« PreviousContinue »
date, the value of foreign investments in the U.S. stocks included in the Standard and Poor's list amounted to $17.0 billion. Foreigners also held investments valued at $7.7 billion in U.S. stocks not included in the Standard and Poor's list. Conversely, the Standard and Poor's list included stocks of 14 foreign corporations, which were, of course, not covered by the survey. In addition, there were numerous differences in the industry distribution of the aggregate value of foreign-held issues, as compared with the industry distribution of the value of outstanding shares of the Standard and Poor's 500 common stocks. The substantial foreign investment in U.S. banks and insurance companies, categories which are not included in the Standard and Poor's list, is considered an important difference. Table 25 summarizes this information.
The differences between the U.S.stocks held by foreigners and the 500 U.S. and foreign stocks covered by the Standard and Poor's index, and the resulting differences in weighting as between the value of foreign holdings and the value of outstanding shares, could result in erroneous valuation adjustments for foreign-held stocks. The comparative data cited above for the end of 1974 suggest that the estimated valuation changes are unlikely to be precise. Thus the valuation adjustments may be the source of a substantial part of the difference between the previous estimates of foreign holdings of U.S. stocks and the data derived from the new survey, although, of course, there is no way of knowing whether the use of the Standard and Poor's index would overstate or understate the value of the foreign portfolio at any given time.
Three major problems are encountered in the report. ing of securities transactions and other current data that are not present to the same degree in the comprehensive benchmark data. One is whether coverage is complete: that is, whether reports are obtained from all of the firms which should be reporting. The second is whether the reports filed by the reporting firms are correct and complete. The third is whether it is possible for reporting firms in every case to know whether a transaction is effected for the account of a foreigner.
The survey questionnaire was directed to the prin. cipal firms in the United States which maintain records of foreign ownership of U.S. stocks, bonds and other long-term obligations—the issuing corporations and the holders of record. It was also directed to Federal, State and local governments and agencies which had issued long-term securities. Except for the effect of the small reporting exemption and the problem involved in foreign holdings of bearer securities, the inventory should in principle be virtually complete, since it covers all significant issuing corporations and Government agencies and all significant holders of record as of December 31, 1974, which could be identified from mailing lists developed from industry and tax sources.
In the current reporting system, it is more difficult to identify all of the significant firms which may effect transactions in long-term securities with foreigners or may have long-term liabilities to foreigners. Many different types of firms are involved, and in many ins. tances, several departments of a given firm. There is no readily available way of uniquely identifying firms that engage in securities transactions with foreigners, or of identifying firms that enter or leave the business of conducting securities transactions with foreigners. There are similar problems in identifying on a current basis nonbanking firms which may have long-term liabilities to foreigners. In short, it is intrinsically more difficult to identify a changing and evolving reporting universe in a continuing reporting operation than to identify firms
Current Reporting Problems For several reasons it is more difficult to obtain complete current reporting of securities transactions with foreigners and long-term liabilities to foreigners than to obtain a complete inventory of foreign portfolio holdings as of a single date through a comprehensive survey.
Table 23.—Comparison of Foreign Holdings of U.S. Stocks with Common Stocks in Standard and Poor's Composite Price Index By Industry." as of
December 31, 1974
Market Value of
of ($ niillion)
CS Stock Held by Foreigner
of ($ million)
100.0 Agriculture, forestry, & fishing .
59.8 Transportation & public utilities.
10.8 Wholesale trade
1.2 Retail trade ..
3.1 Finance, insurance & real estate
11 Federal Government, State & local governments.
38 Industry classification as reported in Foreign Portfolio Investment Survev, except for 11 foreign companies included in Standard and Poor's list on basis of industry classification assigned by Stanglard and Poor's.
Standard and Poor's list includes stock of 14 foreign companies with an aggregate market value of $14.369 million
that held securities and other assets for foreigners at a single date through a comprehensive survey. The best that can be done is to make continuous efforts, to the extent resources are available, to communicate with prospective reporters to ensure that reporting coverage is reasonably complete—that is, that firms which have reportable transactions and long-term liabilities are, in fact, reporting.
