In accordance with the directive in its Terms of Reference (page xiii), the Working Party has examined the effects on balance of payments accounting of both the proliferation of offshore financial centers and the spreading use of new financial instruments. However, as it would go beyond the scope of this report to present a detailed analysis of the complex nature of cross-border transactions involving such centers and instruments, this section looks primarily at the distortions which these transactions may cause in the international investment income account, with brief reference to other types of transactions. Some estimates of the extent to which transactions performed via offshore centers have contributed to the discrepancies in the global investment income account are given in earlier chapters. The net effect of the various adjustments found to be necessary for the income accounts was small in the years covered by our study, but there were sizable omissions of both debits and credits. The size of these omissions has probably grown substantially since 1983. Innovations in financial techniques have created statistical problems of increasing severity that will require some rethinking of the standard categories used in the balance of payments accounts and changes in data collection techniques.


a. Principal Features

Mainly for historical reasons, but also because of different characteristics, two broad categories of international financial centers (IFCs) can be distinguished: the traditional—or classical—IFCs located in Europe and the United States, which evolved during the nineteenth and early twentieth centuries, and the offshore financial centers (OFCs), which are a relatively recent, post-World War II phenomenon.

The traditional IFCs emerged in response to demands for financial services related to interregional and international trade and investment. Their growth was stimulated by access to ample sources of capital, the development of professional and technical skills, good communication facilities; and, more generally, the business and personal connections required to provide the banking, insurance, and other financial services needed by traders, shipping firms, and investors. In addition, these centers enjoyed political stability and laissez-faire policies which allowed unfettered trade and capital flows. Activity in the IFCs—which had dwindled in the years of depression in the 1930s—began to expand gradually after World War II and received a major impetus from the advent of the Eurocurrency markets in the early 1960s, which greatly enhanced their appeal as low-cost and highly efficient conduits for transactions with and among nonresidents.

The OFCs are of relatively recent origin and owe their existence to the deliberate attempts of the authorities in these centers to offer a range of advantages which were not available, or not available to the same extent, in the traditional IFCs. The various incentives provided in the OFCs attracted captive insurance companies, shipping companies, and other nonfinancial firms to set up management and holding companies in these centers and to shift a part of their operations to them. More importantly, however, favorable and flexibly administered tax laws and exchange controls, as well as the virtual freedom from banking and other regulations or restrictions on operations, induced banks to transfer business to these centers. Indeed, the very large volume of international banking business, which is completely unrelated to the size of the domestic market, is the single most prominent feature of all OFCs. The banks which engage in offshore business in these centers can be broadly divided into two groups. First, there are “letter-box” or “brass-plate” companies which act as legal and booking channels for business that is in fact conducted elsewhere—to a large extent in the traditional financial centers located in the countries of residence of the banks which control the letter-box companies. The second group of banks has genuine transactions offices—that is, they conduct banking operations consisting of deposit banking and final lending with business conducted with at least some measure of independent decision making.

While all international financial centers offer certain basic services, such as relative freedom from some taxes and regulations, anonymity, and efficiency in effecting transactions, there is also a considerable amount of specialization in the attractions offered to customers. The older centers provide a wide range of services though, of course, London is clearly the dominant center in the Eurobond market, while Swiss banks are preeminent in offering custodial and other services to their depositors.

Among the OFCs, there is a general tendency to impose no income taxes on financial intermediaries or their customers or to have only nominal income tax rates. Few have any capital gains taxes; commonly there is no taxation of foreign-source income; and individuals taking up residence may be free of estate taxes. In addition to the use of widespread and sometimes specially designed tax incentives, some OFCs have established favorable regimes for particular industries. Thus, for example, insurance operations are centered in Bermuda and, to a lesser extent, in The Bahamas, the Channel Islands, and the Isle of Man, while Liberia and Panama have been established for many years as advantageous countries in which to register ships. In recent years many companies, especially U.S. companies, found that because of the tax situation, it was advantageous to conduct their international borrowing operations via a specialized subsidiary in the Netherlands Antilles or subsidiaries in some other offshore centers. Changes in the tax situation have now greatly reduced that advantage. In general, however, it seems fair to say that any interference with market efficiency that results from attempts to tax or regulate will quickly result in the creation of some special provision in some OFC which will attempt to neutralize the effect of such actions. As these new channels for flows are opened, usually without any attempt at balance of payments accounting, the task of the balance of payments compiler is made harder.

