U.S. Direct Investment Flows with Selected Caribbean OFCs

International Monetary Fund
Published Date:
June 1992
  • ShareShare
Show Summary Details
Werner Hasenberg and Neil Patterson 

The Working Party’s Report on the Measurement of International Capital Flows described problems that arise because offshore financial centers (OFCs) omit from their balance of payments the large volumes of financial flows that pass through their entrepôt facilities.1 OFCs are small countries that consider these flows to be unrelated to activities in their own economies and inappropriate for their balance of payments statements. Also, some OFCs do not prepare balance of payments statements at all. As a result, capital flows in the Fund’s Balance of Payments Statistics Yearbook do not include most financial transactions of OFCs.

Omission of these flows would not affect global balances of the capital account if OFCs acted only as clearinghouses for transactions that pass through their countries. However, transactions change form as they pass through OFCs, perhaps entering as portfolio investment or interbank flows and leaving as direct investment transactions. Moreover, offshore activities may produce sizable net earnings that accrue to direct investors abroad but are retained in the OFCs. Although OFCs omit such flows from their balance of payments statements, partner countries report them as transactions with OFCs. In order to balance flow statistics at the world level by type of capital (such as direct investment or portfolio investment), the OFC entrepot flows must be added to world totals. The Working Party made estimates for these flows—see Chapter 3 (direct investment), Chapter 4 (portfolio investment),Chapter 5 and 6 (other capital), and Chapter 9 (OFC flows) of the Report on Capital Flows. These estimates were derived using data bases on international bond and banking statistics (as compiled by the Bank for International Settlements and the IMF). For direct investment, geographic data were compiled by partner countries.

In order to estimate the missing flows, the Working Party first needed to understand the nature of the OFC flows and the measurement practices of the international organizations and partner countries involved in the compilation process. The Working Party undertook two detailed investigations of certain direct investment flows between several OFCs—the Netherlands Antilles and a group of other Caribbean nations—and the United States. Because these flows—as measured by U.S. compilers—were very large, their omission from OFC reports was seen as a significant source of the global statistical discrepancy on direct investment.

U.S. Direct Investment in Netherlands Antilles


For more than two decades, the Netherlands Antilles has been an intermediate point for direct investment to and from the United States. The largest component of such direct investment transactions occurred because the Netherlands Antilles served as a tax haven. Investors could rely on the income tax treaty between the United States and the Netherlands Antilles, similar bilateral tax treaties between the Antilles and other countries, the existence of bank and commercial secrecy laws, and a well-developed financial system to execute transactions efficiently.

Such transactions have included both direct and other types of investment. The Working Party’s analysis, however, was limited to an exploration of the direct investment of U.S. corporations in finance affiliates in the Netherlands Antilles, especially in light of the statistical treatment of such investments by the U.S. Bureau of Economic Analysis.

Technical Note: The sign conventions used in this paper may need some elaboration. In the U.S. statistics, a direct investment position is described as positive when the aggregate amount of equity capital (including reinvested earnings) and intercompany debt owed by foreign affiliates to their parents is a positive amount; the position is described as negative when net intercompany debt owed by parents to their affiliates exceeds the total amount of the parents’ equity in the affiliates.

For direct investment capital transactions, a negative value usually denotes a net withdrawal of investment by parents from their affiliates. However, if a parent company draws on a loan extended by its affiliate, this transaction is recorded with a negative value under net intercompany debt. Similarly, when the parent company repays the loan extended by the affiliate, this transaction is recorded with a positive value under net intercompany debt.

For income on U.S. direct investment abroad, a negative value denotes an excess of income paid by parents to their affiliates over income received by the parents from their affiliates.

Werner Hasenberg worked as a consultant to the Working Party. Neil Patterson was a member of the Working Party’s technical staff. He is now a Deputy Division Chief in the IMF’s Statistics Department.

