Chapter

17. External Debt Statistics from International Agencies

Author(s):
International Monetary Fund
Published Date:
June 2003
Share
  • ShareShare
Show Summary Details

Introduction

17.1 External debt and related statistics are disseminated by four international agencies:

  • The BIS, from its locational and consolidated International Banking Statistics (creditor reporting) and International Securities Statistics (based on market information), in the BIS Quarterly Review;

  • The IMF, according to balance of payments and IIP (BPM5) framework, in International Financial Statistics and the Balance of Payments Statistics Yearbook;

  • The OECD, through its Creditor Reporting System for the external debt of developing and transition countries, in External Debt Statistics; and

  • The World Bank, through its Debtor Reporting System for the external debt of low- and middle-income countries, in Global Development Finance.

17.2 These collections have developed for different reasons and for different purposes. This chapter outlines the BIS, IMF, OECD, and World Bank reporting systems, as at the end of 2000, and compares the data disseminated by the BIS, OECD, and World Bank with that from the IIP of the IMF. Also, this chapter provides some explanations for the differences between the OECD and World Bank data, and describes the quarterly release, Joint BIS-IMF-OECD-World Bank Statistics on External Debt.

Bank for International Settlements

17.3 The BIS produces two main sets of data: the International Banking Statistics (IBS) and the International Securities Statistics. These data are available at http://www.bis.org/statistics/index.htm, and published quarterly in the BIS publication, Quarterly Review, and in the Joint BIS-IMF-OECD-World Bank Statistics on External Debt (see below).

International Banking Statistics

17.4Table 17.1 shows the coverage of the BIS International Banking Statistics. The IBS system has two main sets of data.1 The first, which was developed in the late 1970s as a by-product of the need for monitoring overall market developments, is based on the country of location, or residence, of creditor banks (termed locational statistics). The second, which was introduced in the wake of the Latin American debt crisis in the early 1980s and was therefore explicitly designed to measure credit risk, is based on the country of origin, or nationality, of creditor banks. Its underlying principle is the worldwide consolidation of the outstanding exposures of reporting banking institutions. While the locational statistics have been available on a quarterly basis since the inception of the system, the reporting frequency of the consolidated data increased from semiannual to quarterly in 2000.

Table 17.1.Coverage of BIS International Banking Statistics
Basis for Defining CreditorBasis for Defining DebtorAvailable Breakdown
Residence/locationResidenceSector, currency, instrument
Nationality/consolidatedResidenceSector, maturity
Nationality/consolidatedNationalityNone

17.5 Although in both sets of statistics debtor counterparties are identified according to their country of residence, regardless of the location of the ultimate guarantor of the borrowed funds, only the locational banking statistics are consistent with the IIP framework. First, creditors are also identified according to their country of location and, therefore, reported by the host lending country (as opposed to the home country of the head office in the case of the nationality/consolidated statistics). This approach permits a statistical reconciliation on a country-by-country bilateral basis. Second, the breakdown by instrument—namely, between loans and debt securities holdings—comes close to the IIP distinction between portfolio and other investment positions. Third, the currency breakdown makes it possible to derive flows from stock data, which can be used as a proxy for measuring balance of payments transactions.2 Also, there is a sectoral breakdown between banks and nonbanks. Keeping in mind that domestic debt compilers face difficulties reporting comprehensively on domestic nonbank financial transactions, this breakdown is particularly useful to national debt compilers for comparative or estimation purposes.3

17.6 In contrast, the nationality/consolidated statistics are not consistent with the IIP framework. Their main objective is to measure the credit risk faced by reporting institutions, with the reporting on a worldwide-consolidated basis being the main underlying principle. Consolidation implies that the country exposure of individual reporting institutions covers that of their affiliates in all countries, including in the debtor country itself. Also as part of the process of consolidation, positions between the related offices of the same banking groups (intrabank positions) are netted out, which eliminates a number of cross-border positions. Finally, country exposure under this reporting system includes local claims denominated in foreign currencies, which clearly fall outside the scope of balance of payments statistics.

17.7 At the same time, the BIS nationality/consolidated statistics provide an insight into some important categories of countries’ external debt not available elsewhere. Prime among these is short-term debt (with a remaining maturity of up to one year), which had not been the original focus of debtor reporting systems. Another important piece of information is the sectoral breakdown (banks, the public sector, and private nonbanks). Moreover, as from end-June 1999, the reporting system includes a reallocation of claims according to the country of domicile of the guarantor, either the head office of the borrowing entity itself (for branches) or of borrowed funds with explicit (legally binding) guarantees—so-called “ultimate risk” data. Also included, in principle, under guarantees is collateral that is liquid and available in a country other than that of the borrower; that is, if the collateral provided is issued by a resident of the United States, then the ultimate risk data reallocates the claim to the United States from the country of residence of the provider of the collateral. This reclassification from immediate to ultimate counterparties will therefore exclude claims with implicit guarantees, or those perceived as such, as is the case of independent banking or corporate subsidiaries (unless explicitly covered by the head office).

17.8 As part of the BIS consolidated statistics, information is available on certain potential claims that do not appear on the balance sheet (“undisbursed credit commitments”). Such off-balance-sheet exposures include legally binding commitments to provide funds, such as the drawdown of loans according to a predefined calendar and the undrawn part of credit lines. Unfortunately, the heterogeneous nature of items covered by the definition (which may, for instance, include certain guarantees) might limit the use of this category for debt-measurement purposes.

17.9 The introduction of data on exposures to ultimate counterparties does not aim to replace those on exposures to immediate counterparties, but to provide a useful complement for the purpose of evaluating country risk. Indeed, in view of the difficulty of measuring where the final risk lies and of the significance of borderline cases, the Basel Committee on Banking Supervision has explicitly recommended that banks calculate their country exposure on both bases (dual exposure measurement).4 The ultimate risk exposure tends to provide a better measure of the ability of creditors to recoup their claims.

