A MONITORING CAPITAL TRANSACTIONS IN THE UNITED KINGDOM, GERMANY, AND JAPAN
The same underlying forces that have affected international capital transactions in the United States—and made information on such activities increasingly important in recent years—have come into play in the economies of the country's major trade partners. In this appendix we consider the reporting systems that three of those partners—the United Kingdom, Germany, and Japan—use to compile their statistics on international capital transactions and how they have sought to adapt them to take account of the new demands on those systems. Whenever possible, we note commonalities and differences with the U.S. system. Together with the United States, these countries account for approximately 80 percent of the world's capital flows.
THE U.K. SYSTEM1
There are two aspects of the U.K. system for monitoring international capital data that are particularly noteworthy. The first is that the U.K. balance-of-payments account is conceptually identical with its overseas-sector account within the national economic and financial accounts framework. The second is that, like the United States, the United Kingdom uses a survey system, generally under the purview of the Central Statistical Office (CSO),
with the Bank of England providing data mainly on banking and capital flow activity. The fact that surveys are conducted to obtain data for broader purposes than just the balance of payments limits costs and adds to the consistency of data across a range of economic statistics, such as the national accounts and the money supply. Because many statistics are collected for purposes other than balance-of-payments ones, however, the system can be slow to adapt to new developments or to adjust reporting requirements to capture the growth of derivative instruments. In addition, problems can arise when the ultimate uses of the data are not clear to filers.
Recently, the view has emerged that the quality of statistical reporting will be improved by the centralization of data compilation in one organization to provide a clear line of responsibility, namely, the CSO. This was the conclusion, for example, of the Pickford Report (Pickford, 1989), which examined the adequacy of government statistics. Its findings led to the transfer of many statistical functions from other government departments to the CSO. In 1990 Chancellor of the Exchequer John Major launched another initiative to improve economic statistics. Among areas that needed improvement were balance-of-payments data. A senior government statistician, Richard Eason, was asked to conduct a thorough investigation of the balance-of-payments compilation system. One theme of his report (Eason, 1991) was the need for centralization of data collection and compilation.
In the United Kingdom, the Bank of England had had a long tradition of producing data for the capital account—more than 60 years. In the 1930s the bank monitored U.K. banking and other short-term liabilities. In 1954 it became responsible for compiling the whole of the balance of payments, which was logical in view of the reliance on exchange control data. In 1960 responsibility for data compilation was transferred to the CSO because of the need to ensure integration of the balance of payments with the national accounts. However, the bank has continued to have a significant role in the production of data.
The Bank of England compiles data on new net issues of securities. This is an area in which the bank's expertise in international markets is of particular help to the CSO. The bank has long had an interest in the efficient functioning of the international financial system: for example, it has in recent years monitored and analyzed developments in Euromarkets, the international debt crisis, and the recycling of the surpluses of oil-exporting countries. The bank's International Division has created an extensive database of announcements of both securities and credits
in the international markets. The database on international bonds alone extends back to 1971 and has more than 26,000 individual records. It is one of the most comprehensive databases of its kind in the world, and data from it are supplied to both the Bank for International Settlements and the Federal Reserve Bank of New York. The bank obtains its data from market sources, and it partly cross-checks them with other sources, such as data from the International Securities Market Association. The database stores information on the residency of the issuer, the names of the lead managers, the interest rate or rates, and the amounts raised, as well as many other attributes. However, these data, which are from market sources, are not used for official statistical purposes, primarily because they concentrate on announcements of facilities and not on amounts drawn. Instead, the data on net new issues by U.K. corporate entities used in the balance of payments are compiled by CSO's Financial Statistics Division, drawing on the bank's expertise and knowledge.
Since the release of the Eason report (Eason, 1991) some of the Bank of England's compilation responsibilities, primarily for nonbank financial institutions, have been shifted to the CSO. The bank's role in the production of data is to be concentrated primarily on those markets and institutions in which the bank has a close policy, operational, or supervisory interest. This mandate covers principally banks, government accounts, and capital markets. In this manner, the bank's expertise in financial markets is still available to the CSO.
THE GERMAN SYSTEM2
Centralization is also a hallmark of the German reporting system, in which the Bundesbank and its branches are preeminent. At the same time, the German system in some respects resembles that of the United States. The German reporting system for international capital transactions, like the U.S. one, is based on two types of reports: the first is concerned with position statistics and treats the external position of banks and nonbanks, that is, claims on and liabilities to foreigners as well as direct investment relations; the second concerns flow statistics and deals with transactions or payments.