Some statistics can be cited that will illustrate the size and growth rates of the international business of the main financial centers. In Table 60 the growth of the foreign assets of banks located in both old and new financial centers can be seen, as they doubled from $1,000 billion in 1979 to $2,100 billion at the end of 1985. The increase in the share of MOB centers can also be seen, and it would have been even greater (and the share of the United States substantially less) except for the switch of funds to head offices located in the United States that occurred when international banking facilities were initially allowed in the United States at the end of 1981.

Table 60


(In percent)

article image
Source: International Monetary Fund, International Financial Statistics (IFS) Yearbook, Vol. 39 (Washington, 1986).Note: Coverage of the series may change over the period.

Including trustee accounts channeled through Swiss banks.

Including international banking facilities (IBFs).

The “major offshore banking centers,” according to IFS definition, are as follows: Hong Kong, Singapore, Bahrain, The Bahamas the Cayman Islands, the Netherlands Antilles, and Panama.

In Table 61, the phenomenal growth of the foreign assets of banks located in each of the major OFCs is shown. This table also discriminates to the extent possible between the assets held by banks carrying out a general banking business and institutions specially organized to carry out offshore business, though the distinction is obviously not always meaningful.

Table 61


(In billions of U.S. dollars)

article image
Source: International Monetary Fund, International Financial Statistics Yearbook, Vol. 39 (Washington, 1986).Note: Coverage of the series may change over the period.

All deposit-taking institutions can engage in offshore banking. No separate category is restricted.

Break occurred in series in 1978.

Includes assets of nonmonetary financial institutions.

No detail available to distinguish between assets of Class A licensed banks, which can engage in both domestic and foreign operations, and Class B licensed banks, which can engage only in offshore operations.

Estimate based on partial data.

Less than $50 million.

b. Relation to Statistical Discrepancy

Several kinds of statistical problems result from the operations of international financial centers, and the problems are likely to grow with the passage of time as the business of international finance becomes increasingly sophisticated. One problem is that some MOB centers do not include in their balance of payments accounts the transactions of certain intermediaries that they regard as essentially nonresident. This aspect is described in the section on MOB centers (Subsection 2.b.vi) in Chapter V and is also discussed in Chapter I, which indicates the adjustments to the international investment accounts that are required in order to bring these economies into line with the reporting done by the rest of the world. About $6 billion was missing from reported income debits and credits in 1983. This amount would be larger now, and it does not indicate how much of an adjustment might be necessary for non-investment income account items such as insurance premiums, fees charged by financial institutions, or revenues of registered ships not reported elsewhere. Table 62 shows for individual MOB centers the extent of the revisions required for bank-related portfolio investment income.

Table 62


(In billions of U.S. dollars)

article image

Balance of payments transactions recorded in the Fund’s Balance of Payments Statistics Yearbook, 1985, Part 2. For Hong Kong, Yearbook figures are Fund estimates, for The Bahamas. Yearbook data were revised by the response to the investment income questionnaire and reclassified on the basis of Fund staff estimates; and for Singapore, Yearbook data were revised on the basis of Fund staff estimates.

Based on amounts outstanding (taken from International Financial Statistics), and on cross-border interest rates. (See Chapter V, Table 38.) Does not include income related to investments in bonds and equities (Table 39).

Does not include interest payments on bank liabilities vis-à-vis nonresident nonbanks.

