I. Introduction

1. On June 3, 2002, the Executive Board held further discussions on possible revisions to the formulas used to calculate members' quotas. ¹ A measure of openness has traditionally been included in the quota formulas in the form of a variable containing the sum of current receipts and current payments, and most Directors reiterated their support for the continued inclusion of such an openness variable. Directors also discussed broadening the openness measure by including in the formulas a variable such as the international investment position (IIP), specifically to capture financial openness. However, as noted in SM/02/132, and during the Board discussion, it would not at this stage be possible to include the IIP for all member countries given that many countries do not at present report comprehensive IIP data. In view of the interest in including the IIP in the quota formulas, many Directors encouraged additional work to improve the collection and reporting of IIP data.

2. The IIP, as a balance sheet, is also important for vulnerability assessment because it provides a snap shot of the levels, sectoral distribution, and maturity of a country's external liabilities (e.g., external debt) and also the size and composition of its external claims (e.g., international reserves and banks' foreign claims) that may be available to meet its external obligations. In their meeting on data provision to the Fund for surveillance on May 10, 2002, Directors noted that data for vulnerability assessment should include frequent and comprehensive data on international reserves; detailed data on IIPs and capital flows; maturity profile and repayment schedules of external and public sector debt; and financial soundness indicators, including corporate sector data. ^{2 3} The objective should be to work toward compiling the above data with the necessary detail depending on country circumstances and characteristics and in line with the relevant statistical methodologies.

3. IIP statistics are also among the items of information that members are required to report to the Fund under Article VIII, Section 5 of the Articles of Agreement. More specifically, Article VIII, Section 5 (a) (vii) requires members to provide information on the "international investment position, i.e., investments within the territories of the member owned abroad and investments abroad owned by persons in its territories so far as it is possible to furnish this information." In recent years, the Fund has impressed upon countries the need to furnish such information and staff reports have informed the Directors of these matters. ⁴ In determining whether a member has complied with its obligation under Article VIII, Section 5, the Fund must take into account the member's ability to provide the relevant information.

4. In light of the prominence given to openness and vulnerability indicators at recent Executive Board discussions, this paper has been prepared for the information of Directors. It discusses the present state of the IIP data and a number of Statistics Department (STA) initiatives to assist countries to compile these statistics. Section II describes the data on IIPs that are currently available to STA, while Section III highlights some of the difficulties that countries face in undertaking new IIP compilations. Section IV describes some of the tools provided by STA and its inter-agency partners that facilitate the compilation of IIP statistics. Section V reviews several initiatives to develop further the IIP statistics in member countries and responds to Directors' request that additional work on data issues related to IIPs be undertaken. Section VI presents conclusions.

II. Present State of the Data

5. The IIP represents at a particular point in time an economy's balance sheet showing its stock of external financial assets and liabilities. The IIP consists of claims on nonresidents, liabilities to nonresidents, monetary gold, and SDRs. The difference between an economy's external financial assets and liabilities is the economy's net IIP. As the IIP is a comprehensive framework covering the complete range of a country's external assets and liabilities, the source data requirements are so extensive that it is most often compiled on an annual basis. However, in some countries it is compiled quarterly.

6. The fourth edition of the Balance of Payments Manual (BPM4), published in 1977, defined the IIP but did not present a framework and a set of standard components. The fifth edition of the Fund's Balance of Payments Manual (BPM5), which was published in 1993, first addressed IIP statistics comprehensively. Twenty-five countries reported IIP data to the IMF in 1993.

7. The IMF began publication of balance of payments and IIP data on the BPM5 basis in 1995. STA regularly invited countries to contribute annual IIP data, including partial data, for publication in the Balance of Payments Statistics Yearbook and in International Financial Statistics. By end-2000, the number of reporting countries had risen steadily, and STA published IIP statistics for 63 countries.

8. STA initiated an outreach program in early 2001 by writing to central bank governors and heads of statistical agencies to encourage those countries not reporting IIP data to bolster their efforts to do so and to encourage those countries reporting partial data to report more comprehensively. As of end-July 2002, STA's database contained IIP data for 80 countries. Three quarters of these countries reported comprehensive data; ⁵ the remainder provided partial data (see Appendix 1).