The German reports on the external claims and liabilities re-
semble those used in the United States, at least with regard to concepts and coverage of assets and liabilities. German nonbanks have to report their external financial and commercial claims and liabilities regularly. The number of reporters is much higher in Germany (about 13,000) than in the United States (about 400), however, and reports must be filed monthly in Germany, not quarterly as in the United States. German banks also have to submit comprehensive and detailed reports on their external assets and liabilities, also on a monthly basis. At the time of this writing, there were about 480 such filers in Germany, compared with close to 1,000 in the United States. Both German banks and nonbanks have to report their direct investment positions annually. The reports consist of major items from the affiliates' balance sheets.
The most important transactions or payments reports filed in Germany concern the purchase or sale of securities. German banks and nonbanks also have to report on all other long-term capital transactions.
All of the reports are filed with one agency, the Bundesbank, and its branches. Because of the legal requirement for such reporting, the use of simple forms, and the existing long-established reporting relationship, the rate of compliance and the quality of the data submitted are high. With such reports, the Bundesbank is able to provide a relatively detailed breakdown of capital transactions of nonresidents 3 to 4 weeks after the end of a given month. Data on direct investment are obtained from the reports of German parent companies or from reports of German subsidiaries of foreign companies on payments of capital. Data are collected on flows of equity capital and long-term intercompany debt. Reinvested earnings, which are shown under direct investment, are calculated from the balance sheets of resident and nonresident associated companies.
It is assumed that almost all direct investment flows are recorded in the German balance of payments, although perhaps after some delay. The Bundesbank has good reporting contacts with large enterprises. As in the United States, however, there may be considerable reporting gaps in the case of real property transactions, which fall under direct investment. Such nonreporting is a particularly acute problem in the case of the real estate investments of private individuals.
In contrast to U.S. practice, the portfolio transactions shown in the balance of payments are based almost exclusively on banks' payment reports. The banks report German residents' securities transactions with nonresidents by submitting special forms, by
providing copies of their securities settlement statements, or by communicating the required data on magnetic tapes. This method of data collection, based on transaction records, is commonly known as the ticket or transaction system.
The ''other capital" transactions are broken down between banks and nonbanks. For banks' long-term capital transactions, payment reports and stock figures from external position are available. With these two sources of data, banks' long-term capital transactions can be accurately monitored. For short-term capital flows, transactions shown in the balance of payments represent changes in claims on and liabilities to foreigners derived from the banks' external position reports. Because of good reporting ethics, a relatively low reporting threshold, and means of double-checking, German authorities are reasonably confident that they have good data on the international capital transactions of Germany's banking sector.
The long-term capital transactions of nonbanks are recorded by means of payment reports. In addition, stock reports on the long-term assets and liabilities (excluding securities) of nonbanks are available. As for the banks' reports, a comparison of the payment and stock reports is possible. This sector appears to be relatively well covered by the reports because the long-term capital transactions mainly involve large enterprises, which generally fulfill reporting obligations.
The short-term capital transactions of nonbanks are identified with the help of stock reports. Nonbanks have to report the level of their external assets and liabilities every month. These data lend themselves to detailed breakdowns of short-term capital transactions. Nonetheless, it is not always possible to differentiate the effects of exchange rate fluctuations and changes in asset prices.
THE JAPANESE SYSTEM3
Institutionally, the Japanese compilation system is also highly centralized. The role of the Ministry of Finance in Japan in many respects parallels that of Germany's Bundesbank and the U.K.'s Central Statistical Office. Moreover, the Japanese system, much like the German one, tends to compile information on all transactions, as opposed to the survey approach used by the United Kingdom and the United States.
The Japanese reporting system seeks to take account of all transactions between residents and nonresidents. It relies on estimation procedures only in a few limited cases for which complete reporting is not feasible. By most accounts, the Japanese system has been quite successful in ensuring accuracy, as evidenced by the very small numbers of errors and omissions in the data. This was particularly the case before 1988; since then, reported errors and omissions have increased (see below).
Underpinning the Japanese system of reporting all cross-border transactions are two principles. One is that all such transactions, or those involving the exchange of foreign currency, must be made through the accounts of banks designated by the Ministry of Finance. The other principle relates to the system of designated securities companies: the purchase and sale of foreign securities must be conducted through securities companies designated by the Ministry of Finance; for other transactions, individuals must submit a report on each transaction. Such arrangements ensure that, under most circumstances, authorities compiling data have information on all international transactions.