Institutional Factors

The existence of Netherlands Antilles finance affiliates (NAFAs) goes back, at least, to the late 1960s and is predicated on the 1948 Income Tax Treaty between the United States and the Netherlands. The treaty was extended to the Netherlands Antilles in 1955. The Antilles treaty was amended by a protocol in 1963 and modified by a convention in 1965. Between the mid- 1960s and the mid-1980s, the Netherlands Antilles served as a major tax haven by virtue of its commercial secrecy laws, its low tax rates for financial institutions, and a network of tax treaties, including the provisions in the U.S. treaty that eliminated or reduced the U.S. tax on “U.S. source income” (such as interest, dividends, and royalties received by foreigners from a U.S. entity). Under the treaty, there was a zero tax rate on all interest and royalties payments, a 5 percent rate on direct investment dividends, and a 15 percent rate on portfolio investment dividends. However, the 1965 modification required that to be eligible for the treaty rates the Netherlands Antilles affiliate receiving the payments had to be subject to the normal Netherlands Antilles income tax (30 percent) rather than to the “offshore tax” (3 percent). In most cases, this differentiation was effectively meaningless, since the taxable income of finance affiliates was very low (largely because most of their costs consisted of interest payments to lenders in the Eurobond market).

In the Deficit Reduction Act of 1984, the U.S. Congress repealed the provisions of the Internal Revenue Code that had imposed a 30 percent nonrefundable “withholding tax” on interest paid by U.S. borrowers to nonresident aliens and to foreign corporations in nontreaty countries. This action removed the comparative advantage of the Netherlands Antilles. Before the 1984 act, over 200 U.S. corporations had established finance affiliates in the Antilles, through which they had borrowed large amounts in the Eurobond market at favorable interest rates.

The largest borrowing by U.S. corporations through their NAFAs took place in the periods 1968-74 and 1978-84. During the earlier period, U.S. corporations borrowed funds in the Eurobond market through their Antilles affiliates and repatriated the borrowed funds to the United States in compliance with conditions imposed on them individually under the Foreign Direct Investment Regulations. (These regulations, which were terminated in 1974, set annual limits on direct investment abroad). Large borrowings in the second period were undertaken by U.S. corporations in response to the credit crunch of the late 1970s and early 1980s. These borrowings occurred mostly to finance domestic corporate operations and generally were not related to other U.S. direct investment activities abroad.

Negotiations for a new tax treaty between the United States and the Netherlands Antilles took place between 1984 and 1987. In mid-1987, the United States moved to terminate the treaty, with 1987 being the last year of full benefits.

BEA Data

Since 1977, BEA has collected data on U.S. corporate borrowing in the Eurobond market via NAFAs as part of its data collection system on U.S. direct investment abroad (USDIA). Previously these data were collected and published as portfolio investment; the change in classification followed the 1977 benchmark survey on USDIA, which indicated that these data would be more accessible as part of the direct investment data series.

The U.S. direct investment data have shown a large “negative direct investment position” for the Netherlands Antilles, preponderantly in the finance sector. BEA has published these data in detail in order to adjust for the negative position amounts in the aggregate date series on USDIA.

Table 1 illustrates dramatically the impact of the Deficit Reduction Act of 1984, with reversals between and 1985 in sign and trend for many of the items. For purposes of analysis, the data have been divided here between intercompany debt and equity capital.

Intercompany Debt

Table 1 presents various aspects of intercompany debt; levels of debt, capital flows on such debt, net interest income on intercompany debt, interest payments on NAFA bonds, changes in U.S. parents’ receivables, and changes in U.S. parents’ payables. For the purpose of analyzing matters related to NAFA borrowing, lines 2c(3), 3b, and 4 are most significant.

“Changes in U.S. parents’ payables” (line 2c(3)) indicate that U.S. intercompany debt increased significantly in 1983 and 1984 (inflows of $5.4 billion and $3.8 billion, respectively) as NAFA borrowing continued. However, as NAFA borrowing began to be repaid, there was a sudden reversal in the debt situation in 1985 (an outflow of $3.5 billion) and continuing large outflows of more than $21.2 billion for the years 1986 through 1990. Line 3b illustrates the corresponding trend in net interest payments: a peak of $5.2 billion in 1984 and a progressive decline thereafter, to $1.7 billion in 1990.