International Securities Statistics

17.10Table 17.2 shows the coverage of the BIS International Securities Statistics, which are derived from a database containing detailed information about all issues of international securities,5 which are obtained from various commercial market sources. Each individual issuer of securities is assigned two country fields. One is location, determined by the residence of the issuer. The second field is nationality, corresponding to the country of residence of the head office or owner of the issuing entity. Thus, debt data are available on both a residence and a nationality basis. However, since holders of debt securities are difficult to identify (not least because international bonds are generally bearer securities), there is no equivalent classification for creditors. As a result, no allowance is made for international securities purchased by residents of the debtor country. At the same time, the fact that only international securities are reported means that domestic securities purchased by nonresidents are not covered by the reporting system.

Table 17.2.Coverage of BIS International Security Statistics
Basis for Defining CreditorBasis for Defining DebtorAvailable Breakdown
Residencen.a.Maturity, currency, instrument, sector
Nationalityn.a.Maturity, currency, instrument, sector

17.11 The statistics comprise four types of basic information, pertaining to individual quarters: announcements of new issues, completions of new issues, net new issues (corresponding to the difference between completed issues and redemptions), and end-quarter stocks. The nationality and residence of issuers are readily available for these four types of basic information, as are the maturity breakdown (remaining maturity) and the sectoral breakdown. In addition, computer programs have been developed to read and aggregate individual issues to produce data such as original maturity and type of issues.

17.12 When aggregating the international banking and securities statistics for the purpose of measuring external debt, the breakdown of the locational (but not the consolidated) banking statistics into bank loans and securities holdings should in principle enable double counting in debt securities to be eliminated. However, the banking data include holdings of an unknown volume of securities issued on local markets (as opposed to international issues), which can be significant and/or volatile in some instances. As a result, the actual size of the overlap between the international banking and securities data cannot be fully ascertained.

International Monetary Fund

17.13 In the field of external debt statistics, the IMF collects and publishes annual and quarterly data on the IIP. These data are published in the monthly International Financial Statistics (IFS) publication, and in the annual Balance of Payments Statistics Yearbook (BOPSY). Data on the IIP were first published in BOPSY in 1984. The recommended concepts for the measurement of the IIP are outlined in BPM5. The concepts are consistent with the 1993 SNA, and hence with the concepts introduced in this Guide. At the time of writing, data were available for 63 countries.

17.14 The IIP is a measure of the stock of a country’s external financial assets and liabilities at one moment in time, such as year-end.6 In other words, the IIP is a statistical statement of the value and composition of the stock of an economy’s external financial assets (that is, the economy’s financial claims on the rest of the world) and the value and composition of the stock of an economy’s liabilities to the rest of the world. The financial items that comprise the position consist of claims on nonresidents, liabilities to nonresidents, monetary gold, and SDRs. In relation to the balance sheet (as delineated in the 1993 SNA) of an economy, the net IIP (the stock of external financial assets minus the stock of external liabilities) combined with an economy’s stock of nonfinancial assets comprises the net worth of that economy.

17.15 The position at the end of a specific period reflects financial transactions, valuation changes, and other adjustments that occurred during the period and affect the level of assets and/or liabilities.7 Because of the consistency of conceptual approach, the financial transactions are those recorded in the balance of payments. The valuation changes in the IIP are holding gains and losses arising from market price changes of such instruments as equities and debt securities, as well as from exchange rate changes. In nominal value terms, changes in the market price of a debt instrument do not affect the nominal amount outstanding. The other adjustments item, which is equivalent to “other changes in volume” in the 1993 SNA, are changes that are not transactions or valuation changes, but items that affect the levels of assets and liabilities, such as reclassifications.

17.16 Thus, the IIP provides a framework that allows transactions in external debt, such as disbursements and repayments of loans, the accrual of interest costs, etc., that are recorded in the balance of payments to be related to changes in outstanding positions in external debt liabilities, as recorded in the change in the IIP between reporting periods. Because stock levels are sometimes utilized in the determination of investment income receipts and payments in balance of payments accounts, consistent classification and valuation throughout the income category of the current account, the financial account, and the position components allows for meaningful analysis of yields and rates of return on external investments. In addition, the reconciliation between the IIP and the rest-of-the-world balance sheet in the national accounts provides a framework for analyzing developments in the IIP in the context of the financial behavior of all institutional sectors of the economy.8 These various reconciliations support debt analysis work.

Organisation for Economic Co-operation and Development

17.17 The OECD collects two sets of data that include information on external indebtedness:

  • Aggregate information on official and officially supported (that is, guaranteed or insured by the official sector) export credits, and individual transactions data on all other official loans from the Creditor Reporting System (CRS)—these data are published in the OECD publication, External Debt Statistics, and the Joint BIS-IMF-OECD-World Bank Statistics on External Debt (see below); and

  • Aggregate data on flows of aid loans and grants, other official flows, private market transactions, and assistance from nongovernmental organizations to each recipient country and recipient countries combined, from the Development Assistance Committee (DAC) annual questionnaire—these data are published in Geographical Distribution of Financial Flows to Developing Countries and in the Development Co-operation Report.

17.18 The main external debt publication of the OECD is the annual publication External Debt Statistics, which provides data on debt for developing and transition economies. These data are based largely on creditor sources, with the CRS data on loans (including export credits), the BIS international banking and security statistics, and the World Bank’s data on multilateral lending providing the core data series. Some additional debtor data are obtained from the World Bank for debt owed to non-OECD official creditors and from various sources for nonresident nonbank deposits in banks. The data are presented by maturity, creditor sector, and/or instrument. The classifications are not the same as in the IIP and, while in theory they should be similar, the external debt data totals in the two presentations will differ because of differences in concepts and methodology used, and in completeness of reporting.

OECD Reporting Systems

17.19 The CRS was established in 1967 with the aim of supplying “participants with a regular flow of data on indebtedness and capital flows.” Consequently, over the years it has become a major source of information not only on official lending but also on the terms and conditions of external lending, as well as the sectoral and geographic distribution of flows to developing economies.

