of total U.S. bonds held by foreigners. The new details provide a more accurate measure of total Eurobonds outstanding and permit an application of interest yields that can be more closely tailored to the market characteristics of the Eurobond holdings.
Newly introduced bond indexes provide a much closer matching of interest rates to the market sectorsdomestic corporate bonds, corporate Eurobonds denominated in dollars, and corporate Eurobonds denominated in foreign currenciesthan in the previous methodology. The new bond indexes also permit a more frequent revaluation of positions and computation of income estimates and therefore result in income estimates that are more sensitive to fluctuations in current market interest rates. Positions are now revalued quarterly; previously, they were revalued only semiannually. When the more detailed prices and yields by market sector and the more frequent revaluations for 1990 to 1994 are applied, the difference between the benchmark survey results for 1994 and the new method is considerably less than the comparable difference under the previous methodology.
Interest payments on U.S. Government securities are revised to incorporate the results of the U.S. Treasury Department's benchmark survey of foreign investment in long-term securities in the United States for 1994 and a new methodology is introduced for estimating interest payments on U.S. Government agency securities. Interest payments on U.S. Government securities are revised down $2.8 billion in 1994, as a $4.7 billion downward revision from a lower estimate of foreign holdings from the survey is partly offset by a $1.8 billion upward revision from a new methodology for estimating interest payments on agency securities.
The benchmark survey shows that foreign holdings of marketable long-term U.S. Government securities were $571.1 billion at the end of 1994, $53.8 billion less than the previous estimate of $624.9 billion. Revisions to U.S. Government interest payments based on the lower positions in the survey are made by country or geographic area and are carried forward through 1997 and back to 1989, the year of the previous benchmark.
A new methodology for estimating interest payments on U.S. Government agency securities applies the market yield on mortgage-backed securities to the outstanding position adjusted to market value. Previously, a mortgage-backed security coupon yield, averaged with a Eurodollar deposit rate, had been applied to positions at face value. The new methodology more closely reflects current market conditions and provides a more accurate estimate of interest payments. The revisions to interest payments caused by the change in estimation methodology for agency securities are made at the country or geographic area level and are carried forward through 1997 and back to 1989. In earlier periods, the new methodology has a negligible effect on the income estimate.
Interest receipts on U.S. holdings of foreign bonds for 199397 are revised to incorporate the final results of the U.S. Treasury Department's benchmark survey of U.S. portfolio investment abroad for 1994. Data on maturities and applicable yields were not available at the time the other results from this benchmark survey were introduced last year. Except for revisions due to the routine updates of data, the position estimates remain unchanged from those introduced in the July 1997 SURVEY.
Results of the survey on the geographic, currency, and sector composition of U.S. holdings led to a lowering of the yields applied to outstanding positions and provided the basis for a more detailed application of interest yields by market sector than was previously possible. The survey results also provided additional data on the maturity structure of these holdings, leading to the introduction of shorter term, and therefore lower, yields for estimating interest receipts. Because the pattern of the newly implemented yields converged with the previous yields for 1997, the revision to interest receipts for 1997 was negligible; annual revisions for 199396 were larger.
There are three sources of revision to the 199497 net capital outflows estimates for U.S. direct investment abroad. First, the estimates are revised to incorporate the results of the 1994 benchmark survey and new and corrected data from the sample surveys for 199597./2/ Second, the revisions also reflect revised estimates of depreciation, depletion, and expensed exploration and development costs, which are used to adjust the reinvested earnings component of fixed capital to a current-cost basis. These two sources of revision result in upward revisions to net capital outflows of $4.7 billion for 1994, $7.9 billion for 1995, $1.7 billion for 1996, and $13.8 billion for 1997. The third source of revision to the 199497 net capital outflows was the reclassification of intercompany debt transactions with foreign affiliates, other than depository institutions, that are financial intermediaries. Previously, these transactions were considered to be direct investment transactions; now, they are considered as nonbank investment transactions. For a more complete explanation of the reclassification, see "Reclassification of intercompany debt and associated interest transactions with financial intermediaries."
Net receipts of income by U.S. parents from their foreign affiliates are revised for 199497 to incorporate the results of the 1994 benchmark survey and new and corrected data from the sample surveys for 199597. The revisions also reflect revised estimates of depreciation, depletion, and expensed exploration and development costs, which are used to adjust the earnings component of income to a current-cost basis, and revisions to related withholding taxes. These sources of revision result in upward revisions to net income receipts of $2.1 billion for 1994, $3.9 billion for 1995, $2.4 billion for 1996, and $2.3 billion for 1997. In addition, the revisions reflect the reclassification of receipts and payments of the interest component of income associated with intercompany debt transactions with foreign affiliates, other than depository institutions, that are financial intermediaries. Previously, these receipts and payments were considered to be direct investment income transactions; now, they are considered as nonbank investment income transactions. For a more complete explanation of the reclassification, see "Reclassification of intercompany debt and associated interest transactions with financial intermediaries."
