Page images

in January 1975, although the decline in the tradeweighted exchange rate of the dollar continued through the end of February 1975. However, if foreign investors anticipated that the exchange rate was close to the bottom, they would have had an added incentive to purchase U.S. stocks. The incentive to buy may have been even greater after February, if investors assumed that the rise in the exchange rate that began then would continue. Since, in the event such anticipations turned out to be correct, the foreign investor not only realized a greater appreciation of his investments (measured in foreign currencies) than did U.S. investors (measured in dollars), but the foreign investor could also expect larger future yields on his investments, as a result of the appreciated dollar rate (again measured in foreign currency terms).

From the end of 1974 through the second quarter 1975, admittedly a very short period, the two balances moved in the same direction, but in the last half of the year the opposite movements appear to have resumed.

These data indicate, therefore, that at least the major movements in private foreigners' net purchases of U.S. stocks generally have gone in the opposite direction from movements in the merchandise trade balance and thus each has tended to mitigate the effects of the other on the exchange markets and—when the par value system was in effect-on the overall balance of payments, as then measured. However, the changes in the balance on transactions in U.S. stocks have been small relative to changes in merchandise trade and in other international transactions, particularly other capital movements. Not surprisingly, therefore, no regular relationship can be found between the balance on private foreign transactions in U.S. stocks and the overall balance of payments during the years prior to the move to widespread floating. In some periods, e.g., 1964-1967, rising private foreign purchases of U.S. stocks were accompanied by declining net purchases of U.S. assets by official accounts. On the other hand, in 1968 large net purchases of stocks were accompanied by large official purchases of U.S. assets; the result was that the U.S. “deficit" on other transactions was larger than the net inflow of funds involved in the stock purchases.

Since March 1973 the exchange rate of the dollar has been determined largely by exchange market conditions. Increasingly, purchases and sales of dollars in exchange for foreign currencies by official agencies of most countries have been conducted largely to prevent disorderly conditions in the exchange markets and not to maintain particular exchange rates. The changes in the exchange rates during the three years 1973 to 1975 do not seem to have been closely related to the changes in the balance on foreign private transactions in U.S. stocks. During 1973 the trade-weighted exchange rate of the U.S. dollar dropped about 6 percent, while net foreign purchases of U.S. stocks increased about $400 million. During 1974 the exchange rate declined about 1.6 percent, while net foreign purchases of U.S. stocks dropped about $2.6 billion, and in 1975, the rise in the exchange rate of 6 per cent coincided with a $2.8 billion rise in net foreign purchases of U.S. stocks.

These coincidences in 1974 and 1975 could have been the result of two developments. The changes in the exchange rate could have resulted in part from the changes in the foreign demand for dollars to pay for U.S. securities. The foreign demand for U.S. stocks may, in turn, have been influenced by economic developments, such as the decline in domestic business activity in 1974 and the rise in 1975 which became evident in the second quarter of the year; but this rise may have been anticipated earlier, as suggested by the rise in domestic stock prices which started around the turn of the year.

Alternatively, the changes in net purchases of U.S. stocks may have been induced by the changes in the exchange rate. In 1974 a declining U.S. dollar exchange rate may have postponed foreign investment in U.S. stocks. Foreign investment in U.S. stocks started to rise

Bonds and Other Long-term Debt Obligations

The total net capital inflow through foreign private purchases of U.S. corporate bonds and through other loans to U.S. corporations during the years 1965 through 1975 amounted to about $15.7 billion, as measured in the balance of payments statistics, about 10 percent more than the net capital inflow through foreign private purchases of U.S. stocks. The capital inflow through the net sales of bonds accounted for $11.5 billion, and the inflow through other loans for $4.2 billion. The conceptual differences between these figures and those obtained in the survey have been pointed out earlier.

Borrowing abroad by U.S. corporations, either through the sales of bonds or through bank loans, became important when capital outflows from the United States to finance direct investments abroad were restricted. The funds borrowed abroad were largely used by U.S. corporations to finance capital requirements of their foreign branches and subsidiaries, particularly those located in Europe. The program to restrict capital outflows through direct investments was initiated on a "voluntary" basis in 1965. In 1968 the program became compulsory and continued until January, 1974, when it was terminated. During the years 1965 through 1973, U.S. corporations obtained from foreign sources through bonds issued by financing affiliates domiciled in the United States and the Netherlands Antilles about $9.6 billion, and through bank loans about $4.1 billion. In the peak year 1972, such funds obtained abroad were about $2.6 billion. (In addition to these amounts, foreign affiliates of United States corporations borrowed funds abroad which were invested in their own organizations or lent directly to other foreign affiliates. The figures included in U.S. balance of payments compilations, therefore, do not reflect the total amount of foreign borrowing by U.S. corporations and their foreign affiliates that was induced by the programs to curtail capital outflows from the United States.)