Table 1.U.S. Direct Investment Abroad: Selected Transactions with and Positions in Finance Affiliates in the Netherlands Antilles, 1983-90(In billions of U.S. dollars)
1.Position (year end)−23.3−25.1−20.8−17.2−14.5−10.3−6.2−1.7
a. Equity15.016.916.716.415.312.111.010.7
b. Intercompany debt
(1) Net−38.3−42.0−37.5−33.7−29.8−22.5−17.1−12.4
(2) U.S. parents’ receivables0.
(3) U.S. parents’ payables−38.7−42.5−38.6−34.2−30.4−23.1−18.6− 15.7
2. Capital transactions (inflows (-))−3.1−
a. Equity capital1.41.0−0.90.7−1.4−2.6−0.6−0.4
b. Reinvested earnings1.00.90.9−
c. Intercompany debt
(1) Net−5.5−
(2) Changes in U.S. parents’
(3) Changes in U.S. parents’
3. Income transactions (payments (-))−3.2−3.4−3.1−2.7−2.1−1.9—1.4−1.0
a. Earnings1.
b. Interest−4.6−5.2−4.8−3.9−3.3−2.8−2.2−1.7
4. Memorandum:
Interest on NAFA bonds (payments (-))−3.5−2.9−2.2−1.7−1.2−0.6−0.3
Source: Lines 1 through 3—Table 5 in U.S. Bureau of Economic Analysis, Survey of Current Business. August 1991; line 4—estimates.Note: This table shows transactions with, and positions in, affiliates primarily established to borrow funds abroad and relend them to their U.S. parents.
Source: Lines 1 through 3—Table 5 in U.S. Bureau of Economic Analysis, Survey of Current Business. August 1991; line 4—estimates.Note: This table shows transactions with, and positions in, affiliates primarily established to borrow funds abroad and relend them to their U.S. parents.

Broad estimates of U.S. interest payments related to NAFA bonds may also be made for each year since 1984. Such estimates can be linked to a baseline estimate of the level of NAFA bonds outstanding at year-end 1984 (about $28 billion) and reduced each year by an amount somewhat below the changes in U.S. parents’ payables. The annual interest payments on the bonds, estimated by multiplying an average NAFA bond interest rate by the derived amount of bonds outstanding, have also declined (from about $3.5 billion in 1984 to about $0.3 billion in 1990) as outstanding NAFA bonds have matured.

Equity Capital

Table 1 shows levels of equity, equity capital flows, reinvested earnings, and equity earnings. Larger reductions in U.S. equity in NAFAs might have been expected after 1984, because large amounts of intercompany debt were liquidated. Most equity investment in NAFAs presumably occurred to conform with certain Internal Revenue Service debt-equity rules as a condition for the deductibility of interest payments. (The debt-equity ratio of the NAFAs, line lb(3) divided by line la, progressively fell from 2.6 in 1983 to 2.1 in 1986 and 1.5 in 1990.) As might be expected, U.S. equity earnings in NAFAs reached their peak in 1984 and 1985, when equity was also at its highest level.

Summary and Outlook

The data examined here describe the peak and the fall of Netherlands Antilles finance affiliates as a major vehicle for U.S. firms’ borrowing in the international capital markets. The highest levels of debt and equity stocks were reached in 1984. The borrowing activities of NAFAs declined after the Deficit Reduction Act of 1984 was passed and the United States terminated its income tax treaty with the Netherlands Antilles in 1987.

By the end of 1990, the U.S. direct investment position in Netherlands Antilles finance affiliates had dropped to -$1.7 billion. Although NAFAs are still used as a conduit for international transfers of funds (significant capital flow activity through NAFAs was recorded in 1991), it may be assumed that by the end of 1992 they will have ceased to function as a major vehicle for offshore borrowing by U.S. corporations—previous borrowing under the tax treaty will essentially have been repaid.