Every cross-border transaction that exceeds a certain amount (about $40,000) must be reported to the Bank of Japan, which has been designated by the Ministry of Finance as the sole compiler of balance-of-payments statistics. In practice, authorized foreign exchange banks serve as the agents of the Bank of Japan and submit appropriate remittance forms on behalf of their customers. The reports specify the amount and purpose: loan, direct investment, or other capital transactions. In addition, they cover invisible transactions, such as the receipt and payment of interest and dividends, grants, and contributions. Accordingly, all balance-of-payments items, with few exceptions, are theoretically covered by these remittance and customs reports. Authorized foreign exchange banks, for their part, are required to submit reports on the monthly movement of their external assets and liabilities.
For some balance-of-payments items, reports other than remittance ones are used; most notably for the compilation of portfolio investment data. Portfolio investment, as part of the long-term capital account, covers most transactions in foreign and domestic securities that have no contractual maturity or have a maturity of more than 1 year.
For outward portfolio investments, the most important sources of data are the reports of designated securities companies, which are required to file monthly statements on the purchase and sale of foreign securities both for their own accounts and those of
their customers. There were more than 100 such securities companies when this report was being written. Like the designated securities companies, some financial institutions are required to file similar monthly portfolio investment reports. In addition, anyone who buys foreign securities (over a specified minimum) directly from overseas intermediaries must submit notification describing the details of each such transaction. Issuers of foreign bonds in Japan are required to submit certain reports to the minister of finance through the Bank of Japan.
Inward portfolio investments, which involve investments by nonresidents in Japanese securities, are classified into stocks, bonds, and overseas bond issues by residents. Data regarding stocks and bonds are collected on a settlement basis from designated securities companies, authorized foreign exchange banks, and individual transactors. Data for bond issues by residents are obtained through special reports called Notification Concerning Issue or Flotation of Securities and Report on Invisible Trade Payment.
The valuation method used is relatively simple. Securities transactions denominated in currencies other than U.S. dollars are converted to dollars using basic exchange rates set by the Ministry of Finance. Outward investments in stocks and bonds are geographically allocated into ten major markets: the United States, the United Kingdom, Germany, France, Luxembourg, the Netherlands, Canada, Switzerland, Australia, and "others." Allocation is made according to the location of the market in which the transaction occurred. Inward investments are allocated by the domicile of the transactor. Accordingly, they can be allocated to individual countries, but they are aggregated into the same geographical breakdown as outward investments. Stock data are compiled from flow data. As a result, most items are recorded at transaction value rather than at market value. An exception is made for Japanese equities held by nonresidents: these values are adjusted periodically to reflect market levels. In addition, adjustments are made in the case of yen-denominated bonds to reflect changes in exchange rates.
Despite the strengths of the U.K., German, and Japanese reporting systems, certain inadequacies in them were becoming increasingly apparent by the late 1980s, as they were in the United States.
In the United Kingdom, for example, it became clear that greater resources had to be devoted to adjust to the rapidly changing do-
mestic and international financial environment. In 1979 U.K. exchange controls were completely removed; this change, together with deregulation of domestic financial markets, led to a surge in financial transactions. Capital inflows and outflows to the United Kingdom were some 14 times larger in the decade after the lifting of exchange controls than before. Errors and omissions also rose: for the 3 years before 1978, errors and omissions in the balance-of-payments accounts were about $4 billion a year; in 1988, they were approximately $18 billion.
A second problem discovered was incomplete reporting by the corporate sector. As in the United States, it was determined that more comprehensive coverage was needed of transactions involving nonbank financial institutions and industrial and commercial companies: considerable data clearly were not being collected on overseas transactions of the corporate sector.
A third shortcoming of the U.K. system that became evident was that there was inadequate data on the financial transactions of the household sector. Although some data on outward portfolio investment could be derived from reports by intermediaries, it was generally seen as unsatisfactory. At the same time, inward portfolio investment was identified as a particularly weak part of the accounts. During the past decade, the United Kingdom has generally had positive errors and omissions, implying either unrecorded credits in the current account, such as investment income, or unrecorded net capital inflows. The Bank of England concluded that it seemed likely the bulk of the errors and omissions were portfolio flows (Bank of England, 1990).
