securities (except for Canadian new issues and a selected list of outstanding Canadian stock issues in which there were large U.S. interests). The effect of the IET was virtually to preclude sales of securities issued by developed countries, other than Canada, to U.S. residents.

Since the elimination of the IET in January 1974, Americans have substantially increased their purchases of foreign securities, especially from developed countries. Table 21, which contains data for 1972-75 on U.S. purchases of foreign bonds that were newly issued in the U.S. market, indicates how strongly developed country issuers have returned to the U.S. market. In 1975, the major foreign issuers in the U.S. securities market were Canada, the IBRD, Israel, European and Latin American regional organizations, Japan, Norway, France, Mexico, and Australia.

Most of the nearly $20 billion in U.S. holdings of foreign-issued bonds are the result of U.S. purchases of new issues which were marketed in the United States by American-based underwriters. The data in Table 22 indicate that U.S. residents purchased more than $14 billion of newly-issued foreign bonds between 1971 and 1975; in the same interval, there was an increase of only about $300 million in U.S. holdings of foreign bonds due to transactions in secondary markets.

Factors Which Influence U.S. Purchases

A number of factors influence U.S. residents to purchase foreign securities. The heavy U.S. concentration on foreign new issues suggests that U.S. holdings of foreign bonds are affected by strong sales efforts by foreign issuers and their U.S. underwriters.

In addition, however, there is a ready market for foreign issues in the United States. This is in part due to the high investment quality of many of the issues placed here. By far the largest portion of all issues placed in the United States are either government issues, or government-guaranteed issues.

Furthermore, American insurance companies set aside a certain small proportion of their portfolios for foreign issues, particularly Canadian ones. As a result they represent a continuing source of demand for new foreign issues, for the replacement of maturing issues, and for the growth of their investment portfolios.

Another factor that encourages U.S. residents to purchase foreign bond issues is the interest rate premium that foreign issues usually carry over domestic issues. Also there are several special factors which encourage the purchase of particular foreign bonds: Certain Canadian issues are probably perceived to be less "foreign" than other foreign issues (just as Canadians view U.S. issues differently from other foreign issues); IBRD bonds have a virtual U.S. Government guarantee; and Israeli bonds benefit from special selling efforts.

Several factors encourage U.S. investors to purchase some foreign stocks, as well as foreign bonds. First, cyclical movements in stock prices often are not synchronized between U.S. and other world securities markets. This lack of synchronization permits sophisticated investors, both in the United States and abroad, to reduce risk (or variability) without sacrificing return. Second, by buying foreign stocks, U.S. investors can participate in foreign markets which are expected to have a greater real trend rate of growth than the United States market.

The ideal foreign market would be one which has both the world's highest growth rate and an inverse correlation with the U.S. market. Based on such analysis, Japan has been attractive to U.S. investors in the past, whereas the United Kingdom has not.

The Relative Smallness of U.S. Holdings of Foreign Securities

U.S. institutions2 have invested abroad relatively little of their total portfolio capital; this was especially true while the Interest Equalization Tax was in effect. This is in striking contrast to the high degree of international diversification routinely practiced by foreign portfolio investors.3

There are several reasons why Americans have shown relatively little interest in foreign securities other than those which have been actively sold in the United States in the form of new bond issues. In the first place, unlike foreigners seeking a diversified and readily marketable portfolio, Americans can acquire such a portfolio entirely in the domestic market; thus there is less need to seek additional investment opportunities abroad. Also, foreign securities markets tend to be less liquid and more volatile; secondary bond markets in other countries tend to be thin or non-existent.

Second, as noted above, the Interest Equalization Tax imposed from mid-1963 until January 1974 on purchases of foreign securities issued by developed countries (except for new issues by Canada) was a strong deterrent to U.S. purchases of such foreign securities.

Third, there are several nuisance factors involved in owning foreign securities which may further reduce their attractiveness to U.S. investors. For example, if the security is not listed on a U.S. exchange, U.S. residents will find little information available on the company, and that information frequently is in a foreign language. Furthermore, U.S. residents will find that in many countries the transfer agents who deliver securities, dividends, and interest perform their functions with considerable delay, so that the purchaser may not receive his stock or bond certificate for several months, and dividend and interest payments may arrive months after payment date. Also, in many cases a withholding tax is imposed by the government of the issuing entity on the payment of interest and dividends. While the United States gives a tax credit for these withholding taxes, there is still an additional form to be filled out and this paperwork undoubtedly makes the foreign securities less attractive. In addition, rules and regulations governing

2 There was no practical way to analyze the behavior of non-institutional American portfolio investors abroad.

3 For example, U.S. securities accounted for 30 percent of the total

foreign securities markets are often complicated, requiring expensive legal assistance.