Another kind of problem is that as funds flow through these centers, their classification is often transformed because of incomplete reporting. For instance, premiums paid from the United States to a captive insurance company in Bermuda and placed in a bank account in London will not be reported at all by Bermuda but will (collectively) appear as an insurance debit in the U.S. current account and a short-term capital account inflow in the U.K. balance of payments. Thus, the OFCs generate discrepancies in both the current and capital accounts, or sometimes asymmetries as between different elements of the current account.

Furthermore, in many cases the MOB centers’ “normal” residents also have substantial foreign interests that are not reported by the MOB centers but are recorded in the accounts of partner countries. These interests include direct or portfolio investments that generate income, including the interest on official reserves, or substantial shipping operations.

Another problem arising from operations of IFCs is the question of whether these countries or economies should be required to report the transactions of the intermediaries operating in their territories on a gross or net basis. In principle, they should report on a gross basis, according to the Fund’s Balance of Payments Manual, but in practice this may tend to inflate these transactions’ significance in the true savings and investment pattern of the world economy. The same question arises, of course, when considering the measurement of capital flows via IFCs or the presentation of banks’ international balance sheets. One alternative would be to introduce net entries under the appropriate transactions accounts in balance of payments statements, but to also supply gross flows as memorandum information.

Finally, the use of IFCs by international investors and borrowers tends to obscure the geographic patterns of capital and income flows, and by the same token it eliminates the possibility of detecting discrepancies by making direct comparisons of bilateral accounts. This problem is far from new, but it is becoming greatly intensified not only by the greater use of IFCs but also by the accelerated introduction of innovative financial instruments, which is discussed below. In combination, these phenomena cast doubt on traditional methods of recording international capital and income flows in the international accounts. Since many OFCs offer opportunities to minimize reporting requirements or obscure the identification of beneficial parties to transactions, it is unlikely that they can be converted to becoming full reporters of their balance of payments accounts. Wherever possible, therefore, the Fund staff should prepare estimates for inclusion in aggregated tables published in Part 2 of the Yearbook.


Since the early 1980s the emergence and growing use of new types of financial instruments or adaptations of existing ones have been important features of the conduct of financial transactions, both in the domestic markets of some major countries and, in particular, in international markets. The report Recent Innovations in International Banking, which was published in April 1986 by the Bank for International Settlements, describes the principal instruments and their background and discusses the implications of financial innovation for macroeconomic policies, prudent bank management, and financial stability. The report also examines some of the consequences of innovation for financial reporting and its impact on international banking statistics, but it does not focus on the issue of whether and to what extent the spread of new financial instruments in cross-border transactions might give rise to discrepancies in the balance of payments accounting for capital flows and associated income streams.

The upsurge of financial innovation in recent years seems to involve a cumulative learning process designed to exploit market opportunities and improving technology, plus reactions to governmental actions with respect to macroeconomic policy and financial regulation. Market participants were prompted to innovate by many factors, such as shifts in the pattern of net flows of international savings and investment, the volatility of nominal and real interest rates and exchange rates, the deregulation of national financial markets, the repercussions of international debt problems on the creditworthiness of borrowers and the attitudes of their main creditors, the introduction of more stringent capital requirements for banks, rapid technical progress in the fields of communications and data processing, and strong competition in international financial markets. Indeed, by allowing interest, liquidity, and credit risks to be priced separately and negotiated in separate markets, the new instruments have greatly enhanced the possibilities for investing, borrowing, hedging, and speculating and have thus contributed significantly to the rapidly expanding volume of international financial transactions in recent years. However, as such transactions became more diversified, with different types of transactors becoming involved and a much greater volume of transactions taking place, it also became more difficult to capture and correctly identify and classify the corresponding balance of payments capital and income flows. Consequently, financial innovation has become a factor in the world current account discrepancy.