Direct Investment in the United States from the U.K. Islands, Caribbean


Data on foreign direct investment in the United States, as compiled by the BEA, indicate relatively large and volatile entries with respect to direct investment from a group of Caribbean nations—called the U.K. Islands, Caribbean (UKIC). The BEA’s grouping of UKIC covers the British Antilles, British Virgin Islands, Cayman Islands, Montserrat, and Turks and Caicos Islands. Most of the UKIC direct investment in the United States comes from the Cayman Islands and the British Virgin Islands; no separate data are available.

This analysis is based on an examination of the U.S. data, especially with respect to the possible identification of ultimate beneficial ownership (UBO) of such investment. As such, it may help to trace counterpart data in the statistics of the UBO countries and to understand the size and volatility of UKIC direct investment in the United States in recent years.


The volatility in UKIC direct investment capital flows into the United States in the 1987-90 period is attributable mostly to capital flows between UKIC parents and their U.S. affiliates in the finance sector. As Table 2 indicates, the data show even greater volatility for U.S. finance affiliates than for U.S. affiliates in all industries.

Table 2.UKIC Direct Investment in the United States, Selected Position and Transactions Data, 1987-90(In billions of U.S. dollars)
Position (year end)
All industries−3.3−3.9−0.1−3.2
Capital transactions (outflows (-))
All industries−−3.5
Intercompany debt transactions
(outflows (-))
All industries−4.4−0.53.4−2.9
Source: U.S. BEA, Survey of Current Business, August 1991.

Suppressed to avoid disclosure of data on individual companies.

Source: U.S. BEA, Survey of Current Business, August 1991.

Suppressed to avoid disclosure of data on individual companies.

At year-end 1990, the UKIC position was -$3.2 billion, but the position for the finance sector was -$4.8 billion. These negative positions were considerably larger than at year-end 1989 (when they were - $0.1 billion and - $1.5 billion, respectively) but virtually the same as at year-end 1987 -$3.3 billion and -$4.8 billion).

Reported outflows in 1990 were $3.5 billion ($3.3 billion in the finance sector), following capital inflows in 1989 of $3.2 billion (almost entirely in finance). Large capital outflows also occurred in 1987 ($3.8 billion, of which $4.5 billion was in finance). Intercompany debt transactions ($4.4 billion outflow in 1987, $3.4 billion inflow in 1989, and $2.9 billion outflow in 1990) dominated the capital flows.

Institutional and Anecdotal Aspects

According to anecdotal information, UKIC direct investment in the U.S. financial sector relates mostly to U.S. finance affiliates established to raise funds in U.S. capital markets (usually through the issuance of commercial paper) for transmission to the immediate UKIC parent. Typically, the UBO of the U.S. finance affiliate is a foreign bank (often in Australia or Western Europe) that has long had a U.S. banking affiliate in the form of a branch or agency. Raising funds through U.S. finance affiliates of UKIC parents, rather than having the U.S. bank affiliate raise funds directly through deposits, presumably is less costly, since finance affiliates do not have to abide by reserve requirements for these funds and do not have to pay Federal Deposit Insurance Corporation premiums for them.

The funds raised by the U.S. finance affiliate may in turn be lent by the UKIC parent to the U.S. bank affiliate or to the UBO parent (or its affiliates in other countries), all of which would be regarded as direct investment transactions. Alternatively, the funds may be used for portfolio investment in the United States or elsewhere, indicating a transformation from direct investment to portfolio investment via UKIC institutions. It may be difficult or even impossible to find data series that would support this anecdotal information, though it may be valid.

Concluding Comment

The preceding investigations illustrate the large volume, great volatility, and considerable diversity of international capital transactions that occur through the entrepot facilities of OFCs. This analysis of U.S. direct investment flows was used by the Working Party to estimate counterpart flows—missing from the global direct investment accounts—of the Netherlands Antilles and UKIC. These estimates formed a significant component of the Working Party’s adjustments—discussed in Chapter 3 of the final report—for missing OFC direct investment flows.

    Other Resources Citing This Publication