Receipts and payments of royalties and license fees between U.S. parents and their foreign affiliates are revised for 199497 to incorporate the results of the 1994 benchmark survey and new and corrected data from the quarterly sample surveys for 199597. U.S. parents' receipts are revised up $2.5 billion for 1994, $1.2 billion for 1995, $0.9 billion for 1996, and $1.5 billion for 1997. U.S. parents' payments are revised up $0.2 billion for 1994, $0.1 billion for 1995, $0.2 billion for 1996, and $0.2 billion for 1997.
Receipts and payments for "other" private services between U.S. parents and their foreign affiliates are revised for 199497 to incorporate the results of the 1994 benchmark survey and new and corrected data from the quarterly sample surveys for 199597. U.S. parents' receipts are revised up $1.4 billion for 1994, $0.5 billion for 1995, $1.0 billion for 1996, and $1.2 billion for 1997. U.S. parents' payments are revised up $1.1 billion for 1994, $0.6 billion for 1995, $0.6 billion for 1996, and $0.8 billion for 1997.
Beginning with 1994, intercompany debt transactions between parent companies and affiliates that are not depository institutions and that are primarily engaged in financial intermediation are reclassified from the direct investment capital accounts to the nonbank investment accounts, where they are combined with other capital transactions between U.S. nonbanking concerns and unaffiliated foreigners. Similarly, interest receipts and payments associated with such intercompany transactions are reclassified from the direct investment income accounts to the "other" private income accounts. Although these transactions are between affiliated firms, in many ways, they are similar to the financial flows that are classified in the nonbank investment accounts. This treatment is similar to the treatment of nonpermanent debt investment and associated income transactions between affiliated depository institutions, and it is consistent with the guidelines in the International Monetary Fund's Balance of Payments Manual (fifth edition), which suggest that these transactions of financial intermediaries are to be excluded from the direct investment accounts./3/ Equity capital transactions with these financial intermediaries, and the associated earnings receipts and payments, continue to be classified as direct investment.
For U.S. direct investment abroad, intercompany debt transactions and associated interest transactions between U.S. parents and the following three groups of nonbank foreign financial affiliates are reclassified: (1) Financial affiliates located in the Netherlands Antilles, (2) financial affiliates whose U.S. parents are depository institutions, and (3) financial affiliates whose U.S. parents are securities dealers.
For foreign direct investment in the United States, intercompany debt transactions and associated interest transactions between foreign parents (and foreign affiliates of foreign parents) and the following two groups of U.S. nonbank financial affiliates are reclassified: (1) Financial affiliates whose ultimate beneficial owner (UBO)/4/ is a foreign depository institution, and (2) financial affiliates whose UBO is a finance or insurance firm./5/
In the process of reclassification, U.S. intercompany debt receivables from both U.S. direct investment abroad and foreign direct investment in the United States are added to claims reported by U.S. nonbanking concerns, and the associated interest receipts from both U.S. direct investment abroad and foreign direct investment in the United States are added to "other" private income receipts. U.S. intercompany debt payables from both U.S. direct investment abroad and foreign direct investment in the United States are added to liabilities reported by U.S. nonbanking concerns, and the associated interest payments from both U.S. direct investment abroad and foreign direct investment in the United States are added to "other" private income payments.
Also in the process of reclassification, certain transactions were discovered to have been counted twice in the compilation of the accounts. This duplication occurred because transactions of U.S. affiliates of foreign depository institutions reporting to BEA on its survey of foreign direct investment in the United States were also covered in the depository institutions' reporting to the Bank for International Settlements and to various foreign central banks, which provide the basis for BEA's estimates of nonbank transactions. Therefore, where duplication could be detected, not all of the capital and associated interest transactions that were removed from the foreign direct investment in the United States accounts were added to the nonbank investment and income accounts. For the capital accounts, the net amount of duplication was relatively small for 199496, but large for 1997. For the income accounts, the amount of duplication was sizable in 1996 and 1997.
Estimates of institutional remittances to foreign residents by charitable, religious, educational, and philanthropic organizations in the United States are raised considerably as a result of improved survey coverage and the use of supplemental annual data from government and selected private philanthropic organizations. These improvements increase the estimate of institutional remittances $0.8 billion for 1997.