Finally, despite the fact that listing foreign securities on U.S. exchanges (see below) may increase their attractiveness, institutional investors, such as trusts, view foreign securities as distinctly different from domestic issues, even in cases where their ratings by U.S. rating companies are comparable to those on the best U.S. securities. This difference in perception results in part from legal considerations: should a foreign company default on a bond redemption or go bankrupt, a trustee holding this company's securities might be faulted by a U.S. court for not having been "prudent" in acquiring the securities, because a U.S. investor would have less recourse to compensation in foreign courts than he would in U.S. courts were a comparable U.S. company to default. Indeed, in many cases there are lists of approved securities which trustees are obliged by State law to consult in making investments. These lists do not in general include foreign securities.

Listing Foreign Securities on U.S. Exchanges

Listing foreign securities on U.S. exchanges generally makes them more attractive and accessible to potential U.S. purchasers, and to an extent such foreign securities may be viewed by Americans as little different from U.S. securities. Price quotations and dividend and interest payments of listed foreign securities are in dollars; also, purchase and dividend transactions are conducted through an American intermediary. Furthermore, since foreign issues listed on U.S. exchanges have to be registered with the Securities and Exchange Commission, and the issuers have to meet essentially the same requirements as their U.S. counterparts (including the provision of annual reports in English), there are considerable amounts of information available on listed foreign companies.

At the end of 1974, there were 182 foreign securities listed on the New York Stock Exchange, of which 148 were bonds and 34 were stocks. Of the bond issues, 51 were issues of foreign governments and 21 were foreign corporate issues. Half of the foreign stock issues were by Canadian companies with the remainder mostly Dutch and British issues. In addition to issues listed on the New York Stock Exchange, many foreign issues are listed on the American Stock Exchange, NASDAQ, and traded in the over-the-counter market. As of the end of 1975, 77 foreign issues were listed on the ASE, of which 71 were equity issues, mostly Canadian.

holdings of a sample of 33 European funds, with total assets of $4.4 billion, interviewed by R. Shriver Associates.

4 See 1975 Fact Book, New York Stock Exchange, p. 33

Section 6 (9) and (10) of P.L. 93-479 directs the Secretary of the Treasury to "study adequacy of information, disclosures, and reporting requirements and procedures", and to "study and recommend means whereby information and statistics on foreign portfolio investment activities can be kept current." The response to this requirement principally involves an appraisal of the adequacy of the Treasury's existing system for statistical reporting of foreign portfolio investment transactions, and the formulation of recommendations for any improvements that appear to be necessary.

Current Treasury Reporting System

The statistical data on current foreign portfolio investment in the United States are produced by the Treasury Foreign Exchange reporting system. Since it was established in 1934, this system has been the source of official U.S. Government statistics on movements of capital (other than direct investment) between the United States and foreign countries, U.S. Government credits and loans, and foreign official investments in nonmarketable U.S. Treasury bonds and notes. Data on U.S. direct investment abroad, foreign direct investment in the United States, and U.S. Government credits and loans are collected by the Department of Commerce. Data on foreign purchases and redemptions of nonmarketable Treasury bonds and notes are derived from the records of the Treasury Department. The Treasury and Commerce reporting systems together provide the data on capital flows which enter into the balance of payments of the United States-both flows of U.S. capital to foreign countries and flows of foreign capital into the United States.

The twelve Federal Reserve banks act as fiscal agents of the Treasury Department in the operation of the Treasury Foreign Exchange reporting system. The Federal Reserve Bank of New York collects and reviews reports from banks in the New York Federal Reserve District and from securities firms and nonbanking firms throughout the United States. The other Federal Reserve banks collect and review reports from banks in

These Treasury Foreign Exchange forms cover U.S. investment abroad as well as foreign investment in the United States. Thus Form S-1 covers transactions in foreign stocks and bonds; Form B-3 covers

their respective Federal Reserve districts. The New York Bank consolidates the data from all of the Foreign Exchange reports filed in all districts and performs central computer services for the Treasury Department.

Data derived from these Treasury reports are published monthly in the "Capital Movements" section of the Treasury Bulletin and in the Federal Reserve Bulletin. The data are also used in the U.S. international transactions and investment position compilations published by the Department of Commerce in the Survey of Current Business.