Some indication of the pace of innovation is given in Table 63, which shows that newer instruments of credit intermediation, such as floating rate notes (FRNs), accounted for only about 6 percent of activity in international credit markets in 1981 but for about 30 percent by 1985. The steep rise in the volume of fixed rate bonds issued was partly attributable to the introduction of new features of bond finance, such as zero-coupon bonds, multicurrency bonds, convertible bonds, and bonds with warrants attached to them but also reflects the increasing use of swaps to tailor the features of these underlying instruments to the needs of market participants. In 1983, the base year for this report, use made of the newer instruments was still rather limited, so that the observations made here are directed primarily to these instruments’ potential for creating discrepancies rather than their possible effects on financial markets in our base year.

Table 63


(In billions of U.S. dollars)

article image
Source: Data supplied by the Bank of England, as shown in Bank for International Settlements, Recent Innovations in International Banking (Basle, April 1986), Table 5.1, p. 130.

Excludes US takeover-related stand-by credits.

Includes revolving underwriting facilities, multiple-component facilities (if they include a note issuance option), and other Euro-note facilities. The facilities may not be fully utilized, or may be utilized at a later date.

The use of new financial instruments is likely to pose two major problems for balance of payments compilers. First, some of the new instruments facilitate international credit flows that bypass banks’ balance sheets, so that flows that had been fairly comprehensively and reliably captured under existing reporting systems based on banks’ records now tend to escape reporting. Second, a number of new instruments, as well as the increased use of swaps, are designed to manage interest and exchange risks; and while their use does not necessarily give rise to errors in capturing capital and interest flows, they generate larger payments and receipts of fees, commissions, and other service payments which should be included in balance of payments reporting but which are difficult to estimate or compile.

The shift toward less easily recordable transactions is clearly demonstrated in Table 63 by the rising share of new lending in the form of FRNs or involving note issuance facilities (NIFs) in total international credit intermediation. Briefly, FRNs carry a rate of interest which is reset at regular intervals in relation to a predetermined reference rate (such as LIBOR) and thus help to reconcile borrowers’ needs for medium-and long-term finance with investors’ preferences for liquid assets with a low price risk. An NIF is a medium-term, legally binding commitment under which a borrower can issue short-term paper in its own name, with the underwriting banks committed either to purchase any notes which the borrower is unable to sell or to provide stand-by credit. For bank borrowers the paper is usually short-term certificates of deposit, while for nonbank borrowers it is in the form of promissory notes (commonly known as Euro-notes). Holders of these notes (whether or not they have underwritten the facility) show them as an asset on their balance sheets, but an underwriting commitment normally does not appear on the balance sheet of the banking institution involved. More recently the arrangement for NIFs has been developing into a Euro-commercial paper market—that is, high-grade borrowers are able to raise the funds in the form of short-term paper without a directly associated credit backing (underwriting commitment) by banks.

There are many variants of short- and long-term financing that now take the form of securities, but the main characteristic that affects balance of payments accounting is that instead of the relatively clear record of cross-border assets and liabilities that appears on banks’ balance sheets when they act as principals in the extension of credit and as deposit takers, there is now only a much more amorphous record related to credit arrangements concluded on behalf of nonbanks, with banks underwriting, or simply accepting a general undertaking to arrange for, the placement of notes for a borrower. Of course, banks themselves also make active use of the securities markets and carry on their balance sheets sizable foreign positions reflecting the issue and purchases of securities.

The substitution of marketable financial instruments for direct bank lending, known as “securitization,” together with the enhanced liquidity of the new instruments, means that it is more difficult at any given time to identify the cross-border assets or liabilities of residents. In fact, often the issuers of marketable paper and sometimes even the holders of such paper are not aware that the counterparty is a nonresident. Banks’ records will still be useful in compiling information on the amount of new borrowing by residents that is to be conducted in international markets, but not necessarily on the amounts actually taken down, the residence of the creditors, or the amount outstanding at any given time. Moreover, whenever securities held in the market are acquired by banks, there will be an increase in bank lending that could be misleading if the disinvestment by the nonbank holder of the security is not also accounted for. Finally, additional difficulties might be created by the issue of securities by special-purpose subsidiaries set up in the foreign financial centers.