17.20 The CRS comprises separate report forms for commitments and loans. Three report forms cover commitments: grants (Form 1A); aid and other official loans excluding export credits (Form 1B); and guaranteed and direct export credits extended for a period of five years or more (Form 1C). Four forms cover loans: the status of individual aid and other official loans, excluding export credits (Form 2); the status of aggregated medium- and long-term guaranteed export credits (Form 3); the status of aggregated medium- and long-term direct export credits (Form 3A); and the outstanding amounts of short-term export credits on an original maturity basis (Form 3B). Form 2 provides individual transaction data, and Forms 3, 3A, and 3B provide aggregate data on outstanding amounts at the end of the period and transactions during the period. Forms 3, 3A, and 3B also provide expected payments.

17.21 Reporting frequency for the CRS differs among forms. Whereas respondents report official loan commitments continuously, and export credit debt semiannually, data on the status of individual aid and other official loans are reported annually. Because these loans are not closely affected by financial market developments, this frequency is considered adequate.

17.22 The annual DAC questionnaires provide aggregated flow data that are based largely on balance of payments principles, with the exceptions noted below.9 Thus, there is a broad correspondence between balance of payments and DAC flows data. Where CRS reporting is incomplete, debt flow data may be obtained from the DAC reporting system, and debt stock data may be estimated on the basis of previous stocks and DAC flows.

Comparison of OECD Data with Balance of Payments/IIP Data

Presentation of data

17.23 Unlike the presentation in the IIP, OECD categories show different types of debt, based partly on the creditor and partly on the instrument. They include official bilateral lending (excluding export credits), official development assistance (ODA)/official aid, officially supported export credits, official multilateral lending, bank lending, debt securities, other claims, and short-term debt.

17.24 Historically, the collection of data on ODA and other official loans has reflected analytical interest in recording development finances, especially aid. ODA is defined as those flows to countries on Part I of the DAC List of Aid Recipients that are (1) provided by official agencies, including state and local governments, or by their executive agencies; and (2) each transaction is administered with the promotion of the economic development and welfare of developing countries as its main objective, and is concessional in character and conveys a grant element of at least 25 percent (calculated at a rate of discount of 10 percent). Flows to countries on Part II of the DAC List (transition countries) that meet the above criteria are classified as official aid.10 Although it is rare, such loans may also be made to the private sector in the borrowing country.

17.25 The collection of export credit data arose from the needs of the OECD Trade Committee to follow the activities of export credit agencies. Also of interest to creditors and debtors is the scale of multilateral lending from the World Bank and related organizations, as are loans made by other non-OECD creditors, although these data are compiled from the Debtor Reporting System (DRS).

17.26 The presentation of the data allows creditors to consider country risk. Debtors and creditors can identify amounts that may be renegotiated in such for a as the Paris Club, the London Club, or may be the object of bilateral debt relief, and examine such questions as burden sharing by creditors, or the relative importance of different categories of creditors in a debtor country’s borrowing.

Concepts

17.27 In both the OECD’s reporting systems, the balance of payments criteria of residence is generally required. Creditors identify their debtor counterparts according to their country of residence, although in cases such as offshore centers, flag of convenience countries, or aircraft leases, the ultimate borrower may be in a third country. In the OECD reporting systems, all debt stocks and flows are valued at face value, unlike market value in the balance of payments and IIP. Although this may seem like a major divergence, there is in practice little difference because nontraded instruments are invariably valued at nominal value in the IIP.

17.28 A significant difference between the OECD and IIP data is that, unlike the IIP, OECD data are not reported on a full accrual basis. Both the OECD and IIP data record disbursements at the time they occur, whereas repayments are reported in OECD data when they take place, not when they are due (as in the IIP). OECD debt stock is calculated as the amount of disbursed principal outstanding plus interest in arrears, whereas the IIP debt stocks are the amounts outstanding including all interest costs that have accrued and have not yet been paid.

17.29 The IIP and OECD data define long-term and short-term debt identically. Thus in the OECD data, short-term debt includes all debt contracted for a period of one year or less plus, wherever possible, arrears of both principal and interest on all debt. In the OECD data, the maturity breakdown is available for only two categories: banks and export credits. For other categories, all debt is classified as long-term. Using data in External Debt Statistics, debt with a remaining maturity of one year or less can be estimated by combining short-term debt with the amount of principal payments due in the next year on long-term debt.

17.30 The nonresident creditor sector is published in the OECD data, whereas the IIP publishes the resident debtor sector. Also, the sector classification in the OECD data does not correspond with the IIP, or the 1993 SNA. In the OECD data, there is the official sector and the private sector, of which banks are separately identified. Although not published, the OECD does have some data on the resident debtor sector. The classification of borrowers is not reported for official lending other than export credits, but it can be assumed that the vast majority of borrowers of these funds belong to the general government sector. In the case of export credits, reporters distinguish between public and private borrowers, although a further distinction between bank and other private sector is not available. Rescheduled export credits are assumed to be claims on public borrowers; the rescheduled debt notified to the OECD is generally debt rescheduled by the public sector of the debtor and the official sector of the creditor.

Specific items

Trade credits

17.31 The concept of “trade credits” is wider in the OECD data than in the IIP, which “only” includes claims and liabilities arising from the direct extension of credit by suppliers and buyers for transactions in goods and services, and for work in progress (or to be undertaken). The OECD data cover three types of export credits—officially supported suppliers credits, officially supported bank credits, and official direct credits. They do not cover private sector credits that do not have official support in the form of insurance or guarantee.

Arrears

17.32 The recording treatment of arrears of principal and interest is the same in the IIP and OECD debt data; arrears arise when payments are past due, and are classified as short-term debt. However, in the OECD data, late interest (interest on arrears) is reported and included in the debt stock only when the interest is capitalized under a rescheduling, whereas in the IIP interest costs accrue on arrears (although BPM5 is not very clear on this issue).