For the past 5 years, BEA has estimated U.S. banks' income receipts on foreign-currency-denominated claims and U.S. banks' income payments on foreign-currency-denominated liabilities separately from banks' income receipts and payments on dollar-denominated claims and liabilities. This change became necessary as the size of foreign-currency-denominated claims and liabilities grew, partly in response to the growth of international transactions themselves and partly in response to shifts in preferences by U.S. residents or foreign residents as to the proportions of their bank claims and liabilities that they wished to hold in foreign currencies or dollars to conduct international transactions.
With this year's revisions, BEA has introduced three improvements to its estimates of banks' foreign currency income for 199297. First, BEA now uses foreign counterpart data from the Bank for International Settlements (BIS) that provide more accurate information on the currency composition of U.S. banks' foreign currency claims and liabilities outstanding. Second, these BIS data provide detail across a larger number of currencies than was available previously. Third, for purposes of computing income, currency translation effects on balances reported in dollars are removed by converting these balances into foreign currencies at the appropriate end-of-quarter rates.
The new approach is based on data collected by the BIS, which is the same data source used by BEA to estimate unrealized gains and losses on foreign currency banking transactions and remove them from the capital flow data./6/ The BIS data show, from the perspective of foreign resident banks, foreign banks' nondollar claims on U.S. banks and foreign banks' nondollar liabilities to U.S. banksthe counterparts of U.S. banks' foreign currency liabilities and U.S. banks' foreign currency claims, respectively. The BIS data on the composition of U.S. banks' foreign currency claims and liabilities are available quarterly.
The percentage composition of foreign currency balances in the counterpart BIS data is applied to the balances reported in the Treasury International Capital (TIC) reporting system, which provides no currency-composition detail, at the end of each quarter to compute the dollar amount of TIC outstanding balances held in each of nine key currenciesthe British pound, Japanese yen, German mark, Swiss franc, French franc, Italian lire, Canadian dollar, the European Currency Unit (ECU), and Special Drawing Rights (SDR's) (for less developed countries that transact largely in nondomestic currencies). For each currency, dollar reported balances at the end of the current and previous quarters are converted into foreign currencies using end-of-quarter exchange rates; then, average foreign currency balances for the quarter are computed.
Monthly interest yields (averaged to a quarterly rate) are prepared by combining local overnight call money rates and 3-month rates for each currency; overnight Eurodollar deposit rates are used for SDR's. The interest yields for each currency are multiplied by average foreign currency balances outstanding for each currency to calculate quarterly income, and the results are converted into dollars using quarterly average exchange rates.
The result of these changes is to raise both U.S. banks' income receipts and U.S. banks' income payments in comparison with the previous estimates for 199297. For 1997, U.S. banks' foreign currency income receipts are raised $0.8 billion, to $2.9 billion, and U.S. banks' foreign currency income payments are raised $1.8 billion, to $4.1 billion.
Computer software royalties and license fees are reclassified to royalties and license fees from "other" private services for 199297. The purpose of the reclassification is to better reflect the acquisitions and sales of rights to use or reproduce computer software as transactions involving intangible, nonproduced, nonfinancial assets and proprietary rights (such as patents, copyrights, trademarks, industrial processes, and franchises) and to combine them with other such transactions. For 1997, the amount reclassified was $2.4 billion for receipts and $0.5 billion for payments.
The operational leasing of transportation equipment without crew is reclassified from the "other" transportation accounts to the "other" private services accounts to consolidate all operational leasing in one account for 198697. In addition, coverage of equipment leasing is now more complete on BEA's Annual Survey of Selected Services. In 1997, the total of the reclassified and the newly collected transactions is $1.4 billion for receipts and $0.4 billion for payments.
Operational leasing of transportation equipment with crew remains in the "other" transportation accounts, because these transactions are closer in nature to the provision of transportation services than to the rental of equipment.
Estimates of freight charges for the transportation of U.S. goods exports by truck between the United States and Canada are revised for 199297. Estimates by Statistics Canada provide the basis for key components of the truck transportation estimates in the U.S. accounts. A new analysis by Statistics Canada indicates an overstatement of freight charges on the transportation by Canadian truckers of U.S. exports from points of origin within the United States to the U.S.-Canadian border (U.S. payments) and on the transportation by U.S. truckers of U.S. exports from the U.S.-Canadian border to points of destination within Canada (U.S. receipts). The new analysis suggests an implied average freight-to-value ratio of about 1.8 percent, compared with an average of 3.5 percent suggested by earlier analyses. Consequently, for 1997, freight receipts are reduced $0.5 billion and freight payments are reduced $0.7 billion.
