Journal Issue
Share
Article

Issues in the Measurement of External Debt

Author(s):
International Monetary Fund. External Relations Dept.
Published Date:
January 1990
Share
  • ShareShare
Show Summary Details

Effective policies for debt management require debt statistics that are accurate and consistent between debtors and creditors

With the emergence, in the early 1980s, of the international debt crisis, issues concerning external debt came to the forefront of economic discussion. A number of efforts have been made in recent years to try to reduce the debt burden of different groups of developing countries. Among them, the Baker and Brady Plans of successive US Treasury secretaries, a Japanese plan, and French President Mitterand’s proposal to reduce developing country debt have gained much attention. The external debt of developing countries had increased from about $64 billion in 1970 to about $380 billion in 1979, and then to some $830 billion in 1982. This produced debt-servicing problems for many of these countries in the wake of the deterioration in the world economy, the sharp rise in interest rates that followed the second large oil price rise in 1979-80, and inappropriate domestic policies. Even with a slowdown in its rate of growth, total debt of developing countries reached some $1,300 billion in 1988.

As more attention was focused on external debt, analysts noted that a number of differing figures emerged for its value because of definitional and measurement problems. The reasons for these discrepancies are: (1) each compiler has specific purposes for the data and frequently uses concepts that differ in ways that may not be immediately recognizable to general readers; (2) there is the fundamental definitional problem of determining which financial instruments should be included in the measure of external debt and debt-related capital flows; and (3) even for well-defined concepts, certain measurement problems exist.

As a consequence of the statistical problems involved in the measurement of external debt, four international organizations—the Bank for International Settlements (BIS), the International Monetary Fund, the Organization for Economic Cooperation and Development (OECD), and the World Bank—formed the International Working Group on External Debt Statistics in 1984. The first results of this collaboration were presented in External Debt: Definition, Statistical Coverage and Methodology, issued in March 1988. Their operational implications are now being worked out.

Implication for policy issues

The formulation and monitoring of desired international economic policy becomes complex in this environment where different statistical systems identify a broad range of data for, basically, the same concept. The monitoring of the Baker Plan to help certain indebted countries can be used as an example. In the table, estimates for net repayments by the 15 heavily indebted countries to international banks vary widely during 1986-88: from $21 billion, as measured by the BIS, to $2.4 billion, as measured by the World Bank. The aggregated balance of payments for the 15 countries shows net lending of $4 billion for all debt-related flows (including nonbank and banking flows).

A longer version of this article, “Methodology and Issues in the Measurement of External Debt” (1989), is available from the authors.

Net lending and estimation of new money flows to the 15 heavily indebted countries, 1985-88 1(In billions of US dollars)
1985198619871988Sum

1986-88
IBS24.0-1.62.3-15.715.0
BIS 30.91.1-7.1-14.9-20.9
World Bank41.0-1.90.6-1.1-2.4
of which: public2.5-0,42.32.14.0
Balance of payments5-6.12.31.74.0
of which:
to resident banks6-4.7-5.9-1.3-7.2
Identified adjustments7
+ debt conversions1.31.64.58.714.8
–interest arrears
capitalization-1.7-.14.3-0.93.3
+ write-downs by
banks1.01.04.05.0
Sources: See footnotes.

Net disbursements less repayments. The 15 countries are Argentina, Bolivia, Brazil, Chile, Colombia, Côte d’Ivoire, Ecuador, Mexico, Morocco, Nigeria, Peru, Philippines, Uruguay, Venezuela, and Yugoslavia.

Sources: See footnotes.

Revised from IMF Survey of May 29, 1989.

BIS, International Banking and Financial Market Developments during the First Quarter of 1989 (Basle: BIS, August 1989).

Publicly guaranteed plus nonguaranteed net lending but not including short-term. WDT (1988-89). Vol. I and II.

Total of all nonreserve capital account debt-creating flows from all nonresident institutions International Monetary Fund, Balance of Payments Yearbook (Washington: IMF, 1988).

Believed to be mostly interbank positions.

Estimates are based on incomplete information (IMF Survey of May 29, 1989).

Sources: See footnotes.

Net disbursements less repayments. The 15 countries are Argentina, Bolivia, Brazil, Chile, Colombia, Côte d’Ivoire, Ecuador, Mexico, Morocco, Nigeria, Peru, Philippines, Uruguay, Venezuela, and Yugoslavia.

Sources: See footnotes.

Revised from IMF Survey of May 29, 1989.

BIS, International Banking and Financial Market Developments during the First Quarter of 1989 (Basle: BIS, August 1989).