After the termination of the controls, overseas bond issues by U.S corporations declined, to $120 million in 1974 and to $200 million in 1975. Funds obtained through long-term loans dropped to about $100 million in each of these two years. (This decline includes net repayments of about $250 million on loans obtained in the Bahamas; omitting these repayments net new loans obtained elsewhere dropped to about $30 million in 1974 but rose again to $325 million in 1975.) The decline in borrowing from foreign sources in these two years taken together was offset in part by somewhat larger purchases by private foreigners of U.S. corporate bonds in U.S. markets. Such purchases (net of sales) were about $480 million in 1974 and $240 million in 1975.

The decline in net capital inflows through private foreign investments in long-term debt obligations in the last two years is clearly related to the termination of the U.S. capital control program. Additional factors may have been the decline in requirements for new capital by the foreign affiliates of U.S. corporations, and, perhaps their policies of reducing their debt obligations relative to their total assets.

The costs of borrowing were usually higher abroad than in the United States during the period when capital outflows were restrained. However, in 1974 the differences between yields on U.S. corporate bonds issued abroad and those issued in the United States narrowed, and in the first half of 1975 yields on outstanding bonds originally issued abroad were actually lower than on those originally issued in the domestic markets. In the second half of the year they were about the same. It is possible, therefore, that borrowing by U.S. corporations in foreign capital markets, particularly to finance investments abroad, may continue, and—as the 1975 experience seems to suggest—the recent drop in such borrowings may be followed by another expansion. It would at best be premature, however, to expect that corporate borrowing abroad will be resumed on a scale comparable to that of the years in which capital outflows were restrained.

purchases, dividend payments will exceed new capital inflows. This would be true of investments in fixed-income securities, such as bonds, as well; in either case the income is paid on the total accumulated investment. But in the case of stocks, since corporations customarily reinvest part of their profits, the rate of return calculated as a percentage of the amount originally invested should also tend to rise over time. In other words, the shareholder is not only receiving a return on his original investment but also on his share of the reinvested earnings. So far, however, this phase of the expected relationship between new investments and current income receipts has not yet been reached.

One factor contributing to the relatively slow rise of dividend payments to private foreign investors relative to their net purchases of stocks is the relatively fast recent rise in such purchases from an annual average of about $900 million in 1965-69 to an average of about $1800 million in 1971-75. As a result, the accumulated stock of investments resulting from lower annual inflows in the past (which determines the amount of dividends paid) is relatively small compared to recent annual inflows.

Another factor contributing to the relatively slow growth in dividend payments is that some foreigners may, in effect, take their potentially larger dividends resulting from the higher earnings attributable to reinvested profits in the form of capital gains. Capital gains or losses by foreigners are not included in the earnings figures in the balance of payments. When foreigners sell securities, the prices at which they are sold and not the original costs are reflected in the data on capital flows. These sales figures include, therefore, the foreigners' realized capital gains or losses. For instance, when foreigners sell stocks at, say, $100 which they had purchased at $80, and purchase other securities at $100, the balance on transactions in stocks, as shown in the balance of payments and used in this chapter, would be zero, and no further entry into the balance of payments accounts would be made for the capital gain of $20. (This example omits the receipts of commissions.) However, both unrealized and realized capital gains (to the extent the latter were reinvested in the United States) are reflected in the survey totals.

Payments of Dividends and Interest on Foreign Private Portfolio Investments in the United States

Balance of payments estimates of dividends and interest paid to foreign investors are not based on reports by those making or transferring such payments, but are calculated on the basis of estimated totals of the outstanding balance of such investments and estimated yields thereon, less an allowance for withholding taxes. The data on foreign portfolio investments and their composition resulting from this survey will be used to revise these estimates. However, even the unrevised data as previously published are sufficiently accurate to draw generally valid conclusions covering the basic trends in these payments and their relationships with other balance of payments transactions.