Write-offs

17.33 A write-off is a unilateral creditor action that is an accounting procedure that removes a debt from a creditor’s books. As such, it should be reflected in the notification of creditors’ debt stocks to the OECD, thus affecting the level of claims. The IIP is silent on the treatment by the debtor, so a discrepancy could arise between the debtor and creditor data. While write-offs are rare for official and officially supported debt, they are a more common procedure for banks.

Debt forgiveness

17.34 In the DAC statistics debt forgiveness is a similar but different concept from that used in BPM5. Unlike BPM5, only relief implemented for the purpose of promoting the development or welfare of the recipient qualifies as debt forgiveness in the DAC data. However, if this condition is fulfilled, like BPM5, a voluntary cancellation of debt within the framework of a bilateral agreement is classified in the DAC statistics as debt forgiveness, and is reported as an ODA grant (capital transfer in BPM5). Unlike BPM5, the DAC system’s concept of debt forgiveness also includes a reduction in the present value of debt achieved by concessional rescheduling or refinancing, and the discount in a debt conversion occurring within the framework of a bilateral agreement between governments (although in certain circumstances BPM5 also records such discounts as debt forgiveness; see Chapter 8 of the Guide, paragraph 8.33).

17.35 Most OECD reporters follow balance of payments principles in reporting forgiveness when debt is canceled—the amount forgiven is valued as the amount of debt stock canceled and is reported in a lump sum at the time the creditor enters the forgiveness on its books. However, a few reporters only report the forgiveness of the debt annually when debt-service payments would have fallen due. This approach results in differences in timing (forgiveness spread out over many years) and amounts (interest not yet due at the time of forgiveness included in addition to principal and interest arrears) between the DAC forgiveness grants and the balance of payments capital transfers. Because some already forgiven amounts may remain included in the outstanding debt stock until the period in which their payments would have fallen due, this approach may also have the consequence of overstating the OECD measured debt stock in the intervening period after the forgiveness agreement.

Debt rescheduling

17.36 Debt rescheduling is reflected in both the debt stock and flow data collected by the OECD. The rescheduling flows are recorded at the time of actual implementation of the rescheduling, which should correspond to the time they are entered into the books (of both the creditor and the debtor), the same approach as the IIP. The rescheduling of any future maturities is therefore recorded at the time of the actual implementation of their rescheduling, rather than when the rescheduling as a whole is agreed. When short-term debt, including arrears, is rescheduled into long-term maturities, this is reflected in the OECD data, as in the IIP. Also as in the IIP, if the rescheduling involves a shift in creditor or debtor sectors—for example, a Paris Club rescheduling of debt lent by the private sector (under a creditor government guarantee) to the private sector (under a borrower government guarantee) could become government to government—the OECD data record the legal change of ownership.11 However, when the rescheduling is within the official sector, only the capitalization of interest is reported as a flow (in order to avoid two offsetting entries for the principal rescheduled). While rescheduled export credit debt owed to public creditors can be identified as such in the OECD database, in External Debt Statistics it is classified under nonbank export credits.

17.37 Although there is much similarity in the principles, in practice the complexity of restructuring makes full and correct reporting difficult to implement for both creditors and debtors and can give rise to discrepancies between OECD data and the IIP. Genuine differences in the timing of book entries between creditors and debtors, and practical difficulties in tracing restructuring that can lead to problems such as misclassification of arrears and rescheduled debts, and omission of capitalized interest, sometimes produce different figures in creditors’ reports and the debtor’s IIP.

Debt conversions

17.38 In OECD data, when official debt is exchanged for equity or counterpart funds to be used for development purposes, this should be reported as an ODA grant for debt conversion, with debt forgiveness recorded only if there is a discount on the exchange. Also when, in the framework of a bilateral agreement for development purposes, the official sector sells debt at a discount to a private sector entity that is then exchanged for equity or counterpart funds to be used for the benefit of the private sector entity for development purposes, the official sector’s loss should be reported as debt forgiveness. In both cases, and as in the IIP, the debt stock is reduced by the value of the debt converted.

World Bank

17.39 The World Bank collects data on external indebtedness from debtor countries through the Debtor Reporting System (DRS). These reported data form the core of the detailed country-level debt stock and flow data that are published annually in the Global Development Finance (GDF) publication (formerly World Debt Tables). Selected debt data are also available in the annual World Development Indicators publication, and in the Joint BIS-IMF-OECD-World Bank Statistics on External Debt (see further, below).

17.40 The World Bank’s interest in debt statistics is both analytical and operational. At the analytical level, the Bank is a leading international source of information and analysis on the economic situation of developing countries. Bank staff make extensive use of debt statistics in analyzing the economic prospects, financing needs, creditworthiness, and debt sustainability of developing economies. At the operational level, the lending and borrowing activities of the Bank demand a close monitoring of the overall financial situation of each borrower, such as debt-servicing capacity. To this end, the Bank’s General Conditions (of borrowing) require a borrowing or guaranteeing member country to report external debt information to the Bank. As a condition of presentation of loans and credits to the World Bank’s Executive Board, each borrowing or guaranteeing country must submit a complete report (or an acceptable plan of action for such reporting) on its external debt.

Debtor Reporting System

17.41 The DRS was established in 1951 and is the World Bank’s principal means of monitoring external debt. Through the DRS, countries—typically low and middle-income—that borrow from the Bank report data on long-term external indebtedness.

