and investments, they are carried separately in the capital movement table (VIII) but, as stated above, they are excluded from the estimates of private foreign investments in the United States.

Residual in Relation to Unreported Capital Movements and Investments. Payments for net gold and silver imports in 1940 exceeded net receipts on trade and service account by more than $2,809,000,000. Assuming that the trade and service estimates are reasonably correct, this difference indicates that during the year there must have been a net capital movement into the United States of approximately that amount. This figure would, in principle, represent the net result of all kinds of capital transactions-the net rise in foreign balances in United States banks; net sales of securities, either domestic or foreign, by United States investors to foreigners, or vice versa; net reductions in commercial credits granted to foreigners; advance payments by foreign purchasers for Ünited States merchandise; and similar transactions.

The difference between the $1,492,000,000 net inflow accounted for by available data on capital movements and the $2,809,000,000 indicated by the other balance-of-payments items cannot be explained in quantitative terms. This difference (shown in table I as "Other transactions and residual") is believed to represent chiefly unrecorded capital transactions, partly because experience has shown that the residual has been the largest during periods of heavy flights of capital and gold movements from foreign countries.**

Currency depreciation, exchange restrictions, political instability with resulting changes of governments, international crises, and war preparations, have all been conducive to a flight of capital to the United States each year since 1934. Under such conditions there was bound to be considerable emphasis on the concealment of assets and of their transfer to the United States, with the result that data collected through normal channels were certain to be incomplete. For example, funds actually owned by foreigners, whether for business or for propaganda purposes, may be held in this country under domestic names and not be recognized as foreign by reporting banks and brokers. United States and foreign securities may be bought through domestic accounts, or foreigners may place United States currency in safe-deposit boxes in this country and thus conceal the fact of foreign ownership. Capital coming into this country under such circumstances would not be reported as a foreign capital movement but would form part of the residual. It also presents a problem for those estimating the total of foreign investments.

Movements of the funds of refugees account for part of the unreported capital movement even without involving any direct effort at concealment. Before coming to this country the refugees may have had deposit accounts in one or more of our banks, accounts which were registered under foreign addresses. Upon arrival of the owners, the accounts were likely changed to domestic addresses, thus reducing actual foreign-owned deposits. Accounts shifted in this manner no

While several of the trade and service items have been the subject of intensive inquiry in recent years, their nature is such that perfectly accurate data cannot be obtained. Te results of the investigations have been sufficiently definitive, however, to Justify the statement that, relative to the existing residual, the aggregate errors in the statistics of merchandise exports and imports and the estimates of the various service items probably are not material. See also footnote 4, p. 4.

longer represent foreign investments in the United States. The effect of such a change in the reported statistics of capital movements is the same as if the funds were actually withdrawn from the United States through the exchange market, yet in reality no outward movement occurs. It was estimated in the balance-of-payments study covering 1939 that such transfers were between $200,000,000 and $300,000,000 in that year.45

Other shifts of this general character tend to reduce the estimates of foreign investments in the United States without involving any movement of capital. For example, securities held by refugees before coming to this country would likely enter into the investment estimates, but when brought to the United States and registered under a domesic address they would not be classed as foreign investments. In this manner foreign investments are reduced but the movement of the securities would not be recorded as a capital movement.

It is probable, as was often reported in the press, that there was a considerable unrecorded export of United States currency in 1940. To the extent that such currency may have been obtained by the use of foreign balances, the movement would have the same effect on the recorded capital movement as the change of address of refugee funds. Axis funds, whether registered as such or not, may have been withdrawn in this manner-possibly for export to Latin America—in anticipation of the possible extension of the freezing orders.

The inflow of hot money and funds expatriated from abroad by refugees to the United States during the last 5 or 6 years presumably included some funds for investment in real estate and in small business enterprises. Such transactions are not included in the regular reports of capital movements and no adequate basis exists for a reasonable estimate of their volume. The net effect on the balance of payments is that of unreported capital imports whether or not the funds enter into the reported foreign balances before the transaction is consummated. Complete data on this subject would tend to increase the estimates of foreign investments in the United States.