Concurrent with the reclassification of operational leasing without crew from the "other" transportation account to the "other" private services accounts, operational leasing of transportation equipment with crew has been shifted to the freight component of the "other" transportation accounts. Thus, "other" transportation receipts and payments each now have only two subcomponentsfreight services and port servicesin table 3 of the standard presentation of the international accounts, rather than the three shown previously.
BEA and the Bureau of the Census seasonally adjust the goods export and goods import estimates at the five-digit end-use commodity category level, which is the most detailed level of end-use classification available. Aggregate goods seriestotal exports, total imports, and all major end-use categoriesare derived as the sum of detailed seasonally adjusted series. An alternative set of aggregate series can be derived by directly seasonally adjusting each of the aggregate series. Comparisons of the directly adjusted series with the corresponding series that are derived as the sum of individually seasonally adjusted series show differences that are sometimes called "residual" seasonality. The amount of "residual" seasonality is usually small, so no adjustments are made to the aggregate series. However, in recent years, the amount of "residual" seasonality for total exports and for capital goods exports has increased. Consequently, a concerted effort was made last year and this year to reduce the "residual" seasonality for goods exports. Little "residual" seasonality exists for goods imports.
Last year, some progress was made in reducing the amount of "residual" seasonality by combining several of the individual machinery export series into a single category and by developing seasonal factors based on that category, rather than on each component series separately. This year, the addition of trading-day factors to many export series and changes made to many trading-day groups led to a further reduction in "residual" seasonality.
In addition, this year, the Bureau of the Census and BEA are introducing monthly seasonal adjustment for exports of civilian aircraft. This adjustment reduces the amount of "residual" seasonality and improves the reliability of the quarterly international transactions accounts and gross domestic product estimates. For 1997, the adjustment raised seasonally adjusted total exports significantly in the first and third quarters and lowered seasonally adjusted total exports significantly in the second and fourth quarters, relative to the previously published estimates.
The monthly pattern of exports of civilian aircraft exhibits high variability and was not previously adjusted, because it did not meet all of the usual criteria for seasonal adjustment of individual series. This year, however, after considering the contribution of adjusting civilian aircraft to improving the accuracy of seasonally adjusted total exports, a decision was made to adjust this series. At this time next year, as part of the regular review of all goods series, BEA and the Census Bureau will assess the effects of seasonally adjusting exports of civilian aircraft on the monthly and quarterly goods statistics to determine whether to continue the practice.
For many years, a part of imports of prepackaged computer software has been valued at media valuethat is, the value of the carrier medium itself, such as CDROM or disketterather than at market value, which reflects both the value of the carrier medium and the data or instructions included on the medium. This valuation practice is consistent with a General Agreement on Tariffs and Trade decision in 1984 that permits countries to value imports of prepackaged computer software for customs purposes at either media value or market value.
BEA has prepared an estimate of the dollar amount that is necessary to bring computer software imports reported at media value to the full market value required for the international transactions accounts and the national income and product accounts. This amountwhich was small for 1995, 1996, and 1997will be added as a balance-of-payments adjustment to the Census basis data reported to BEA. A comparable adjustment to exports of prepackaged software is not required, because exports of prepackaged computer software are already reported to the Census Bureau at market value.
Table 3, page 1
Table 3, page 2
1. See Report on the Measurement of International Capital Flows (Washington, DC: International Monetary Fund, September 1992); Behind the Numbers: U.S. Trade in the World Economy (Washington, DC: National Research Council, 1994); and Following the Money: U.S. Finance in the World Economy (Washington, DC: National Research Council, 1995).
2. A more detailed explanation of the benchmark revisions will appear in the October 1998 SURVEY. See also U.S. Department of Commerce, Bureau of Economic Analysis, U.S. Direct Investment Abroad: 1994 Benchmark Survey, Final Results (Washington DC: U.S. Government Printing Office, May 1998).
3. The U.S. international transactions accounts are consistent in all major aspects with the principles set forth in the Balance of Payments Manual (fifth edition).
4. The UBO is that person, proceeding up a U.S. affiliate's ownership chain, beginning with and including the foreign parent, that is not owned more than 50 percent by another person.
5. Ideally, the reclassification should not have been extended to affiliates of UBO's that are insurance companies. However, the industry of UBO for both foreign finance and insurance firms is reported under a single industry code, so it is not possible to exclude these affiliates from the reclassification. However, most of the UBO's appear to be in finance rather than in insurance.
6. For more information see, Christopher L. Bach, "U.S. International Transactions, Revised Estimates for 197496," SURVEY 77 (July 1997): 53.