17.42 The number of countries covered and the data to be reported have expanded over time. At the time of writing, 136 countries submitted two kinds of reports: loan-by-loan data on long-term debt of the public sector and debts guaranteed by the public sector; and summary reports on long-term debt of the private sector that is not publicly guaranteed. The data are supplied on special reporting forms. For public and publicly guaranteed debt, individual new loan commitments are reported (quarterly) on Forms 1 and 1A, and the status of each loan at the end of the recording period and the transactions recorded during the recording period are reported on Form 2. For private nonguaranteed debt, aggregate figures on the stock of debt, transactions during the recording period, and future debt services are reported on Form 4. Short-term debt data are either obtained from the country or estimated separately using creditor and other sources, the most important source being remaining-maturity data from the BIS consolidated International Banking Statistics, which are adjusted to come into line with the original-maturity concept.12

17.43Form 1 is used for reporting the terms and conditions of each external public and publicly guaranteed debt obligation incurred during a calendar quarter with an original maturity of more than one year. This report allows a wide range of information to be captured and disseminated for statistical and analytical purposes within and outside the World Bank.

17.44 Information is collected on creditor name, type, and residence and is used in classifying external debt owed to official and private creditors, assessing creditor exposure, analyzing net resource flows from official and private creditor sources, and identifying Paris Club debt eligibility.

17.45 On the debtor side, borrower name and type, guarantor name, economic sector of borrower, and whether funds for debt servicing are to come from the budget of the central government are also reported on Form 1. This information is used in several ways, including measuring public and private sector borrowings, identifying uses of funds, and assessing the central government’s debt burden.

17.46 Form 1 allows compilation of detailed information on loan terms, including interest rates and spreads, grace period, maturity, debt-service pattern, and currencies in which the loan amounts are denominated and repaid. This information is used in calculating the grant element component, projected debt service, present value of debt, and other debt and economic indicators.

17.47Form 1A captures future payments due when the terms of repayments cannot be adequately described in Form 1, and amounts rescheduled in multiyear rescheduling agreements that will become effective on future specified dates.

17.48Form 2 is used for reporting the annual status of each external debt liability with an original maturity of more than one year. This annual summary report presents stock and flow information for each public or publicly guaranteed debt extant at the end of the reporting period or repaid or canceled during the period. For each debt, the amount of debt committed, undisbursed, and outstanding and disbursed is presented along with the transactions that have taken place during the year. Also presented is information on any accumulation of arrears and debt reschedulings. All amounts are reported in the currency in which the debt is payable. Based on the Form 2 report, a wide range of stock and flow accounts as well as economic indicators are derived and disseminated in the GDF. The report is due within three months after the end of the reporting period.

17.49Form 4 is used for submitting annual information on the status of private sector external debt that has an original maturity of more than one year and that does not have a public sector guarantee. The information is aggregated by type of debtor institution—commercial banks, direct investment enterprises, and other—and a separate form is submitted for each type of institution. The creditor information for each type of debtor institution is provided for the following types of creditors: private banks and other financial institutions, foreign parents and affiliates, exporters and other private, and official (governments and international organizations).

17.50 Form 4 contains both stock and flow accounts and, for each type of debtor institution, estimated future payments of principal and interest for the first ten years following the end of the reporting period.

Comparison of World Bank and Balance of Payments/IIP data

Presentation of data

17.51 The presentation of debtor data, as shown in the GDF, responds to a different set of analytical requirements from that of the IIP. The aim is to provide a detailed view of a country’s borrowing activities, accessibility to external funds, and borrowing costs, as well as to facilitate a comprehensive analysis of the debt burden, debt-servicing capacity, financing needs, and creditworthiness of the country. To this purpose, both stock and flow data are provided at different levels of breakdown. The first breakdown is that between long- and short-term debt, and the second between public (and publicly guaranteed) and private borrowing. Special attention is paid to identifying private borrowing with direct government guarantees. Also, projected repayment profiles are viewed as critical to analysis and management of obligations, and these are included in the presentation of data.

17.52 The creditor breakdown goes beyond a breakdown by instrument. For instance, for official creditors, multilateral and bilateral, the more detailed breakdown identifies concessional lending by this sector. These data are particularly useful in debt work. Official credits with an original grant element of 25 percent or more using a 10 percent rate of discount are characterized as concessional (as defined by the DAC). The exception are credits from major regional development banks—African Development Bank, Asian Development Bank, European Bank for Reconstruction and Development, and the Inter-American Development Bank—and from the IMF and World Bank, where concessionality is determined on the basis of each institution’s own classification of concesssional lending.

17.53 The disaggregation of private lending is a mixture—by institution, such as banks, and by instrument, such as bonds. Trade-related borrowing, such as export credits and supplier credits, is included within “other private,” and so not separately identified. The presentation of the data distinguishes between private-sourced debt that is owed by public entities or owed by private entities but with explicit government guarantees, and that which is owed by the private sector.

17.54 Projected debt-service payments and debt-disbursement profile are based on current debt transactions and loan terms. Projected debt-service payments are projections of payments due on existing debt outstanding, including undisbursed amounts of existing external debt, taking account of implemented multiple-year restructuring agreements. Future disbursements and debt-service payments refer only to existing debt and do not reflect any assumptions about future borrowing.

Concepts

17.55 The principal concepts used by the DRS in compiling debt stocks are consistent with the conceptual framework of the Grey Book (BIS and others, 1988) and there is consistency with several elements of the IIP as well. The level of detail of the information required from reporting countries and the presentation of debt data are influenced by the analytical and operational application of the data (see paragraph 17.40). The DRS includes all debt with an original maturity of more than one year owed to nonresidents and short-term debt. Total external debt is derived as the aggregate of long-term and short-term debt (and use of IMF credit).

17.56 Like the IIP, external debt statistics in the DRS are compiled on a residence basis (as opposed to a nationality basis)—external debt is that owed by entities physically located in the reporting country to entities located outside the reporting area, irrespective of nationality. So, branches of foreign banks are resident to the reporting country, whereas foreign offices of domestic banks are not. Also, bank deposits held in domestic banks by nationals living abroad are included in external debt data.