In the last 7 years the residuals have aggregated about $4,500,000,000 and, as stated above, a considerable part of this item is composed of unreported capital transactions although some of it may be due to errors and omissions in the merchandise and service estimates in the balance of payments. The probable existence of large unreported capital transactions between the United States and foreign countries does not necessarily mean that the net creditor position is in error to the extent of $4,500,000,000, nor even to the extent of the unreported capital inflow, whatever amount that may be. Some of the residual might be accounted for by the inflow of American capital from abroad not previously reported in the estimates of United States investments in foreign countries. There are also the funds of refugees, which had their status changed from "foreign" to "domestic" investments by a simple change of address by the owners. In addition, there may be some unreported and unestimated Government transactions that affect the balance of payments in the same manner, but do not change the estimated totals of private foreign investments.

45 It is appropriate at this time to point out specifically what is implied in this paragraph, namely, that the so-called "capital movement" statistics. insofar as they relate to short-term funds, are not actually measures of the flow of such funds. Instead, they represent the differences from date to date in the outstanding amounts of foreign-owned short-term-balances reported by the banks and other reporting agencies.

Table IX.-International Investment Position of the United States,' End of 1938, 1939, and 1940

1 See tables XI, XII, and XIII, pp. 52, 56, and 61 for additional details regarding these private foreign investments.

2 Revise.l.

3 Using market values instead of par values for bonds and preferred stocks, the net creditor position of the United States in 1938, 1939, and 1940, would be reduced by $633,000,000, $698,000,000, and $1,064,000,000, respectively.

NOTE: These estimates are exclusive of intergovernmental debts-war debts and the direct loans of the Export-Import Bank.

Of equal significance in this connection is the fact that the estimates of long-term investments presented herein, and in other recent bulletins of the Department of Commerce, were not based on the capital movement statistics. They were based, rather, on independent investigations calculated to permit reasonably accurate estimates of the value of the investments on specified dates in accordance with ccrtain stated methods of valuation. The resultant values did not and could not correspond with the capital movement estimates required for balanceof-payments purposes. The investment estimates were affected by changes from year-end to ycar-end in the market values of the securities held by foreigners while the capital movement data are an aggregate of the cash values at which transactions took place. For these and other reasons, the residual is not as important in relation to the creditordebtor position as it is in relation to the annual transactions comprising the balance of payments.

Creditor Position at New Low.

As of the end of 1940, the net creditor position of the United States on private account amounted to only $1,486,000,000 (see table IX), the lowest in many years. Our total assets abroad, at $11,181,000,000, were only 15 percent in excess of our liabilities. On an income basis, however, this country still remained a net creditor by a substantial margin. During 1940, it is estimated, the United States received dividends and interest from its investments abroad amounting to $525,000,000, and paid $195,000,000 to foreigners from their holdings here (see table X). Our receipts of income, therefore, were 169 percent greater than our payments.

Less than 4 percent of United States investments in foreign countries were short-term holdings, while this same category constituted 41 percent of total foreign investments in this country. In most parts of the world, short-term money has during recent years had but little earning power. Political instability and uncertainty as to the future have universally resulted in a premium on liquidity. Vast transfers of funds to what appeared to be the safest financial centers were in

fluenced much less by prospective yield than by the desire for security. New long-term investment has been supplied largely by government funds rather than by private capital. The relative superfluity of short-term funds has caused a great gap to exist between its earning power and that of existing long-term capital investment.

The income from the large foreign deposits which came to this country for safekeeping was almost nil. At the same time the rate of yield on foreign long-term holdings in this country-highly concentrated in manufacturing, public utility, and transportation securities-was less than that on United States long-term foreign holdings abroad, which included a high percentage of enterprises in the extractive industries. These facts explain most of the wide difference that exists between the net creditor position of the United States as shown, respectively, by the investment data and by the estimates of income receipts and payments (tables IX and X).