Although dividend payments to private foreigners increased steadily, and more than doubled from 1965 to 1975, there were only a few years, 1965, 1966 and 1974, in which these payments exceeded the net inflow of capital (net of commissions) through foreign purchases of U.S. stocks. Except for these years, the balance on all transactions related to foreign private investment in U.S. stocks resulted in net receipts. In fact, the trend in that balance during these years showed that net receipts were rising. In the long run, assuming a constant flow of net

Impact of the Changing Balance
on Portfolio Investment Transactions
on Exchange Rates, on International Trade,
and on the Domestic Economy

Changes in capital flows associated with foreign investments in the United States and in income payments on these investments can be compensated for by opposite changes in capital flows associated with United States investments abroad and in income receipts. To the extent that this occurs, the balance on other transactions would not be affected. The following discussion deals with the problems that may arise if changes in capital flows and income payments do not offset each other.

An increase in the foreign demand for U.S. securities which exceeds the rise in the net outflow of U.S. funds through payments of interest and dividends would tend

to raise the exchange rate of the dollar and, in turn, tend To the extent that the capital inflow stimulates to lower the balance on trade in goods and services. The domestic consumption directly, or is used for investment decline in the trade balance could result from a decline to produce consumer services (including consumption in exports, an increase in imports, or both. The induced expenditures by governmental organizations), it would changes in the balance on trade do not imply, however, tend to support an excess of domestic consumption over that foreign purchases of U.S. securities must have an domestic production so long as capital imports exceed adverse effect on the U.S. economy and on U.S. produc- income payments on the debt. However, when interest tion and employment.

payments on the debt exceed new capital inflows, if there In periods of rising or high levels of production rela- had in the meantime been no expansion in output, tive to the domestic productive capacity, the rise in the domestic use of domestic production for consumption exchange rate would tend to dampen increases in and investment would have to be reduced in order to domestic prices and indirectly might also contribute to a make more of the domestically produced goods available reduction in the pressures for higher wages. In that case, to foreign buyers. This shift in the allocation of domestic the impact on domestic price movements could be con- production between domestic and foreign buyers may insiderably larger than the combination of the changes in volve a downward movement of the exchange rate of the exchange rates and the share of foreign trade in the dollar. The sequence might be that the excess supply of economy would suggest.

dollars resulting from the income payments would tend In addition, the impact on domestic production and to depress the dollar on the exchange markets, thus leademployment must not be judged on the basis of the ing to a rise (fall) in net exports (imports)—the reverse of changes in the trade and services balance alone, even if what happens during periods when capital inflows exthat is the only explicit item for international transac- ceed income payments. tions included in the customary GNP calculations. As If capital inflows are discouraged through policy discussed in more detail in the last section of this measures imposed by the United States (or through chapter, capital inflows can lead to increased domestic policy measures to discourage capital outflows imposed investment, thus filling the gap in production and by other countries), the demand for dollars on the interemployment that may have been opened up by the national exchange markets is reduced. The result is a decline in the trade balance. However, the possibility decline in the exchange rate of the dollar which will conalso exists that, indirectly, the funds coming in from tinue until buyers are found who use the dollars to pay abroad might finance consumer or government expen- for additional merchandise or services obtained in the ditures, instead of increased investment. An increase in United States, or until the supply of dollars in internacapital inflows through larger purchases of corporate tional exchange markets is reduced as U.S. imports are stocks and corporate debt obligations may perhaps have discouraged by the higher dollar prices that have to be a more direct effect on domestic investment in produc- paid for goods and services purchased abroad. tion facilities than a capital inflow through larger Under these circumstances the trade balance would purchases of banking or government obligations, but the be larger than if capital inflows had taken place, and final distribution of the funds among those who use them would be lower later as a result of the failure of payto finance the expansion of productive facilities and ments of income on foreign investments to rise. While those who use them to finance other expenditures would both current and subsequent adjustments in the trade be difficult to determine either statistically or balance and in the relationship between domestic contheoretically.

sumption and production would be avoided, there may As explained earlier, during any given time period, all be a potential loss of an opportunity to acquire capital the transactions in the balance of payments must add which could stimulate domestic investment and thus acalgebraically to zero; that is, the debits and credits must celerate the growth of the economy and of domestic inbe equal. Thus, if income payments rise relative to the comes, even if a part of these additional incomes later volume of new capital inflows, unless there are changes has to be transferred to foreigners through payments of elsewhere in the capital accounts, exports of goods and