17.57 In a few cases the DRS deviates from the residence criteria, and hence the IIP framework, for analytical and operational reasons. For instance, the DRS excludes from a country’s external debt the indebtedness of banks located in a resident offshore banking center; this indebtedness can often be very large in relation to the host economy.13

17.58 Debts payable in both foreign and domestic currency to nonresidents are required information to be reported under the DRS. In practice, the DRS’s focus is on foreign currency debt; domestic currency debt owed to nonresidents has not been included. This is a departure from the IIP framework. Also, currency—notes and coins—held by nonresidents are not captured by the DRS.

17.59 A point of departure from the IIP is the valuation of stocks. The DRS measures all stocks at nominal value rather than at traded or current market value. For nontraded or nontransferable debt instruments such as loans and deposits, there is in practice little difference because nontraded instruments are invariably valued at nominal value in the IIP. However, this is not true for traded debt instruments.

17.60 Short-term and long-term debt are similarly defined in the DRS and the IIP: short-term debt includes all debt with an original maturity of one year or less, and long-term debt includes all debt with an original maturity greater than one year; interest arrears are included under short-term debt. There is a difference in the treatment of principal arrears; the DRS classifies these arrears by the original type of the debt, whereas the IIP classifies them as short-term debt.

17.61 The DRS sectoral classification of external long-term debt has two categories: debt of the public sector and private debt with a public sector guarantee; and all other private, nonguaranteed debt. This classification is not equivalent to the IIP breakdown, although with the available information it is possible to relate the DRS’s debtor classifications—the nine types are central government, local government, central bank, private bank, direct investment, public corporation, mixed enterprise, official development bank, and private—to those of the IIP. Within the debtor category the DRS provides a further breakdown by creditor sector. The IIP does not provide a creditor sector breakdown.

17.62 The DRS measures debt stocks and flows on a cash transaction basis as opposed to the accrual method recommended in the IIP. Thus, reported flows are the result of a cash (or in-kind) transaction, such as an actual loan disbursement or repayment, and debt outstanding is the amount disbursed less amount repaid (and any interest arrears). Projections are on a debt-due basis. In the IIP framework, disbursements are recorded when they occur, but repayments are reported when due. Debt stocks in the IIP include interest costs that have accrued and have not yet been paid.

Comparison of World Bank Data with OECD Data

17.63 There are notable differences for data users, both in presentation and in the recording of categories of debt between the debt statistics of the debtor and creditor data. This is because the breakdowns chosen by the different reporting systems reflect the analytical requirements of users. This section discusses and explains the reasons for some of the differences.

17.64 One classification that is consistent in both systems is the concept of short- and long-term based on original maturity. Both systems also provide data on long-term debt due within the year, as well as short-term debt data on an original-maturity basis, thus allowing for a measure of remaining maturity to be estimated.

17.65 The compilers of the DRS and CRS do compare the two sets of data series to see why there are differences in reported debt figures. From their work, certain reporting differences have emerged.

17.66 First, certain borrowing countries apply the definition of short- and long-term debt differently from creditors. For instance, certain creditors may classify short-term loans that are rolled over as long-term loans. Second, the DRS does not classify arrears of principal as short-term debt but, as noted above, by the original type of debt. However, this does not pose a problem for reconciling debtor and creditor data because principal arrears are separately identified in the debtor data so to allow comparability with creditor information.

17.67 Second, bilateral ODA in the CRS and bilateral concessional debt in the DRS are not entirely comparable. The difference emerges because of the coverage of loans. In the debtor data, bilateral direct export credits may be included under bilateral concessional loans if the grant element on the loans is 25 percent or more, whereas they are classified as export credits and not ODA in the creditor data. When export credits are subsidized by ODA loans—mixed credits—the subsidies for such credits would appear as ODA loans in the creditor data.

17.68 Third, loan-by-loan comparisons between the DRS and CRS have sometimes shown a different perception of the timing of disbursements and repayments between debtor and creditor, resulting in a difference in the reported outstanding debt at any given time.

17.69 Fourth, differences arise due to restructuring. In the case of forgiveness, the DRS may, for analytical purposes, anticipate the timing of cancellation, whereas the creditor usually waits for the signing of the bilateral agreement, which may entail delayed parliamentary approval. In the case of rescheduling of guaranteed export credits, the rescheduled loan may remain classified as an export credit rather than a new official loan on the creditor side, whereas the debtor records it as a bilateral official loan.

Joint BIS-IMF-OECD-World Bank Statistics on External Debt

17.70 The Joint BIS-IMF-OECD-World Bank Statistics on External Debt were first released on March 15, 1999 on the website of the OECD14 with hyperlinks available from the websites of the BIS, the IMF, and the World Bank.15 This release is an initiative of the Inter-Agency Task Force on Finance Statistics (TFFS); it is updated quarterly. The purpose of the site is to facilitate timely and frequent access by a broad range of users to one dataset that brings together external debt data that are currently compiled and published by the contributing international agencies (BIS, IMF, OECD, and World Bank).

17.71 The types of debt primarily covered in the Joint Statistics comprise loans from banks, debt securities issued abroad, Brady bonds, officially supported nonbank trade credits (that is, export credits extended by nonbank institutions of the exporting country), multilateral claims,16 and official bilateral loans (loans provided mainly for development purposes excluding export credits). Data on total liabilities to banks and on officially supported bank and nonbank trade credits are available as the memorandum items. The BIS, IMF, OECD, and World Bank provide the data. The statistics are mostly from creditor and market sources but also include data provided by debtor countries. At the time of writing, data were available for more than 175 countries. Data are also shown on external financial assets in the form of claims on banks and holdings of international reserve assets, which are prepared by the BIS and the IMF, respectively. A detailed description of the data in the Joint Statistics as at mid-2001 is provided in Box 17.1.

Box 17.1.Joint BIS-IMF-OECD-World Bank Statistics on External Debt

The sources, definitions and coverage of individual series are explained in detail in this box. See also Table 17.3 and http://www1.oecd.org/dac/Debt/index.htm.