Experience with the U.S. system, which is based on reporting of resident/nonresident positions by banks and nonbanks, already showed evidence of difficulty in recognizing cross-border borrowing and lending some years ago, when most lending was done via banks. In this case the difficulty often arose because U.S. borrowers did not know whether the actual lender was the U.S. office of a U.S. or foreign bank, or a foreign branch or head office with which a U.S. banking office might be affiliated. The expansion of cross-border financing with less information appearing on banks’ records in a usable form will exacerbate this problem. Other countries with different systems for collecting data may perhaps be able to obtain better coverage, especially if the number of potential transactors is small. However, even as countries try to cope with this problem with different degrees of success, the scope for discrepancies in the reporting of capital flows will grow.

The difficulty of identifying outstanding cross-border assets and liabilities may also be increased by the growing use of interest rate and currency swaps, which have been instrumental in opening up markets to borrowers for whom direct access would have been more difficult or costly. A “currency swap” normally involves the exchange of instruments payable in two different currencies with repayments to be made over time according to a predetermined schedule encompassing both interest payments and amortization of principal. In an “interest rate swap” no actual principal is exchanged, either initially or at maturity, but interest payment streams of differing character—usually interest payments on a floating basis against fixed rate interest payments—are exchanged according to predetermined rules and based on an underlying notional principal amount.

The basic motivation behind swap arrangements is that they allow both parties in the transaction to exploit their comparative advantages in different markets and to trade the obligations to mutual benefit. As each party raises funds in the market where it can obtain the best terms, the exchange of the debt-service payments enables each party to acquire the desired kind of financing on terms and conditions which it could not have obtained directly. Apart from the cost aspect, swap arrangements also provide a useful tool to hedge interest and exchange rate exposures.

From the point of view of compiling the balance of payments accounts, swap transactions present several difficulties. Apart from the further twist in the already confused puzzle of identifying outstanding foreign assets and liabilities, there is greater ambiguity about the value of the stream of interest and principal payments between residents and nonresidents. To complicate the accounting situation further, the standard accounting rules in the United States, and probably also in other countries, become very complex when dealing with swaps, treating them essentially as hedges and therefore tending to spread any gains or losses over time and to assign them to either translation accounts or current income, depending on circumstances. It is not yet clear how these rules may be carried through into balance of payments accounting or even whether the corrections that may be necessary are significant on a global scale. At the very least, however, some elaboration of the appropriate sections of the Fund’s current Balance of Payments Manual would seem to be called for.

The balance of payments implications of the spreading use of new hedging devices, such as foreign currency and interest rate options and forward rate agreements (FRAs), are somewhat different and probably less severe for current account discrepancies than the implications of the trend toward securitized credit flows. More specifically, as is the case with interest rate swaps, hedging with options and FRAs does not involve an exchange of principal—that is, it does not give rise to changes in cross-border positions, but it may entail payments of fees, commissions, and other service income between residents and nonresidents.

Foreign currency and interest rate options are contracts conveying the right, but not the obligation, to buy or sell a specified financial instrument at a fixed price before or at a certain future date. There are two parties to an .option contract: the buyer purchases from the writer a commitment that the option writer will stand ready to sell or purchase a specified amount of the underlying instrument on demand. The option buyer’s cost for this right is paid to the option writer. The option extends, or is “alive,” until a set date. On the expiry date, the option owner can exercise his right to buy or sell the underlying asset; can let the contract expire; or, under certain conditions, can sell the option contract in the market.

Forward rate agreements are arranged between two parties wishing to protect themselves against a possible future movement in interest rates. They agree on an interest rate for a specific period of time (with a specified future settlement date) and based on an agreed principal amount.