9. In addition to the 80 countries that report IIP data to STA, the majority of other IMF member countries compile data on selected components of their external sector position, and these data provide useful information with which to construct important elements of an IIP statement. As well, these partial data are often available with greater periodicity and timeliness than the full IIP statement. These include data on the following:

• International reserves.

• Other monetary and financial statistics, which includes data on the foreign assets and liabilities of the monetary authorities, deposit money banks, and nonbank financial institutions.

• External debt statistics. Although data may not be available for the private nonbank sector, many countries compile data on the external debt of the government sector and frequently also cover the external debt of government enterprises.

• More selectively, countries also gather external position data from administrative information, such as portfolio investment positions from stock exchanges, direct investment information from investment approval agencies, etc.

III. Difficulties in Compiling IIP Data

A. Survey on Implementation of BPM5

10. The difficulties that countries face in compiling IIP statistics, including the considerable resources involved, have been emphasized on various occasions. In early 1995, STA conducted a survey of its members to assess progress they were making in implementing the guidelines of the BPM5. Of the more than 50 responses received, many of the developing and emerging market countries indicated that they found or anticipated finding the greatest difficulty in implementing IIP collections, noting especially the lack of reliable data on private sector investment positions, particularly in view of the liberalization of capital controls that had taken place in recent years. They also cited the need to develop new surveys, to secure greater intragovernmental cooperation, and to overcome difficulties in estimating the market value of certain international assets. ⁶

B. Conference on Capital Flow and Debt Statistics

11. More recently, compilers had an opportunity to voice concerns regarding calls for improved external debt statistics, which are an important component of the IIP. In February 2000, the IMF, in cooperation with the Financial Stability Forum's Working Party on Capital Flows, sponsored a conference on capital flow and debt statistics. At this conference some 120 senior-level data users, policy makers, and data compilers considered what actions and resources would be required to provide better and more timely data on capital flows and external debt and what actions should be given priority. ⁷

12. At the conference, policymakers, data users, and some compilers stressed the need for more comprehensive and timely capital flow and external debt statistics. Data compilers from industrial countries and offshore centers generally viewed the dissemination of such data as a lower priority, which reflected national statistical priorities as well as concerns regarding resources, compilation difficulties, and respondent burden. Most data compilers felt pressured to produce more timely and comprehensive data, but many lacked the necessary additional resources to undertake the work. Compilers pointed out that respondents to national statistical surveys had expressed concerns about increased demands. ⁸ A number of participants also pointed out that a key issue to address was the undercoverage of the activity of the nonbank private sector, but this sector was often reluctant to provide data. ⁹ All of these issues must be addressed when compilers decide to compile IIP statistics.

IV. Tools to Facilitate Compilation of IIP Statistics

13. A number of tools have been developed to assist countries in their efforts to compile and disseminate IIP statistics. These include statistical manuals and compilation guides, the conduct of coordinated surveys under the auspices of the Fund, and research activities undertaken by international statistical committees and task forces. In the five examples given below, STA strives to coordinate these and other initiatives to ensure consistency across countries and conformity with agreed international statistical guidelines.

A. Methodological Materials, Training, and Technical Assistance

14. The IMF and its inter-agency partners facilitate the efforts of countries to compile IIP statistics through the provision of methodological materials, training, and technical assistance. In addition to the BPM5, the IMF published in 1995 the Balance of Payments Compilation Guide, which provides practical guidance for using sources and methods to compile the balance of payments and IIP statistics. The Guide presents the relative strengths and weaknesses of a variety of approaches and notes source data adjustments required to derive, in conformity with recommendations of the BPM5, data on flows and stocks. The BPM5 and the Guide, together with the Balance of Payments Textbook, form a trilogy of publications that provide a comprehensive range of information on the compilation of balance of payments and IIP statistics. ¹⁰ The IMF uses this trilogy in its training program at headquarters and in regional training institutes as well as in its technical assistance program.