dividends. services (exclusive of payments of investment income) The lower exchange value of the dollar relative to will tend to rise relative to imports. This change will what it would be if capital imports had not been dismean that, in absolute terms, an increasing amount of couraged would also contribute to higher domestic domestic production will be made available to prices, because exporters who can sell abroad at higher foreigners. However, this can be accomplished without a dollar prices, and importers who would pay higher doldecline in the absolute amount available to domestic lar prices to foreigners, would charge higher prices to buyers if the increase in domestic production can be their domestic customers. Policies to discourage capital speeded up as a result of the investment of the inflow of inflow, whether they originate in the United States or foreign capital in productive facilities. In fact, it is likely abroad, would tend to raise domestic prices. If such price that such investment will also raise domestic produc- movements contribute to increases in nominal wages tivity, so that a rise in the trade balance can occur and thus to further price increases they may stimulate simultaneously with a rise in production available for inflationary developments. This may appear as a remote domestic use. Presumably, the increase in the trade danger, but with foreign trade comprising a rising share balance could be achieved under these conditions with a of our economy, and involving a rising share of basic minimum of adjustment in exchange rates.

commodities, such as foodstuffs and fuels, domestic price

[blocks in formation]

Net foreign private

purchases of U.S.
stocks and corporate

and other hondsa 2,190 Estimated price changes Combined effect on inter

national investment
position of the
United States ...... -2.626

2,269 2.219

4.507 4.183

4,056 -6,036

672 -9,370,

13,694 -8,568


[blocks in formation]
[blocks in formation]

As should be clear from the immediately preceding discussion, foreign portfolio investment is not likely, in the short run, to have any effect on the overall investment position of the United States. Instead, it is more likely to reflect a shift in the types of U.S. assets held by foreigners and a shift among foreign holders. This was particularly true prior to 1973 when private purchases of U.S. assets tended ceteris paribus, to be offset by sales of officially held assets. This would not occur, of course, if the private purchases were directly connected with forces causing an offsetting change in some other balance of payments transactions; as explained above, this seems frequently to have been the case.

It is also possible that the foreign portfolio investment may cause (or be associated with) an outflow of U.S. capital of some kind. In this case, too, the overall investment position of the United States would not be immediately affected.

Even if the immediate effect on the investment position is zero or negligible, any change in the pattern of foreign investment, or in the relationship between foreign investment in the United States and U.S. investment abroad, could subsequently affect the investment position, as conventionally measured. This is due mainly to the fact that stocks and marketable debt are measured at market values and that the value of stocks tends to fluctuate more than debt prices. Over the long run, foreign purchases of equities will tend to have a greater negative effect than foreign purchases of bonds on the net international investment position of the United States because the reinvestment of corporate profits will tend to raise stock prices. Also, of course, our net asset position will be less strong to the extent foreigners succeed in concentrating their purchases in times when stock prices are relatively depressed.

Moreover, to the extent that inflation tends to raise stock prices more than debt prices, foreign purchases of stocks would affect our international position (in a negative direction) more than an equivalent purchase of bonds in an inflationary period. However, taking current yields and capital appreciation combined over a period of time, such an effect could be offset by the higher interest rates on bonds that tend to accompany inflation.

The fact that the survey was taken at virtually the lowest point reached by the U.S. stock market in more than a decade thus tended to make the net international investment position of the United States appear stronger than it was, for example, at the end of 1975.

The influence of price changes of U.S. securities on the international investment position of the United States in recent years is shown in the following tabulation:

In most years (although not over the period as a whole), price changes had more effect on the investment position than did actual transactions.

Changes in the place of residence of portfolio security holders also affect the international investment position without a recorded balance-of-payments capital flow. U.S. nationals abroad were reported, as of December 31, 1974, to hold $2.7 billion of U.S. portfolio securities. These holdings represented 7 percent of all U.S. portfolio securities held by private foreigners and 9 percent of the $23.7 billion in private foreign positions in U.S. equity securities. It is likely that a significant share of these holdings of U.S. citizens residing abroad was acquired while the owners were U.S. residents.

If the foreign portfolio investment is related to changes in the trade balance, either because of a causal relationship or because both are affected by common forces, the U.S. international creditor position is "worsened." That is, foreign investment in the United States rises, and correspondingly the cumulative export surplus is smaller than it would have been in the absence of the capital inflow. To the extent that capital inflows and the change in the trade balance are related, the U.S. economy, at any given level of national output, has more goods and services available for domestic consumption and investment than it would otherwise have had. Assuming the domestic environment is favorable, the net result could be less (positive) net international investment and more domestic (real) investment. But whether the excess of goods and services (over domestic output) is consumed or invested, the net international investment position will show smaller net claims on the rest of the world.

