1. Coordination among official agencies

6.5 If the responsibility for debt compilation is shared between several agencies, it should be clearly established which agency has the primary responsibility for compiling and disseminating the complete set of public sector debt statistics—hereafter referred to as the central compiling agency. Responsibility may be assigned through a statistical law or other statutory provision, interagency protocols, executive decrees, etc.

6.6 This Guide does not recommend which institution should be responsible for compiling and disseminating public sector debt statistics; this depends on the institutional arrangements in each country. Nonetheless, it is likely that the central compiling agency is the central bank, the ministry of finance, an independent debt management office, or a national statistical agency.¹

6.7 In whatever way the statistics are to be collected and compiled—and invariably a range of methods and approaches will be adopted—the process will be resource intensive. Thus, where there is more than one agency involved in the compilation of public sector debt statistics, there should be a coordinated and cooperative effort, avoiding duplication of effort and ensuring as far as possible consistency of approach across related data series. With computerized techniques, different units can be connected through computer networks facilitating the specialization of the different institutions concerned, without hindering data reporting and compilation. In this regard, procedures to ensure, as far as possible, smooth and timely flows of data between data compiling agencies are essential.

6.8 It is essential that adequate agreements or procedures exist to facilitate data sharing and coordination among the central compiling agency and other data-producing agencies (i.e., source data producing agencies or decentralized compiling agencies). Among other things, these procedures should provide for the effective and timely flow of source data to the central (or decentralized) agencies that compile public sector debt statistics.

6.9 It is also important to ensure that well-established contacts between the staff of the different agencies exist. This will facilitate dealing with any problems, or difficulties, in an expeditious manner, and avoid duplication of data coverage in the different institutions. One way of encouraging cooperation, developing contacts, and resolving problems that arise is to hold regular meetings among staff of the various agencies, at the working level. Not only could these meetings help resolve problems that might be arising, but there would also be an opportunity to notify each other of upcoming developments and possible future enhancements or changes to collection systems. This type of cooperation helps ideas to spread and improvements to be made, allows institutions to understand each other’s position, and helps build important personal contacts.

6.10 If public sector debt statistics are compiled by different agencies, the following considerations must be borne in mind:

• The concepts used and instruments presented should be consistent, or at least reconcilable. When merging together various sources, the central compiling agency must ensure that other contributing agencies are aware of, and supply statistics that are consistent with, core concepts and presentation requirements (such as residence, valuation, etc.), as outlined in this Guide. Indeed, the central compiling agency should develop expertise in these standards. In a sense, the central compiling agency should act as the guardian of these standards to ensure that statistics supplied by the other agencies meet the requirements in terms of the coverage, concepts, valuation principles, basis of recording, as well as the periodicity and timeliness on which these statistics have to be provided.

• It is recommended that the central compiling agency carries out, as far as possible, cross-checking of data on a regular basis (at least once a year). Such cross-checking will involve comparing data with those of creditors as well as other macroeconomic statistics, such as external statistics (International Investment Position and external debt) and monetary and financial statistics. The comparisons can either be undertaken on an individual instrument basis (for example, specific government loans) by the agency responsible for compiling these statistics, and/or at an aggregate level using other macroeconomic statistics datasets and compare, for example, external debt statistics of the government sector.

• There should be mechanisms to ensure that the disseminated public sector debt statistics continue to meet the needs of policymakers and other users. Meetings could be periodically convened with policymakers and other data users to review the comprehensiveness of the public sector debt statistics and to identify any emerging data requirements. New initiatives could be discussed with policy departments and statistical advisory group(s); such discussions provide scope for seeking additional resources. From these discussions, and in consultation with both users and other compiling agencies, the central compiling agency might devise a strategic plan to improve the quality and coverage of public sector debt statistics.