Similarly, in the German case, reporting gaps became evident in the late 1980s when German private individuals increasingly started to avail themselves of foreign banking services, which led to various kinds of errors. For example, if a German investor bought a German government bond through a Luxembourg bank, the transaction then appeared in the German balance-of-payments statistics as a purchase of German securities by a nonresident. As in the United Kingdom, Germany's errors and omissions in the balance of payments had significantly increased by the end of the 1980s. In theory, the German reporting system should have been able to close such gaps. In the example above, the private German investor should have reported the purchase of German or foreign securities through the Luxembourg bank. In practice, however, either deliberately or through ignorance, the reporting requirement often is not met. The result of purchases of deutsche mark securities by German investors through nonresidents, therefore, is an overestimation of Germany's external liabilities or, in the
case of purchases of foreign securities, an underestimation of the stocks of foreign securities held in Germany.
Japanese financial observers have encountered similar difficulties in recent years. The Japanese system was quite successful in ensuring the accuracy of statistics until recent years, as evidenced by the very small errors and omissions in their data up to 1988. Since then, however, problems have arisen with the capital account data. Errors and omissions increased dramatically in 1989 and 1990, reflecting the growing proportion of international transactions conducted outside the reporting system. Clearly, the increased scope and sophistication of the activities of Japan's financial institutions have placed a substantial burden on the reporting system. At the same time, these developments have made it necessary for compilers to obtain more information directly from nonfinancial private institutions. In addition, Japan is one of the industrial countries that does not include reinvested earnings from direct investment in investment income.
Like their counterparts in the United States, compilers in the United Kingdom, Germany, and Japan have also had to deal with the difficult problems of valuation and geographic allocation. When exchange rates fluctuate, the exchange rate used to convert foreign currencies to national ones for statistical purposes can differ substantially from the rate that prevailed at the time of the transaction, causing discrepancies in statistics. Similarly, allocating transactions geographically has become problematic.
The United Kingdom, Germany, and Japan all have begun to adjust their reporting systems in an attempt to reverse the perceived decline in the quality of data and ensure that the systems keep pace with the rapidly evolving international financial markets.
In the United Kingdom, such efforts have centered on major reviews of the adequacy of the U.K. balance-of-payments data and, when possible, corrective actions. Impetus for such reviews had come from a combination of rising errors and omissions, on one hand, and a growing current account deficit, on the other. U.K. efforts to improve the capital account data have, pursuant to the recommendations of the Eason report (Eason, 1991) led to a greater centralization of data collection. In addition, steps have been taken to deal with incomplete reporting by the corporate sector. The result has been a more comprehensive coverage of nonbank financial institutions and commercial companies. This has been accomplished by collecting more data on overseas transactions directly from the corporate sector and by reducing reliance on data from intermediaries. For example, a new survey of securities
was conducted in 1989. It was followed by an expansion of coverage of the financial assets and liabilities of nonfinancial institutions at the end of 1990 and an exploratory survey of fund managers in 1991. Consideration also has been given by U.K. authorities to improving data on transactions in the household sector. The proposal is to survey individuals to learn more about their overseas transactions in securities. Another proposal under consideration is to assess these transactions through surveys of fund managers' private client business. Since the measures have been undertaken, the net errors and omissions have decreased.
For data on inward portfolio investment—that is, data on foreign investment in U.K. equities—improvements were made as a result of the introduction of a share register survey at the end of 1989. This survey, carried out on behalf of the CSO by a private company, showed that the overseas sector held about $33 billion more equity than had previously been estimated from data on transactions provided by financial intermediaries. In early 1990, the method of compiling data on foreign investment in U.K. company bonds was changed from a reliance on survey data from intermediaries to calculating net overseas investment as a residual from total net new issues by U.K. companies obtained by the Bank of England, primarily from market sources and net acquisitions reported by domestic sectors.
German compilers likewise took steps to adjust to the changes in international capital markets that occurred during the 1980s. They sought to deal with the increasing volume of securities transactions, especially those involving nonresidents, by relying more heavily on electronic data processing. Difficulties related to the recording of data on new financial instruments and related innovations was dealt with by seeking closer cooperation with banks. The German compilers believe that major reporting gaps did not arise from the internationalization of institutional investors' business because the number of such institutions in Germany is generally limited, and they are generally sympathetic to statistical reporting requirements. The German compilers also believe that data gaps attributed to the growing international diversification of private individuals' portfolios can be improved as long as these transactions are handled and reported by the German banking system.
In Japan, efforts are also under way to adjust to the new reporting environment. Now under consideration are proposals to make more frequent use of sample estimations of certain balance-of-payments items, such as international investment position and reinvested earnings.