Publicly guaranteed plus nonguaranteed net lending but not including short-term. WDT (1988-89). Vol. I and II.

Total of all nonreserve capital account debt-creating flows from all nonresident institutions International Monetary Fund, Balance of Payments Yearbook (Washington: IMF, 1988).

Believed to be mostly interbank positions.

Estimates are based on incomplete information (IMF Survey of May 29, 1989).

With the possible exception of the World Bank statistics, the data in the table not only measure the underlying cash transactions relating to the external debt of these countries but also include various balance sheet adjustments resulting from changes in stock positions that are not due to cash transactions (e.g., write-downs). To produce a better estimate of the underlying private banking flows (i.e., net new lending) to the 15 heavily indebted countries, we have attempted to take into consideration possible balance sheet adjustments stemming from debt conversions, capitalization of interest arrears, and write-downs for adjusting the changes in net liabilities to banks, as reported by the Fund’s International Banking Statistics and by the BIS. The estimates shown in the table are incomplete and are somewhat conjectural in nature, since there is no systematic reporting of these items. However, they may indicate the order of magnitude for the adjustments: roughly $17 billion for 1986-88.

Defining and measuring debt

Several of the definitional and measurement problems are common with respect to stock and flow data. This article will therefore first discuss some of the problems related to debt stock data before returning to the more complex questions related to data on debt flows.

Stock of external debt. The International Working Group defined gross external debt as “the amount, at any given time, of disbursed and outstanding contractual liabilities of residents of a country to nonresidents to repay principal, with or without interest, or to pay interest, with or without principal.”

The two key elements within the definition are the types of financial instruments involved and residency.

The types of financial instruments that fall under the classification of debt are those contractual liabilities that involve the payment of interest and the repayment of principal. The transaction must also take place between a resident and a nonresident. The definition of residency is given in the Fund’s Balance of Payments Manual. The rule of thumb is that a person residing in a country for 12 months or longer is generally considered a resident of that economy. It is important to recognize that the external nature of a debt is not determined by the currency of denomination of the debt or by the nationalities of the transactors. Countries often incorrectly classify domestic foreign currency loans as external debts. Such loans are relevant for foreign exchange management of countries with nonconvertible currencies. However, external debt-related flows are balance of payments items that represent transactions between a country and the rest of the world—they are independent of the currency of payment. Both short- and long-term liabilities with varying interest rates and payments, in cash or in kind, are treated equally under this definition.

Definitional problems. The residency criterion as defined in the balance of payments methodology produces some contentious results in cases in which a country provides special treatment for emigrants’ deposits, which are external liabilities of the receiving bank in the home country. Where such deposits are used primarily for domestic consumption (for example, when the emigrant returns home after several years abroad to purchase a house, or when the remittances are being used by the emigrant’s family), they take on the characteristics of domestic monetary variables. In some cases where the national authorities believe that emigrants’ deposits will be used for domestic consumption, they have chosen to treat them as domestic banking liabilities. Such deposits can be quite substantial—for example, they amounted to about $4 billion in Turkey, about $8 billion in Portugal, and up to $11 billion in Yugoslavia in 1987.

Problems also arise in classifying the liabilities of offshore banking units (OBUs). These are banking entities granted special licenses by their host economies that may exempt them from some degree of domestic banking regulations, but may also constrain their contact with the host economy. The difficulty in classifying their liabilities in external debt statistics is the varying relationships of these units with the host economy and the huge size of their gross liabilities. OBUs incur sizable external obligations by foreign borrowing, but counterbalance their positions by foreign lending. For example, the gross foreign liabilities of banks in the Cayman Islands at the end of 1987 were $235 billion—about twice the size of the foreign debt of Brazil—whereas the net foreign assets were some $55 billion.

Similarly, industrial countries’ domestic banks, operating under domestic regulations, may be engaged in large extraterritorial deposit-taking and lending. Given the attempts of OBUs and international banks in industrial countries to maintain offsetting balance sheet positions, it is reasonable to argue for the inclusion only of their net foreign liability position in external debt statistics. In addition, industrial countries typically have large direct investments included in their external positions. Since the broad and highly integrated financial markets of industrial countries often allow greater substitutability between debt and equity instruments in industrial countries than in less developed countries, an analysis of the external situation of an industrial country should focus on the country’s net investment position rather than solely on its external debt.