2. Resources

6.11 Resource allocation decisions for the compilation of statistics are the preserve of the authorities in each country and should be periodically reviewed. Nonetheless, the authorities are encouraged to provide, at least, adequate resources to perform existing tasks—that is, adequate staff, financial, and computing resources. In particular, key staff should be knowledgeable and well-versed in public sector debt concepts and compilation methods, and a core contingent of trained public sector debt statisticians should be retained at any point in time. Instructions for performing existing tasks should be maintained. New compilers could be provided formal and on-the-job training in public sector debt compilation methods, including international statistical standards and system procedures for handling and processing of data.

3. Legal framework

6.12 Obtaining appropriate legal support for statistics collection could be a complicated and lengthy process, likely to be undertaken infrequently. Given this, a first step should be to determine whether there is any existing legal support for statistics collection that could be employed to acquire the required information. If not, there may be a need to seek additional legal support.

6.13 The legal framework for the collection of statistical information varies from country to country, depending, not least, on the institutional arrangements and the historical development of statistical gathering. Nonetheless, some elements typically covered include:

• A designation of the type of entities that the central compiling agency can approach for data (for example, entities in the public corporations sector) and for what purpose (for example, to compile flows and stock positions in financial assets and liabilities).

• The boundaries of the compiling agency’s responsibilities, without being so restrictive that the agency does not have the freedom to adapt as new developments emerge.

• A provision for the possibility of imposing penalties on respondents for nonresponse, which should be accompanied by an appropriate legal mechanism for enforcement.²

• While generally not applicable to general government units, because of their accountability function to the public, a clear statement may be needed that information supplied by individual public sector entities outside the general government sector would not be separately disclosed and would only be published in the form of statistical aggregates (except, perhaps, where explicit permission is given from an individual entity to disclose information), along with appropriate penalties for the compiling agency and, in particular, individual employees, if such information is disclosed.

• A prohibition on the use by the authorities of information supplied by individual entities for any purpose other than statistics compilation, thus establishing the independence of the statistics compilation function from other government activity.³ The prohibition should be supported by penalties and a mechanism for their enforcement.

• A prohibition on other government agencies influencing the content of statistics releases.⁴

• The establishment of an oversight committee of independent experts to help ensure the professionalism and objectivity of the compiling agency.

6.14 Within such a legal framework, the public sector debt statistics compiler would have the necessary support for the collection of information from, for example, public corporations or extrabudgetary government units. However, compilers should use the legal framework to help and encourage the public sector entities to report rather than solely relying on the legal backing.

4. Dissemination of public sector debt statistics

6.15 The compilation of public sector debt statistics is undertaken for the ultimate purpose of making them available to policymakers and other users. Comprehensive and timely statistics on public sector debt:

• Allow users to monitor the evolution of the public sector’s debt liabilities and its debt-service obligations over time;

• Can provide early warning signals of possible debt-servicing problems;

• Serve as an indicator of the sustainability of government and public corporations’ policies; and

• Serve as essential inputs for government budget preparation, for approval by parliament, for execution, for forecasting, as well as for compiling other macroeconomic statistics.

6.16 Statistics should be publicly disseminated on a frequent and timely basis, preferably according to a well-established, pre-announced release schedule. The dissemination of statistics could be in print and/or electronic form. As part of the dissemination process, the concepts, definitions, classifications, and methodology used should be documented and disseminated in publication form, at regular intervals.⁵ This metadata could also identify any significant deviations from internationally accepted standards, biases in the statistics, and information about response rates, if any surveys are employed in collecting public sector debt statistics. The IMF’s Special Data Dissemination Standard (SDDS) and General Data Dissemination System (GDDS) provide further guidance on good practices for the timeliness and periodicity of public sector debt statistics being disseminated.⁶

6.17 Often, to meet the legitimate needs of users, statistics will be published that could well be subject to later revision. In such cases, users should be alerted that the initially published data are preliminary and may be subject to revision. If revised data are later published, users should be informed of the revisions, with explanations appropriate the size of the revisions. Also, if major changes to the statistical methodology are to be implemented, it is strongly recommended that users be given advance notice, and sufficient back runs of data provided after the revisions have been published. In general, providing the user with such information is likely to engender greater confidence in the statistics and may help encourage a “culture of reporting” to the compiling agency(ies).