Table X.-Receipts and Payments of Income on United States International Investments, 1938-40

Income to foreigners from short-term investments in the United States.

UNITED STATES INVESTMENTS ABROAD

United States investments in foreign countries are of three main types. The largest group consists of "direct investments"-investments through which a good measure of control over business enterprises in foreign countries accrues to United States investors. Next in size are the "portfolio investments"-principally holdings of foreign dollar bonds and miscellaneous securities which do not involve any element of control over the foreign borrowing entities. The smallest group, the so-called short-term investments, includes primarily loans to foreign banks and other short-term obligations. The estimated total of these three types of investment at the end of 1940 was $11,181,000,000, and their average yield in 1940 was 4.7 percent (tables IX, X, and XI).

Direct investments, estimated at the end of 1940 at $7,180,000,000,** yielded an income of $395,000,000, equivalent to an average return of

46 A new questionnaire study of United States direct investments in foreign countries is being made. The results may alter the estimates made at this time.

5.5 percent. These investments have been subjected in recent years to diverse influences, such as expropriation, the freezing of capital, restrictions and prohibitions in the transfer of earnings, on the one hand, and the stimulation of business arising out of war preparations, on the other hand. Although most of these factors have been unfavorable, the total annual income yielded has declined by less than 10 percent since 1937. The decline has not been related to the volume of these interests, inasmuch as new direct investments abroad, except in the form of advances to established enterprises (see appendix A, table 1), as well as resales to foreigners, were almost nil.

47

Portfolio investments, estimated at $3,591,000,000 in 1940, returned an income of $124,000,000, or an average yield of 3.45 percent. Of these "portfolio" investments, $3,141,000,000 were holdings of foreign dollar bonds, of which 31.3 percent, or $983,000,000, were in default of interest payments. These bond holdings were reduced during the year by the expenditure of about $59,000,000 by foreigners for the redemption of bonds held in this country, and by another substantial sum, recorded to the extent of $25,000,000 but undoubtedly much higher, spent to repatriate foreign dollar bonds.

Short-term investments were not only reduced to the very low level of $410,000,000 (table XIII), but, as a result of the abundance of such money, the estimated average rate of return of 1.46 percent was likewise very low.

More than seven-tenths of the United States investments abroad are located in the Western Hemisphere-about equally divided between the areas to the north and those to the south-while less than twotenths are in Europe. In other words, only a small percentage of our foreign holdings are in the area of present hostilities and the bulk of them are in neighboring or contiguous countries. For several years the concentration of income from the countries of the Western Hemisphere has been even more notable than that of investments, as a result of the greater severity of restrictive measures in continental Europe, which affected both major types of investments. Exchange controls prevented the transfer of much of the income from direct investments, and caused defaults on a large part of the dollar bonds held in this country.

Since the start of the war, publicly offered (private) foreign loans have been almost completely eliminated, partly because of the restric tive effects of the Johnson Act and the Neutrality Act. In addition, private capital has been unwilling to assume the risks involved in loans to those borrowers outside the scope of the restrictive legislationprincipally the Latin American countries-because of the large volume of issues at present in default.

For these reasons, the Reconstruction Finance Corporation (through the Export-Import Bank) and the United States Stabilization Fund have made virtually all of the recent foreign loans. The Bank's commitments rose by $300,000,000 during 1940 to $655,000,000, and total loans outstanding rose from $65,000,000 to $131,000,000 during the same period. Even larger increases occurred during the first 6 months of 1941, and in addition, the Stabilization Fund contracted two loans of $50,000,000 each-one to Argentina and one to China.

For additional details see FOREIGN COMMERCE WEEKLY, July 19, 1941, pp. 3-5, "Status of United States Investments in Foreign Dollar Bonds at the End of 1940.'