Definitional problems also arise from the inclusion of certain types of financial instruments in the gross external debt of a country. Many of the new types of financial instruments, such as note issuance facilities (NIFs), are not recorded on balance sheets and become part of debt only when credits are actually provided. Other examples of problematic instruments would be the recording of certain leases and intracompany debt. The balance of payments methodology distinguishes between operational and financial leases, depending on whether the intent is to lease the equipment (an operational lease) or to purchase it (a financial lease). Financial leases are treated as a debt instrument. Further, the balance of payments methodology classifies intracompany debt as direct investment, although the instrument has all the characteristics of external debt. The Working Group classified such loans as external debt.

The obvious measurement problems include lack of data, most often for short-term debt, nonguaranteed private debt, and military debt. Additional measurement problems include discrepancies that can arise from using creditor-reported data and certain valuation issues.

The external debt of a country can be measured either from debtor data or from creditor data. Debtor countries usually have relatively precise knowledge about their long-term public and publicly guaranteed debt, but often are unable to monitor their private sector and short-term debt, which usually comprises bank credits and export credits. For the latter categories, creditor-based data can be particularly useful.

One of the issues that arises in reconciling creditor-based and debtor-based debt data is the valuation of debts. These issues fall into two categories: market valuation and exchange rate valuation. Market valuation refers to the market value of a debt instrument, as opposed to its face value. When determining the debt of a country, debtors will normally cite the face value of their debt, whereas creditors apply the market value of certain debt instruments, for example claims on developing countries that have been bought at a discount on the secondary market. Reconciling or combining both sources of information can therefore be problematic. The exchange rate valuation issue exists because total debt is reported in a single currency (often US dollars), but debts exist in many currencies.

Debt-related flows. The manner in which cross-border financial flows are defined and recorded complicates the interpretation of debt-related flow data. In general, as illustrated above, measurement issues arise out of the following questions: (1) Do the data represent genuine flows, or are they simply changes in stock positions? (2) Whose data are being used to calculate the flows—the debtor’s or the creditor’s? and (3) What type of accounting system is being used—accrual or cash accounting?

Debt-related flows could be defined as any transaction that gives rise to a change in (debt) liabilities to, or (debt) claims on, a nonresident. Like the measurement of debt stocks, debt-related flows can be measured either from creditor or debtor records. In economies without exchange controls, it is usually not feasible to measure actual transactions directly. Instead, flows have to be estimated from stock data by calculating the change in balance sheets after adjustments for exchange rate valuation and other effects, where possible. The Fund’s International Banking Statistics (IBS) and the BIS’s quarterly banking data estimate flow data from changes in balance sheet positions. Nevertheless, these exchange rate adjusted changes in stocks, used by many countries to calculate their balance of payments flows, are influenced by changes in the stock data that are not based on actual flows.

Some analysts are interested in the cash flow between borrowers and creditors. Exchanges of financial instruments between a creditor and debtor may cause balance sheet adjustments, which give the appearance that flows of cash have taken place when flow data are estimated from stock data. For example, in the exchange of short-term for long-term debt instruments, a repayment and new disbursement are incurred. Another case of an imputed flow arises when a debtor fails to pay interest charges and incurs interest arrears.

The use of debtor-based data versus creditor-based data for flows creates different types of problems, because all transactions between debtors and creditors may not be treated the same. For example, when a lender writes down a debt, a “repayment” takes place if the estimation is based on creditor’s stock data. However, the full face value is still considered a claim against the debtor in the borrower’s debt data. Similarly, if a bank is indemnified for the guaranteed portion of a loan, it will remove this portion from its balance sheet which means that bank lending to the country will show a decline; the transfer to the guarantor might escape recording in creditor sources.

Different accounting systems for recording flow transactions also produce confusion in measurement of debt. Cash accounting, in which entries are recorded when transactions take place, is used by the World Bank’s Debtor Reporting System and the Fund’s Government Finance Statistics. Cash accounting provides information about the actual cash flows, while payments arrears and debt restructuring are not easily identifiable. Alternatively, the balance of payments methodology and most enterprises use accrual accounting, in which entries are recorded when payments fall due, whether they have been made or not. The use of accrual accounting in balance of payments provides consistent overall balances, and will provide information about arrears, rescheduling, forgiveness, and debt instrument exchanges to the extent this information is available.

Conclusion

The lesson to be drawn is simple: the nature of data and the analytic purpose for which they are used must be kept in mind in assessing them as accurate measures of external debt. The wide range of estimated flows to the 15 heavily indebted countries illustrates the serious problems involved in measuring trends in commercial banks’ lending to developing countries. The challenge for policymakers is to formulate economic policy in this unclear situation and for balance of payments and external debt compilers to develop more reliable statistical measures.

Other Resources Citing This Publication