1. Debt office(s)

6.23 A debt office, or an asset and liability management office/department, would usually be the main source for stock (and flow) data on debt securities and loans. Many countries have one debt office—the government debt office—covering the debt securities issued by, and loans of, various components of the public sector. However, it is possible that state and/or local governments and public corporations may have their own debt offices if they are issuers of debt securities and/or large borrowers from abroad. While the remainder of this section discusses a government debt office as data source, the same principles would apply to debt offices at other levels of the public sector.

6.24 Typically, a government debt office is either within the ministry of finance or constituted as a separate agency within the government sector or the central bank. Whether the government debt office is the central compiling agency of public sector debt statistics or not, it is, for reasons outlined earlier in this chapter, also important that the government debt office coordinate, as appropriate, with any other agencies involved with the compilation of public sector debt statistics.

6.25 A government debt office is responsible for debt management, which involves seven functions. These are discussed in detail in the annex to this chapter. Proper records of debt (see paragraphs 6.42–6.44) are fundamental for effective debt management and the availability of accurate and up-to-date statistics determines how effectively the debt office can carry out its other functions—whether they are operational or analytical. It is critical to the smooth functioning of a government debt office that the compilation, recording, and dissemination (if the debt office is the central compiling agency) of debt statistics be undertaken in a timely and comprehensive manner.

2. A balance sheet

3. Questionnaires and periodic surveys

6.32 Information collected at the level of the individual debt instrument is preferable but may not always be possible for a variety of reasons. Balance sheet information may either not be suitable or may be unavailable. If so, compilers will have to use questionnaires and/or periodic surveys to obtain source data for the compilation of public sector debt statistics. Questionnaires may be used to collect source data from extrabudgetary government units, state governments, local governments, and/or some public corporations. Periodic surveys would typically be used when the number of reporting entities is large, such as with local governments.

6.33 In determining the reporting population for a periodic survey covering a large number of entities, such as local governments, various approaches are possible:

• Census: Including in the survey all members of the population;

• Partial coverage collection: Including in the survey all entities above a certain threshold; and

• Stratified random sample: A procedure that groups population components according to the size of selected activity so that entities within different strata have different probabilities of selection.

  It is important that partial coverage collections or sample surveys are supplemented with a census—at least, every three years.

6.34 The design of the questionnaire or survey form is particularly important because it needs to meet all foreseeable data needs—it is unlikely that the form can be changed very frequently, not least because respondents will develop systems to compile the required information—and incorporate quality-control features (for example, cross-checks on the form itself or with related data collections). However, if the survey form is too complex or detailed, there could be a negative impact on quality as respondents may have difficulty supplying the required information.

6.35 Questionnaires and surveys should cover all debt instruments. If the information on debt instruments is provided on an instrument-by-instrument basis, details collected could include name of lender, residence and type of lender, currency, amount outstanding, start date of contract, due date of contract, scheduled payments of principal, interest payments, put options, and relationship between borrower and lender. Similar information could be required for debt securities, although the identity of the lender may be unknown to the borrower. Although this information is detailed, it should be readily available to the entity for its own accounting purposes and, in some instances, may be public knowledge. Also, if possible, it is preferable to collect data on debt liabilities and financial assets in the form of debt instruments together on the same survey form, to ensure consistency in the calculation of net debt statistics.

6.36 Questionnaires and surveys should include clear reporting instructions to ensure high response rates and high-quality responses. Also, seminars and workshops explaining the reporting requirements for respondents are of value to both respondents and the compiling agency, and are encouraged by this Guide.

6.37 There are, at least, three important steps that can be taken to encourage response to the compiling agency:

• There should be legal backing for the surveys, so that as a last resort the compiler has some means of redress if the respondent proves unwilling to report. This legal backing must make clear that any data supplied will be used only for statistical purposes, and this statement must be honored in letter and spirit by the compiling agency. Some respondents may well be reluctant to supply data if they believe their individual data will be shared among other agencies.