The diversity of potential reporters and the impracticability of maintaining continously the level of publicity and communication that attended the benchmark survey, suggest that the coverage of current reporting is likely to be less complete than that of the benchmark survey. While the firms currently reporting transactions in long-term securities with foreigners include all of the large securities firms in the United States, not all securities firms are in the reporting universe. The reporting exemption excludes many small firms, and some larger firms may not be aware of the reporting requirements. It is not certain that all banks that act directly by order of foreigners in transactions in long-term securities in the United States are reporting them, although the coverage of reporting by banks of their long-term liabilities to foreigners, such as longterm certificates of deposit, is considered to be satisfactory. Maintaining adequate coverage of nonbanking firms with respect to their long-term liabilities to foreigners involves a number of difficulties, and requires periodic general canvasses of the business community to publicize the reporting requirements.
It is also more difficult to obtain complete and accurate reporting on the current reports than on the benchmark survey. The survey was a count of securities and other long-term assets owned by or held for foreigners on a single date. While it contained complexities and required the reporting of a large mass of data, the concept was relatively straightforward and considerable time was available to answer questions and correct errors. The work of ensuring the correctness and completeness of current reporting, on the other hand, requires both a continuous review and questioning of the monthly and quarterly reports on a short time schedule, and periodic comprehensive discussions with reporting firms to scrutinize the contents of their reports and the adequacy of their reporting procedures. The reporting of current transactions in long-term securities requires the monthly cumulation by the reporting firms of data on every transaction for foreign account, a procedure which involves greater possibilities of error and omission than the determination of the amount of foreign holdings as of a single date. The requirement that reports must be filed within fifteen days of the end of the month increases the possibility of error, and the monthly reporting and publication cycle limits the time available for scrutiny of the reports and correction of errors. Moreover, correct reporting by large banks, brokers, and other firms requires constant attention to the adequacy of internal communications among their various departments to ensure that they report all transactions carried out for foreigners and all long-term assets held for foreigners by
the various parts of the organization. Experience has shown these to be difficult and persistent problems.
The third type of problem is encountered in the reporting of transactions with foreigners in long-term securities, and arises from the widely-used practice of holding securities under "nominee" or "street" names. The survey indicates that two-thirds of the stocks held for foreign account are held by U.S. “nominees” or custodians who act as holders of record; consequently, these stocks are recorded as being U.S.- rather than foreign-owned on the books of the issuing corporations. The reporting instructions require the bank, broker, or dealer acting directly by order of a foreigner-often a holder of record—to report transactions for his account as foreign. If the holder of record fails to do so, and does not inform the broker who executes a transaction that it is for foreign account, that transaction will be omitted from the current reports. On the other hand, it is possible that duplicate reporting can occur when two institutions are involved, unless there is adequate communication between them and they both give sufficient attention to the reporting instructions.
The tendency toward underreporting of current transactions that results from these reporting problems is unavoidable, inasmuch as complete solution of these problems would impose unacceptable costs on both the Treasury and the business community. The underreporting can be kept within acceptable limits, however, and an unreasonable level of cost avoided. The underreporting of securities transactions applies to foreign sales as well as to foreign purchases of U.S. securities. Since there have been persistent foreign net purchases of U.S. securities over the period between benchmarks, the underreporting of securities transactions is considered to have imparted a downward bias to the previous estimates based on the transactions data.
Improvements in Current Reporting System
As noted above, the principal area in which the survey suggests problems is in the current reporting of transactions with foreigners in U.S. stocks, with much smaller problems indicated in the reporting of transactions in bonds. The Treasury has been working actively on problems in this area for some time, independent of the work on the benchmark survey. This work has mainly consisted of contacts with large banks and brokerage firms, beginning in early 1975, to discuss problems and review procedures for reporting securities transactions.