The fundamental problem arising from these new hedging devices is that they may generate cross-border income streams which are not easy to capture and aggregate in a standardized fashion. In the case of options, for instance, records are needed to capture both the payment for an option and the transaction that occurs when the option is exercised—if it is exercised. Similarly, some special record appears to be required to account for the payments and receipts made on an FRA at the time of settlement. While it is not clear to what extent these cross-border transactions related to banks’ off-balance-sheet business are included in balance of payments records of capital flows, it should be emphasized that not all the associated service payments and receipts are necessarily incorrectly recorded.

As the BIS report Recent Innovations in International Banking emphasizes, financial innovation creates new challenges for bank supervision and for the evaluation of the macroeconomic credit situation and macro-economic policies. Some suggestions are given in that report for dealing with the statistical difficulties stemming from the securitization of bank lending and the disappearance of activity from banks’ balance sheets. However, the accounting problems that are accumulating for balance of payments compilers are not discussed in any detail in that report.


a. Conclusions

The increasing use of international financial centers by participants in international financial transactions, coupled with the introduction of new instruments and techniques for conducting international financial transactions, does not create entirely new problems for compilers of the balance of payments accounts, but it heightens the significance of longstanding sources of error or obscurity in the international accounts. In its research covering 1979–83, the Working Party found that increasing amounts of investment income attributable to residents of the offshore centers were being omitted because these centers provided little or no information on the transactions of entities operating in their territories that they did not count as residents; similarly the earnings of financial and other types of businesses established by foreign parent companies, or of personal holding companies established by nonresidents, were commonly omitted. These omissions create large discrepancies in both recorded credits and debits, though the discrepancies in the two sides tended to offset each other in 1983. Because of these omissions, there is a danger that without continuing adjustment the IFCs will add to the global discrepancy on income account in the coming years.

In addition to the problem of omitted transactions, the IFCs make it increasingly difficult, if not impossible, to trace financial and income flows between investors and final users of the funds. This lack of geographic information does not necessarily add further to the world discrepancy, but it severely limits economic analysis and the ability to unearth the sources of error in the accounts.

There seem to be two main statistical problems stemming from the acceleration of innovation in financial techniques: first, the increased speed and complexity of transactions and, second, the number of stages or transitions that may be involved in consummating a single operation. These developments put great pressure on any statistical system in which either a transactor or a financial intermediary is expected to report the amount, time, and nature of an international transaction. Furthermore, as new instruments enter the market and are traded or undergo transformations of terms or currency and of ownership, it becomes increasingly difficult to determine the residence of the creditor—an essential criterion for balance of payments accounting—at any given time.

The Working Party believes it can suggest several actions that would improve the situation (which are listed in abbreviated form below), but would caution that some difficult problems may be encountered that cannot be solved at reasonable cost.

b. Recommendations

(1) Compilers in international financial centers or other important intermediary economies should be encouraged to prepare complete balance of payments accounts, even if they are satisfied with an alternative presentation for their own use. Where such accounts are not prepared, the staff of the Fund should develop the necessary data sources and prepare proxy accounts consistent with the data being reported by partner countries.

(2) For the period immediately ahead, until such time as the preceding recommendation can be implemented, the Fund should carry forward the adjustment procedures employed by the Working Party and should improve on them as time permits.

(3) The Fund should institute a number of activities designed to bring the practices of balance of payments accounting into closer accord with the realities of contemporary financial markets. A review of the Fund’s Balance of Payments Manual, which is now being planned in the context of the System of National Accounts, should include specifically an updating of the sections relevant to the treatment of the newer financial instruments, the effects of rapid technological change in financial markets, and the preparation of balance of payments accounts by all economies that serve as financial intermediaries. Since these are highly technical subjects, the Working Party recommends that the Fund organize a consultative group of experts, both in balance of payments compilation and in actual market practice, that could serve as a source of advice in matters of accounting principles and also in terms of the practical application of such principles. Such a standing group could also be useful in keeping up with the continuous change that is characteristic of financial markets.