The columns of the table cover stocks—the amounts outstanding at the end of each period—and flows—disbursements net of repayments during the period. Flows are available for debt securities, Brady bonds, multilateral claims, and bilateral loans (lines B, C, E, and F of the table). For the banking and trade credit figures (lines A, D, J, L, and M of the table), the change in stocks, adjusted for changes in exchange rates to the U.S. dollar during the period, is given. For other series, flow data are not available.

Line A: Bank loans

Line J: Total liabilities to banks (locational)

Line M: Total claims on banks (locational)

Data above are derived from the BIS locational banking statistics.

Line B: Debt securities issued abroad

Line H: Debt securities issued abroad (due within a year)

Data are derived from quarterly BIS statistics on international securities.

Line C: Brady bonds

Brady bonds comprise commercial bank debt restructured under the Brady Plan, introduced in early 1989. Data on Brady bonds are provided from the World Bank’s Debtor Reporting System (DRS). Annual data on stocks and flows (issues less repayments) are as reported by the debtor country and include buybacks. Quarterly data on stocks and flows are estimates based on repayment terms of the bonds and reflect adjustments for buybacks during the quarter. In the World Bank’s Global Development Finance (GDF), data are included (but not shown separately) under public and publicly guaranteed debt.

Line D: Nonbank trade credits

Data are derived from the semiannual reports to the OECD made by OECD member countries’ export credit guarantee agencies. Nonbank trade-related credits comprise official export credits, which are long-term, and officially guaranteed or insured suppliers’ credits, which are credits extended by exporters to importers abroad. They also include arrears and officially rescheduled amounts on officially guaranteed or insured financial credits, since these are taken over by export credit agencies from the original bank creditors. Guaranteed financial credits made by banking institutions that do not report to the BIS are also included here. These data only cover trade credits that have been guaranteed or insured by the official sector in the creditor country. They include credits extended to both the public and private sector in the borrowing country.

Line E: Multilateral claims

Multilateral claims cover data for African Development Bank (AfDB), Asian Development Bank (ADB), Inter-American Development Bank (IADB), IMF, and World Bank claims. Stocks are the total of loans from AfDB, ADB, and IADB, use of IMF credit, and IBRD loans and IDA credits from the World Bank. Flows are the sum of disbursements less principal repayments on loans and IDA credits, and IMF purchases less IMF repurchases.

Line F: Official bilateral loans (DAC creditors)

This line shows the outstanding debt from the OECD’s Credit Reporting System (CRS) on loans, other than direct export credits, extended by governments that are members of the OECD’s Development Assistance Committee (DAC). Direct export credits extended by the official sector are included in nonbank trade credits (lines D and I). In addition to straightforward loans, official bilateral loans include loans payable in kind, and eligible loans in Associated Financing packages.

Line G: Liabilities to banks (due within a year)

Line K: Total liabilities to banks (consolidated)

Data are derived from the BIS consolidated banking statistics.

Line I: Nonbank trade credits (due within a year)

These data are derived from the OECD’s CRS. They comprise official and officially guaranteed or insured suppliers’ credits extended by exporters to importers abroad that have a remaining maturity of one year or less. They include (1) export credits with an original maturity of one year or less and (2) the amounts of principal due in the next year on credits with an original maturity of over one year. These data only cover trade credits that have been guaranteed or insured by the official sector in the creditor country. They include credits extended to both the public and private sector in the borrowing country.

Line L: Total trade credits

These data are derived from the OECD’s CRS. This line covers all official and officially supported trade credits; that is, trade credits that have been guaranteed or insured by the official sector in an OECD reporting country. The credits include those extended to both the public and private sector in the borrowing country. In addition to the nonbank trade credits shown in line D, this line includes financial or buyer credits extended by banks that are guaranteed or insured by an official export credit guarantee agency. These guaranteed bank credits are also included in the amounts shown in line A (Bank loans), line G (Liabilities to banks), line J (Total liabilities to banks—locational), and line K (Total liabilities to banks—consolidated).

Line N: International reserve assets (excluding gold)

Data are those published in the IMF’s International Financial Statistics (IFS).

17.72Table 17.3 (on preceding page) shows a sample table from the Joint Statistics: the stock of debt, with a minimum two-month lag, for the past five quarters and the previous December; and flow figures for the latest complete two years and two recent quarters. Whenever available, data on short-term debt, based on the remaining-maturity concept, are also provided. Free access to an on-line database, which provides longer time series and permits manipulation of the figures, is also available. Some of the data are only available semiannually, and no attempt is made to provide quarterly inter- or extrapolations of these data. The data are published 22 weeks after the end of the quarter.17

Table 17.3.Example of Joint BIS-IMF-OECD-World Bank Statistics on External Debt (1)
Stocks (end of period)Flows (2)
2000 December20012002 March2000 Year2001 Year2001
(In millions of U.S. dollars)MarchJuneSeptemberDecemberThird QuarterFourth Quarter
COUNTRY A
External debt—all maturities
A Bank loans (3)
B Debt securities issued abroad
C Brady bonds
D Nonbank trade credits (4)
E Multilateral claims
F Official bilateral loans (DAC creditors)
Debt due within a year
G Liabilities to banks (5)
H Debt securities issued abroad (6)
I Nonbank trade credits (4)
Memorandum items
J Total liabilities to banks (7) (locational)
K Total liabilities to banks (6) (consolidated)
L Total trade credits
M Total claims on banks (8)
N International reserve assets (excluding gold)
COUNTRY B
External debt—all maturities
A Bank loans (3)
B Debt securities issued abroad
C Brady bonds
D Nonbank trade credits (4)
E Multilateral claims
F Official bilateral loans (DAC creditors)
Debt due within a year
G Liabilities to banks (5)
H Debt securities issued abroad (6)
I Nonbank trade credits (4)
Memorandum items
J Total liabilities to banks (7) (locational)
K Total liabilities to banks (6) (consolidated)
L Total trade credits
M Total claims on banks (8)
N International reserve assets (excluding gold)
COUNTRY C
External debt—all maturities
A Bank loans (3)
B Debt securities issued abroad
C Brady bonds
D Nonbank trade credits (4)
E Multilateral claims
F Official bilateral loans (DAC creditors)
Debt due within a year
G Liabilities to banks (5)
H Debt securities issued abroad (6)
I Nonbank trade credits (4)
Memorandum items
J Total liabilities to banks (7) (locational)
K Total liabilities to banks (6) (consolidated)
L Total trade credits
M Total claims on banks (8)
N International reserve assets (excluding gold)
Source: OECD, on the Internet at http://www1.oecd.org/dac/debt.