• Government departments that have a policy interest in public sector assets and liabilities should be made aware of the reporting needs and encouraged to promote the need for good reporting, whenever possible, when dealing with public corporations. Better data helps promote better-informed policymaking. In other words, the authorities should build the idea of good reporting into their policy objectives in this field. Often, those with policy responsibilities have access to senior officials in public corporations or other levels of government and may deliver the message of good reporting at a more senior level than might be available to the central debt compiling agency.

• The compiling agency, along with other agencies responsible for statistics, should encourage a “culture of reporting.” This is usually achieved in the medium to long term. Steps to encourage a culture of reporting include meeting potential respondents and discussing issues of concern; developing report forms that are in line with management reporting systems and are not overly complex; disseminating and promoting the final output in a transparent manner; and illustrating the usefulness of the final output.

6.38 Even if data are supplied, how can they be confirmed to be reliable? First, if data are supplied in a balance sheet that is part of an audited set of integrated financial statements, this adds a degree of consistency in its own right. Also, if a publicly quoted company supplies data, published accounts from the company are likely to be available against which data can be checked.¹² Second, wherever possible, data should be cross-checked with other sources. For example, transactions data can be compared with changes in stock position data if different sources are used. Net borrowing data from revenue and expenditure accounts, or profit and loss accounts of companies, can be compared with the buildup of net financial assets and liabilities because the two are interrelated. Data on revenue in the form of property income could be compared with stock position data to see whether the implied rates of return on assets are realistic. Data on expense in the form of interest and other property expenses should reflect the cost of liabilities.

4. Other sources

6.39 If no information is available from the debt office, balance sheets, or questionnaires and surveys, other data sources have to be used. In particular, data could be obtained from sources other than those discussed above for liabilities of (i) public sector employees pension funds managed by a public corporation, (ii) unfunded public sector employees pension schemes, and (iii) provisions for calls under standardized guarantee schemes. For example, debt statistics compilers may request actuarial companies to estimate the liabilities for unfunded government employer pension schemes.¹³ To ensure consistency, it is important that the same sources and methods be applied over time.

6.40 In countries with some form of exchange control, the central bank requires approval or registration of external borrowing. In such cases, the central bank can provide information on external borrowing—particularly in the case of public corporations.

1. How should data be collected and compiled?

6.42 To establish a proper debt record, detailed information about debt instruments and related flows needs to be compiled. For those countries that may not have proper records of debt data, there may be a need to first compile a thorough inventory of existing debt (and metadata) in order to establish the debt stock positions, including any arrears that have accumulated on principal and interest. Once the debt stock position is known, procedures should be put in place to obtain, on a regular basis, information on existing and new borrowing, as well as information on other transactions and other economic flows that affect the debt stock position. There may be a need to establish formalized institutional arrangements for the comprehensive and timely flow of information to the debt office. Table 6.2 lists information that should be collected for each debt instrument. This table is explained in more detail below.

Table 6.2.

Information to be Compiled on Each Debt Instrument

Table 6.2.