The survey provided some fairly specific indications as to areas where current reporting of securities transactions might be improved. The questionnaire included a question designed to identify firms which during 1974 had engaged in transactions in long-term securities with foreigners but had not been reporting on Treasury Form S-1. The replies to this question indicated that 405 firms, on their own behalf or on behalf of customers, had engaged in transactions in long-term securities with foreigners during 1974; 105 of these firms stated that they had been filing reports on Treasury Form S-1.
The names of the 300 respondents which had not been reporting were screened to determine which firms were potential Form S-1 reporters. It was recognized that there could be many valid reasons why such firms might have no obligation to report, including the possibility that the transactions they handled fell below the exemption level, were not otherwise reportable, and were reportable by another U.S. firm through whom they were executed or cleared. Finally, there was a possibility of error in the responses to the question itself. A number of names were struck from the list of potential reporters on the basis of checking the initial responses for obvious errors. Other names were deleted and some were added as a result of further research and crosschecks agn: + the list of existing Form S-1 reporters.
Letters were sent in late January and early February 1976, to some 125 firms that were considered to be prospective Form S-l reporters, requesting them to review their records to determine whether a reporting responsibility existed. Only a few of the responses have indicated any reportable transactions. In most cases it was found that the securities transactions with foreigners noted in the survey were covered by Form S-1 exemption provisions, or that the transactions were not otherwise reportable on Form S-1. While these letters did not produce large additions to the data, they did bring the reporting requirements to the attention of the recipients, which should help ensure the reporting of future transactions where required.
The reports filed in the survey by holders of record provide a useful context for discussion of the “nominee" and "street name" practices mentioned above. Because the survey reports show securities held for foreigners while the Form S-1 reports show the amounts of transactions for foreigners, it was not expected that there would be any measurable quantitative connections between the amounts reported by a given firm on the two forms. Discussions with a number of holders of record have confirmed the lack of such measurable relationships. The discussions, however, indicated that these financial institutions are well aware of the reporting procedures required by the practice of holding securities under ''nominee" or "street" names, and that the reporting of transactions for foreign account is generally handled adequately when the holder of record is a bank or broker, presumably accounting for the bulk of such transactions. Such discussions will be continued with additional firms to explore these matters more thoroughly.
In addition, the survey results will be examined for indications of broad areas of possible problems in current reporting. For example, a detailed comparison of the amounts of U.S. stocks, bonds and other debt obligations reported as held for the residents of individual foreign countries with available investment position and transactions data may indicate specialized problems in the reporting of transactions with particular countries. Similarly, analysis of the country data in terms of the types of foreign holders may provide useful insights.
Apart from the assistance and clues that may be derived from the survey, the ongoing work of the Treasury reporting staff to improve the current reporting will continue. Over the past year, Treasury staff members have visited 22 banks and 21 brokerage firms in New York, Boston, Chicago, Los Angeles and San Francisco, to ensure that the reporting requirements are understood. The discussions involved representatives of the various departments of the firms which may be concerned with securities transactions with foreigners. As a result of these visits, more than $1.2 billion of additional net purchases of long-term U.S. securities have so far been identified and entered into the securities transactions statistics, covering transactions from January, 1974 through March, 1976. These net additions to the statistics include nearly $650 million of foreign purchases of U.S. stocks and some $600 million of U.S. corporate and other debt securities. Some further additions to the net purchases figures may be forthcoming from several large institutions as a result of the visits.
In addition to these visits to banks and brokers, the Treasury and the Federal Reserve banks acting for the Treasury have communicated with a number of other banks and brokers for the same purpose. Most of these firms have replied that no changes in their reporting practices were called for. These efforts have not only produced significant amounts of previously omitted data but have also improved the financial community's understanding of the reporting requirements and the importance of the data. Effective liaison with the reporting firms is the key to effective reporting, and the Treasury will endeavor to give priority to this work.