From creditor and market sources, except for data on Brady bonds which are from debtor sources, all currencies included.

Flow data for items B, C, E, F and L; exchange rate adjusted changes for items A, J, and M; no data available for items D, G, H, I, K and N.

From BIS locational banking statistics, which are based on the country of residence of reporting banks.

Official and officially guaranteed. Break in series at end-1998 due to reallocation of rescheduled export credits from line F to line D.

From BIS consolidated banking statistics, which are based on the country of head office of reporting banks and which include banks’ holdings of securities.

Including debt securities held by foreign banks, which are also included in line G.

From BIS locational banking statistics, which are based on the country of residence of reporting banks and which include banks’ holdings of securities.

Source: OECD, on the Internet at http://www1.oecd.org/dac/debt.

From creditor and market sources, except for data on Brady bonds which are from debtor sources, all currencies included.

Flow data for items B, C, E, F and L; exchange rate adjusted changes for items A, J, and M; no data available for items D, G, H, I, K and N.

From BIS locational banking statistics, which are based on the country of residence of reporting banks.

Official and officially guaranteed. Break in series at end-1998 due to reallocation of rescheduled export credits from line F to line D.

From BIS consolidated banking statistics, which are based on the country of head office of reporting banks and which include banks’ holdings of securities.

Including debt securities held by foreign banks, which are also included in line G.

From BIS locational banking statistics, which are based on the country of residence of reporting banks and which include banks’ holdings of securities.

17.73 With a view to making users aware of the data limitations and promoting best practice in using the data, a set of metadata has been prepared, along with the data, indicating how the data relate to internationally agreed concepts. These data are mostly from creditor and market sources but also include information provided by the debtor countries themselves. These data do not provide a completely comprehensive and consistent measure of total external debt. For example, these data do not cover (1) nonofficially guaranteed suppliers credit not channeled through banks; (2) direct investment: intercompany lending; (3) private placements of debt securities; (4) domestically issued debt securities held by nonresidents; (5) deposits of nonresidents in domestic institutions; and (6) amounts owed to non-DAC governments. Nevertheless, the Joint Statistics do bring together the best international comparative data currently available on external debt that are compiled and published separately by the contributing institutions.

17.74 The user needs to be careful in comparing data series. For instance, there are overlaps between data sources such as the international securities data and the nationality/consolidated banking statistics, which indistinguishably include securities. Thus, for debt due within a year, the data relating to debt securities issued abroad include securities held by foreign banks that are also included under the data relating to liabilities to banks. Also, there can be inconsistencies. For example, the data on loans from banks and on total liabilities to banks due within a year are drawn from different data sources—the BIS locational and consolidated international banking statistics, respectively. Thus, creditor and market-based statistics are not a substitute for setting up appropriate reporting systems by the debtor countries themselves.

See BIS (2000a). Although the BIS also collects data on syndicated loan facilities, this information cannot be used for measuring external debt. First, facilities may be used as a backup for other types of fund-raising and may therefore remain undrawn or only partially used. Second, in some instances the funds are used to replace past banking debt, without therefore entailing any increase in borrowers’ debt. Third, syndicated loans are but one of the various forms of international bank lending. Thus, whereas syndicated loan data may help to assess current market conditions, they cannot be used to measure external debt.

Changes adjusted to exclude the impact of currency movements on stock data using average exchange rates for the period under consideration can only serve to approximate actual transactions.

See also IMF (1992), pp. 54–62.

See Basel Committee on Banking Supervision (1982). Also, in this context and as mentioned in Chapter 12, the collection of semiannual statistics on open positions in the global over-the-counter (OTC) derivatives market was introduced by the BIS in June 1998. However, these data are not available with a country-by-country breakdown of counterparties.

International securities issues are defined as those raised outside the debtor country itself, whether in the international bond (formerly Eurobond) market or in foreign markets, such as the Yankee bond market.

For a full description of the IIP, see Chapter XXIII of BPM5.

While the financial transactions are shown in IFS and BOPSY as part of the balance of payments statement, the valuation changes and other adjustments are not collected or published by the IMF.

There are differences in classification between the rest-of-the-world account and the IIP that reflect, inter alia, differences in analytical requirements. For instance, the national accounts focus on instruments, while the IIP focuses on functional categories. The detailed reconciliation is provided in Appendix IV.

Further information on the DAC reporting system is available on the Internet at the following website:http://www1.oecd.org/dac/htm/crs.htm.

The DAC list essentially includes all non-OECD members and some OECD members.

The counterentry to the governments’ assumption of external debt might well be a claim on their private sector or capital transfer. Because of guarantees or insurance provided by the government’s export credit agency, the creditor government may acquire the claim from their private sector.

Described in detail in the notes and definitions to the Global Development Finance report; on the Internet, see http://www.worldbank.org/prospects/gdf2002/vol1.htm. The GDF database is also available on-line, by subscription, at http://publications.worldbank.org/ecommerce/catalog/product?item_id=1023868.

The same can be true for countries that sponsor “flag of convenience” companies.

At the time of writing, the multilateral claims covered by the data in the Joint Table were loans from the African Development Bank, Asian Development Bank, and Inter-American Development Bank, use of IMF credit, and IBRD loans and IDA credits from the World Bank.

The lag refers to the BIS International Banking Statistics, which are the core series in the Joint Statistics.

    Other Resources Citing This Publication