Data on current foreign portfolio investment in the United States, other than in the nonmarketable Treasury issues mentioned above, are derived from three of the Treasury Foreign Exchange forms. Form S-1 produces monthly data on aggregate foreign purchases and sales of long-term U.S. securities: U.S. Treasury and agency bonds and notes, corporate stocks, and corporate and other non-Federal bonds. Form B-3 provides end-ofmonth data on the outstanding amounts of long-term liabilities to foreigners reported by banks in the United States, such as long-term certificates of deposit held by foreigners. Form C-1/2 provides end-of-quarter data on the outstanding amounts of long-term liabilities to foreigners reported by nonbank business enterprises and nonprofit institutions, such as borrowings by such enterprises from foreigners. The transactions data reported on Form S-1 represent the bulk of foreign portfolio investment activity.1

Monthly reports on Treasury Foreign Exchange Form S-1 are required from all persons in the United States who, on their own behalf or on behalf of their customers, engage in transactions in long-term securities directly with foreigners in amounts in excess of the reporting exemption. Reports are filed principally by banks, brokers, and dealers in the United States; other firms are required to file reports on Form S-1 when they buy or sell long-term securities directly with foreigners without using a bank or broker in the United States to effect the transactions. In practice, reports are not filed by individuals. Firms whose, aggregate purchases and sales in transactions with foreigners each averaged less than $100,000 per month during the six months ending with

bank long-term loans to foreigners; and other long-term claims on foreigners, and Form C-1/2 covers nonbank claims on foreigners, as well as the liability ns mentioned above.

the reporting month are exempt from reporting for that month.

The data reported on Form S-1 cover transactions during each month, and do not provide information on outstanding balances of the various types of securities held by foreigners. The reports provide data on aggregate purchases and sales by residents of each foreign country or area listed on the report form, on the basis of the records of the reporting firms. The amounts reported represent the total amounts of money debited or credited to the foreign account as of the ledger dates of the transactions-i.e., the costs of purchase plus commissions, or the proceeds of sales less commissions, tax, etc. The reports on Form S-1 do not identify the foreigners who are involved in the securities transactions or the particular securities traded.

Reports of long-term bank liabilities to foreigners are filed monthly on Treasury Foreign Exchange Form B-3 by banks in the United States, including agencies, branches, and subsidiaries of foreign banks, and the Federal Reserve Bank of New York. Reports are required from banks whose aggregate long-term liabilities to foreigners averaged $500,000 or more in the six months ending with the report date, on the basis of monthly closing balances. The data reported represent amounts outstanding as of the end of the month.

Reports of long-term nonbank liabilities to foreigners are filed quarterly on Treasury Foreign Exchange Form C-1/2 by business enterprises other than banks and brokers. Reports are required from enterprises whose aggregate end-of-quarter liabilities-both short-term and long-term-to foreigners averaged $500,000 or more in the quarter ending with the report date and the preceding quarter, on the basis of closing balances. The data reported represent amounts outstanding as of the end of the quarter.

Comparison of Survey Results with Previous Estimates

Considering the long period of time-33 years since the previous benchmark survey and the numerous reasons for expecting differences between the current and benchmark data, the difference between the new benchmark data and the previous estimates of foreign portfolio investment suggests that the current reporting system is operating reasonably satisfactorily.

The table in Chapter 2 comparing the results of the survey with the previous estimates is repeated here for convenience (in billions of dollars):

A part of the difference between the previous estimates and the survey results may arise from deficiencies of the earlier benchmark survey. The 1941 survey may have been less complete than the 1974 survey because the business community may not then have been as aware of its reporting obligations; also the techniques used in this survey and in updating of the 1941 benchmark in 1949 (described below) were probably more subject to error than the new survey. Moreover, a 1971 Treasury survey used as the basis for the series on marketable Treasury bonds and notes was less complete than the new survey. The differences in the figures probably also arise in part from definitional differences, problems in valuation adjustments, and problems in current reporting.

The principal current reporting problems suggested by the survey results lie in the reporting of transactions in stocks. As the table indicates, the difference in the data on bonds other than Treasury securities is small. The difference for other debt obligations is large enough in relative terms to suggest some problems in the current reporting in that area.

Benchmark Data vs. Transactions Data

The previous estimates of foreign holdings of stocks and bonds other than Treasury issues were calculated by applying estimated valuation adjustments and current net transactions data to previous benchmark figures. The previous estimates are based on the 1941 Treasury census of foreign-owned assets in the United States, updated in 1949 on the basis of U.S. withholding tax returns which showed the residence of the foreign persons, banks, or other institutions in whose names the securities were registered and to whom dividends and interest were remitted.2 The estimate of foreign portfolio investment in U.S. stocks is based on the application to the updated benchmark of quarterly valuation adjustments based on Standard and Poor's composite price index of 500 common stocks (industrial, public utility, and railroad), and the net transactions data reported on Treasury Form S-1. The figures for corporate bonds held by foreigners are similarly calculated, using annual valuation adjustments involving a factor computed from the Standard and Poor's composite price series for high grade corporate bonds, and the Treasury transactions data.

The figures for estimated foreign holdings of marketable U.S. Treasury bonds and notes are based on a 1971 Treasury survey of amounts held for foreigners by banks and brokers reporting on the Treasury Foreign Exchange forms, to which the monthly transactions figures are applied.

There are several reasons for the differences between the results of these calculations and the data from the new benchmark survey. There are a number of differences between the current survey and the earlier benchmarks which probably produced some differences and U.S. citizens residing abroad, whose income was not subject to withholding.