Information to be Compiled on Each Debt Instrument

Type of Information	Description
I. Details of Borrowing Instrument
Purpose of borrowing Agreement date Type of instrument Effective date Type of borrowing Amount borrowed Currency of borrowing Participants: • Borrower • Implementing agency • Creditor • Disbursement agency • Creditor insurer Guarantee status	Descriptive title Date agreement has been signed Type of borrowing instruments Date borrowing becomes effective Whether single currency or multicurrency or multi-tranche Original amount borrowed or revised amount after cancellation or enhancement Original currency and currencies of disbursement and repayments Whether government, public corporations, or private sector Agency in charge of implementing project Name and type of creditor (multilateral, bilateral, etc.) Name, if different from lender Name (and country, if nonresident) Borrowing by general government units, public corporations, or the private sector guaranteed by government, and percentage guaranteed
Insured	Whether borrowing is insured by export guarantee agency in creditor country and percentage guaranteed
Economic sector	Economic sector receiving the borrowing (for example, general government, public nonfinancial corporations, public financial corporations, etc.)
Use of funds	Whether to finance a project, etc.
II. Disbursements
Disbursement period Method of disbursement Expected disbursement pattern/profile Actual disbursement	Period during which disbursement is to take place Such as direct disbursement or reimbursement Forecast of how the borrowing will be disbursed Currencies and amount of each disbursement taking place
III. Borrowing Terms
Interest	Information on interest should include: • Interest type: fixed or variable rate • For variable rate: specify interest base/reference and margin/spread • Interest period: dates of payments • Basis for interest calculation (conversion factor: daily, monthly, semiannual, annual, etc.) • Months: actual number of days or 30-day month • Days in interest year (360/365)
Commitment fee Penalty fees Other fees Principal	Rate levied on undisbursed (full or partial) amount Charges for late payment of interest and principal Such as agency fee, management fee, front-end fee Maturity: repayment period/profile Type of repayment: bullet, equal or annuity-based, etc.
IV. Actual Debt-Service Payments	For each payment (of interest, principal, other charges) made: • Date, currency, and currency of transaction; amount of transaction in original currency, currency of transaction, domestic currency, and perhaps U.S. dollar and SDR For multicurrency borrowing: equivalent amount paid in borrowing currency
V. Exchange Rate	Exchange rates on each transaction date for relevant currency vis-à-vis the local currency Exchange rates for end of period (daily, weekly, month, quarter, year)
VI. Interest Rates	Prevailing variable interest rates of base/reference rate used by the creditor for each interest period
VII. Debt Restructuring	Changes in terms as a result of debt reorganization, through rescheduling, refinancing (voluntary or involuntary), write-off, etc. Date required: - Debt concerned; arrears, consolidation period - Debt-relief terms (debt forgiveness, reschedule) - Terms for rescheduled debt (applicable interest rate, repayment profile) - Transactions on actual debt-service payments or for rescheduled debt - Other transactions from buy-back or conversion/swap
VIII. Financial Derivatives	•Transactions arising from financial derivative contracts • Positions measured both in market value and notional amounts in forwards (including swaps) and options

View Table

Table 6.2.

Information to be Compiled on Each Debt Instrument

Type of Information	Description
I. Details of Borrowing Instrument
Purpose of borrowing Agreement date Type of instrument Effective date Type of borrowing Amount borrowed Currency of borrowing Participants: • Borrower • Implementing agency • Creditor • Disbursement agency • Creditor insurer Guarantee status	Descriptive title Date agreement has been signed Type of borrowing instruments Date borrowing becomes effective Whether single currency or multicurrency or multi-tranche Original amount borrowed or revised amount after cancellation or enhancement Original currency and currencies of disbursement and repayments Whether government, public corporations, or private sector Agency in charge of implementing project Name and type of creditor (multilateral, bilateral, etc.) Name, if different from lender Name (and country, if nonresident) Borrowing by general government units, public corporations, or the private sector guaranteed by government, and percentage guaranteed
Insured	Whether borrowing is insured by export guarantee agency in creditor country and percentage guaranteed
Economic sector	Economic sector receiving the borrowing (for example, general government, public nonfinancial corporations, public financial corporations, etc.)
Use of funds	Whether to finance a project, etc.
II. Disbursements
Disbursement period Method of disbursement Expected disbursement pattern/profile Actual disbursement	Period during which disbursement is to take place Such as direct disbursement or reimbursement Forecast of how the borrowing will be disbursed Currencies and amount of each disbursement taking place
III. Borrowing Terms
Interest	Information on interest should include: • Interest type: fixed or variable rate • For variable rate: specify interest base/reference and margin/spread • Interest period: dates of payments • Basis for interest calculation (conversion factor: daily, monthly, semiannual, annual, etc.) • Months: actual number of days or 30-day month • Days in interest year (360/365)
Commitment fee Penalty fees Other fees Principal	Rate levied on undisbursed (full or partial) amount Charges for late payment of interest and principal Such as agency fee, management fee, front-end fee Maturity: repayment period/profile Type of repayment: bullet, equal or annuity-based, etc.
IV. Actual Debt-Service Payments	For each payment (of interest, principal, other charges) made: • Date, currency, and currency of transaction; amount of transaction in original currency, currency of transaction, domestic currency, and perhaps U.S. dollar and SDR For multicurrency borrowing: equivalent amount paid in borrowing currency
V. Exchange Rate	Exchange rates on each transaction date for relevant currency vis-à-vis the local currency Exchange rates for end of period (daily, weekly, month, quarter, year)
VI. Interest Rates	Prevailing variable interest rates of base/reference rate used by the creditor for each interest period
VII. Debt Restructuring	Changes in terms as a result of debt reorganization, through rescheduling, refinancing (voluntary or involuntary), write-off, etc. Date required: - Debt concerned; arrears, consolidation period - Debt-relief terms (debt forgiveness, reschedule) - Terms for rescheduled debt (applicable interest rate, repayment profile) - Transactions on actual debt-service payments or for rescheduled debt - Other transactions from buy-back or conversion/swap
VIII. Financial Derivatives	•Transactions arising from financial derivative contracts • Positions measured both in market value and notional amounts in forwards (including swaps) and options