Effects on U.S. Financial Markets

Foreign portfolio investment in the United States as defined for purposes of this study—is a small but not insignificant source of capital for American firms. However, foreign holdings of U.S. Government longterm debt accounted for a significant portion-30 per cent—of the long-term public debt held outside the Government and the Federal Reserve System.

The extent to which these foreign investments provide additional funds to American financial markets as a whole is a question of complex analysis, involving ceteris paribus assumptions that must be carefully spelled out,

Corporate shares ...

697.6 billion Credit market instruments a (excluding household debt)

$1,673.7 billion Total. .....

$2,371.3 billion Foreign portfolio investment (excluding time deposits at banks)..

$66.4 billion Ratio of foreign to total.....

2.8 percent a The term does not include deposits at financial institutions.

including the nature of the exchange system (flexible vs. fixed rates) and the response of monetary authorities (United States and foreign) to exchange market and balance of payments developments. As discussed earlier in this chapter, during any given time period, the balance of payments on current account (net exports or imports of goods and services plus unilateral transfers) must be equal to and opposite in sign from the balance on all capital transactions, private and official, involving U.S. investment abroad and foreign investment in the United States. Thus, in the very short run at least, foreign portfolio investment may not add to overall funds available for real investment in the United States; foreign purchases of U.S. stocks, for example, may merely result in a decline in short-term dollar assets held by the same or other foreigners. Indeed, some simultaneous, exactly offsetting, capital transaction must occur, unless it is assumed that some change in the balance of exports and imports is either caused by the capital inflow, or is simultaneously induced by the same general economic forces that produced the capital inflow (again see the earlier discussion).

Another important ceteris paribus assumption relates to the actions of the U.S. monetary and fiscal authorities: presumably they would take offsetting action if foreign capital flows were having undesired effects on domestic interest rates or other financial market factors.

Keeping the foregoing limitations in mind, it can probably be safely said that foreign portfolio investment in the United States, on balance and over time, has a small, but not insignificant, influence on U.S. capital markets, tending to raise the prices of securities and lower their yields. Such a development should tend to increase real domestic investment, thus adding to the output and productivity of U.S. industry.

Foreign investment might also affect the relative cost of capital among sectors of the U.S. economy. As detailed below, however, the pattern of foreign equity holdings by industry is not significantly different from the overall pattern of stock outstanding. The structure of foreign investment in U.S. bonds, however, is much more heavily weighted in favor of Government debt (because of the large official holdings); and private holdings of U.S. bonds are concentrated in so-called offshore issues, which provided relatively limited funds for domestic investment, as described earlier. Thus private foreign portfolio investment could be said, if anything, to have more of a favorable effect on the cost of equity capital to U.S. corporations than on the cost of borrowed funds.

However, the influence thus far of foreign portfolio investment on U.S. interest rates and other security yields should not be overestimated. At the end of 1974, these nvestments amounted to only about 2.8 per cent of total rutstanding equity issues and credit market instruments mitted by U.S. non-financial businesses and governnent sectors, as shown by the flow of funds accounts:

These figures are only a rough indication of the overall importance of foreign portfolio investment. The ratio is somewhat understated, in that "credit market instruments" include some short-term debt, particularly short-term bank loans, a type of claim not included in foreign portfolio investment.

Whether foreign investment has, on balance, influenced the relative availability of equity capital among U.S. industries can best be determined by comparing the ratios of foreign holdings to total stock outstanding by ndustry categories. This comparison has been made for all companies reporting in the survey. The results are presented in Appendix Table A-16. Although there are some noticeable variations by industry group, there is no readily explainable pattern. It seems more likely that such variations result from the size patterns of companies within an industry. As discussed in Chapter 2, there seems to be clear evidence of foreign preference for stocks of large (and therefore well-known) corporations.

In any event, any differential impact by industry of the foreign funds placed in the U.S. stock market would be overwhelmed by the effect of the preferences of U.S. shareholders, given the extremely small proportion of foreign holdings.

Foreign Trading in U.S. Stocks

In any given time period, both the gross amount of foreign trading in U.S. stocks as well as the amount of net purchases (or sales) may affect market developments—and of course are affected by them.

For instance, it appears that foreigners as a whole engage in more trading on the U.S. stock exchanges in relation to the size of their portfolio than do American investors as a whole. In 1974, the ratio of total transactions (in dollars) on the New York Stock Exchange to the end-year market value of all listed shares was about 19 per cent; the comparable ratio for foreign transactions (gross purchases plus gross sales divided by two) was about 30 percent. In the years 1970-73, the disparity was even larger, as shown in Table 16.

« PreviousContinue »