The survey results also suggested problems in the current reporting of long-term debt obligations other than bonds, i.e., long-term liabilities of banks and nonbanking firms in the United States. The difference between the previous estimates and the survey results for debt obligations other than bonds is probably mainly in the nonbank area. The reporting of long-term debt obligations by banks is considered to be satisfactory. Reports of liabilities to, and claims on, foreigners filed by banks on the Treasury Foreign Exchange forms are of high quality, and are considered the best statistical data in the Treasury system. The reports filed by nonbank firms and nonprofit institutions are substantially less satisfactory than those filed by banks, largely because the problem of communication is much greater in the nonbank area. The survey questionnaire required banks and nonbank firms to indicate whether they had liabilities to, or claims on, foreigners and if so whether they had been reporting on the Treasury Foreign Exchange forms. On the basis of a preliminary review, the responses by banks do not suggest significant omissions. The responses by nonbank firms, however, suggest the possibility that a number of these firms should be reporting. More than 700 firms indicated that they had liabilities to, or claims on, foreigners during 1974 but had not reported on Treasury Form C-1/2. These responses will be screened, as in the case of the reponses on securities reporting, and the Treasury will communicate with the remaining firms to inform them of the details of the reporting requirements and to determine whether they should be reporting. This procedure should have the same effect as a general canvass of the business community, the method which has been used successfully in the past, and should significantly improve the data reported on Treasury Form C-1/2.
The Treasury Department's current statistical reporting system was designed to provide reasonably accurate information on current movements of capital between the United States and foreign countries, for use in the balance of payments and for the formulation of international financial and monetary policies. The conceptual basis and institutional structure of the system are well established and adequate for its purposes. To the extent that current transactions result in changes in holdings for foreign account, the current reports are useful in maintaining a running inventory of total foreign holdings between benchmarks; but this running inventory will not be precise because the value of foreign holdings at any given date may be affected significantly by factors other than money flows, such as valuation changes, movement by owners and other factors mentioned above. The results of the survey provide general indications which are useful in appraising the adequacy of the current reporting system, but the relationship between the two sets of reports is too indirect to permit the survey to be used for precise identification of problems in the current reporting. At best, the survey can suggest areas that should be explored.
Similarly, no precise method of determining standards for appraising the current reporting exists in the balance of payments data, or elsewhere in data presently available. The most effective means of ensuring the adequacy of the current data are constant monitoring of current reports and ongoing communication and liaison with reporting firms and the financial community.
The current reporting system does not purport to cover the entire universe. Exemptions are provided
which exclude many firms for the purpose of reducing the reporting burden on the business community. The limitations on the resources available for this work, both in the business firms and in the collecting agencies, have a considerable effect on the extent to which the practical problems which arise in the reporting of this complex array of data can be solved.
The results of the new benchmark survey suggest that there is an underreporting bias in the current reporting system, as was to be expected. However, the differences between the survey totals and the previous estimates do not appear unduly large, in view of the long period of time that has elapsed since the previous benchmark survey and the significance of the non-transaction factors affecting the survey totals. It is of interest that the difference is substantially larger in the figures on equity holdings, where the valuation adjustment problem is greatest, than it is for holdings of debt securities. Also, the reported large holdings of stock by U.S. citizens residing abroad suggest that migration abroad of owners of stock which had been purchased in the United States could be a significant factor in the difference. Because the survey results do not appear to raise major questions about the current reporting system, no specific recommendations are presented for steps to improve the system.
The Department of the Treasury will carry forward its work on communicating effectively with the financial community to inform reporters and prospective reporters of the reporting requirements, to solve reporting problems as expeditiously as is feasible, and to keep abreast of developments affecting the data. In this work the Treasury will utilize fully the analytical opportunities offered by the new benchmark survey in tracking down reporting problems and in identifying areas that need further work. The perspectives provided by the survey will be of considerable assistance not only in providing clues for investigation but also in generating wider understanding of these matters in the financial community, which will materially facilitate the liaison with reporting institutions which is essential for adequate reporting.