6.43 Data compilation is best undertaken on an instrument-by-instrument basis, tranche by tranche, and in its original currency. For each borrowing instrument, there are basically three types of information that need to be compiled: (1) the core details and terms that will produce the amortization and disbursements tables, as well as the stock position data; (2) details on actual disbursements, as well as the changes in the committed undisbursed amount if, say, there are cancellations and/or increases (for example, with a project loan); and (3) details on actual debt-service transactions. There are other types of information required, and these are described below (under the heading “Additional Data Requirements”).

6.44 If the debt instrument is tradable, additional information will be required in order to attribute ownership by residence. This information may come from a different agency, which is responsible for capturing information on the nonresident ownership of traded securities. Methods of capturing information on nonresident ownership of traded securities are set out in Chapter 7.

2. How should information be stored?

6.55 A debt office should store information in an efficient and comprehensive computer-based debt management system (CBDMS) that can undertake a number of tasks and so support both operational and policy functions. Typically, a CBDMS should be able to do¹⁶:

• Debt recording (loan-by-loan);

• Debt reporting (loan-by-loan and on an aggregate basis);

• Debt analysis; and

• Linkages with other packages and systems of the public sector unit.

3. How should data be validated?

6.56 Data validation is crucial in ensuring the compilation of reliable public sector debt statistics which, in turn, are essential for the management and formulation of a country’s fiscal and other macroeconomic policies and strategies. For this reason, the Guide recommends that procedures be put in place at various stages of the data compilation and recording process to ensure that all data captured are properly validated and reconciled with other data sources. Although data provided to and supplied by the different institutions and departments—both international and domestic—should be checked for mutual consistency, these data may not be identical. But the data validation process should ensure that where differences do exist, the underlying factors for the differences are identified and explained to users of the data.

6.57 Among the various procedures and actions that can support data validation are:

• Verification of data recorded in data entry sheets with data extracted from debt instrument agreements, statements, and other documentation;

• Systems with inbuilt validation procedures to check for inconsistencies at the time of the recording of the information in debt recording and management systems;

• Description of procedures for treating different types of debt and their components, including sources of data in a Debt Procedures Manual—a “how-to” manual that accumulates knowledge and passes on experiences;

• Periodic reconciliation of data obtained from one source with other sources of information—for example, data on debt-service payments can be checked with records kept by the foreign exchange payment department in the central bank; loan balances could also be verified with creditors and debtors on a regular basis; cash flows could be reconciled with bank accounts and with accrual records; and

• An audit mechanism that is consistent with the general rules of public finance control.