15. To train compilers on the concepts and methodologies set out in the BPM5, the first IMF Institute balance of payments course based on the new methodology was offered in late 1992. ¹¹ In the following years, the training program was expanded to encompass the various regional training institutes that had opened. In 2000, training in external debt statistics was introduced. In 2001, STA conducted 14 training courses and seminars in external sector statistics. In total, some 350 persons received training in balance of payments, IIP, and external debt compilation in that year. In 2002, there are plans to conduct 13 courses and seminars in external sector statistics. The number of training courses and seminars offered in each of the past two years was more than double the number offered in each of the preceding four years, which reflected increased training due to preparations for the 2001 Coordinated Portfolio Investment Survey and the introduction of the new external debt guide (see below).

16. In addition to training, STA has an intensive program of technical assistance in balance of payments statistics and has responded positively to requests from member countries for technical assistance in the development of IIP statistics and other facets of implementing BPM5 guidelines. On average, STA has conducted about 30 short- and long-term technical assistance missions a year in external sector statistics since the end of 1996.

17. Related to these activities, the number of countries reporting balance of payments statistics in the format of the BPM5 has grown to 144 ¹² and, as previously indicated, 80 countries currently report IIP statistics.

B. IMF Committee on Balance of Payments Statistics

18. The terms of reference of the IMF Committee on Balance of Payments Statistics, which was formed in 1992, includes a requirement to "bring to the attention of the IMF new developments that impact on the compilation of statistics on cross-border transactions or related stocks of financial assets and liabilities, and work with the IMF in determining how these activities should be treated in accordance with the BPM5." Each year, the Committee reviews a number of methodological issues that have an impact on external sector statistics. In 2001, these included, inter alia, the statistical treatment of nonperforming loans, employee stock options, and repurchase agreements and securities lending, and the statistical definition of resident units, such as special purpose entities in offshore financial centers. The Committee's program of work helps STA to stay abreast of statistical developments in the rapidly changing environment of globalization and in turn to inform national compilers in the Committee's Annual Report and in the Balance of Payments Statistics Newsletter, both of which are on the IMF's external website.

C. Coordinated Portfolio Investment Survey

19. Portfolio investment is a significant component in many countries' IIP, and the Fund-sponsored Coordinated Portfolio Investment Survey (CPIS) helps countries develop stock data on portfolio investment assets on a consistent basis in accordance with the BPM5 methodology. The 2001 CPIS is underway and preliminary results are expected towards the end of 2002. Sixty-nine countries and jurisdictions, including a significant number of offshore financial centers, participated in the 2001 CPIS compared with the 29 major investing countries that participated in the first survey (end 1997). The 2001 CPIS will provide more comprehensive global information on cross-border ownership of portfolio investment securities (equities and debt securities). The objectives of the surveys are (i) to collect comprehensive information on the geographical distribution of cross-border holdings of foreign portfolio investment securities for use in the compilation and/or improvement of IIP statistics and (ii) to exchange, with the assistance of STA, the bilateral data collected in the national surveys. The data exchanged provide creditor-source measures of countries' external indebtedness reflecting nonresident holdings of debt securities issued by individual countries.

20. In May 2002, the IMF wrote to its member countries to inform them of the decision taken by the IMF Committee on Balance of Payments Statistics to conduct the CPIS on an annual basis and to invite countries' continued participation. The response received to date has been overwhelmingly positive. STA expects that the 2002 CPIS will reflect further improvements in the coverage and methodology, which will provide more authoritative global statistics as well as improved data for IIP compilation at the national level. The CPIS contributes a very important component in the development of a comprehensive IIP statement by countries. The Coordinated Portfolio Investment Survey Guide (2^nd edition, 2002), which was prepared to assist countries to participate in the CPIS, provides practical guidance on conducting different kinds of portfolio investment surveys.

D. Inter-Agency Task Force on Finance Statistics

21. Two important initiatives to improve access to and understanding of data on external debt were launched in 1998 in conjunction with the work program of the Inter-Agency Task Force on Finance Statistics. ¹³ First, since March 1999 the Joint BIS-IMF-OECD-World Bank Statistics on External Debt, ¹⁴ which are quarterly releases of statistics on external debt for 176 developing and transition countries, have been disseminated on the websites of several of the Task Force members. Although these data do not provide a comprehensive and consistent measure of a country's external debt, they do bring together the best and most timely internationally comparative data currently available on external debt. Perhaps the most important component of these series is the BIS international banking statistics, which provide a perspective of countries' external liabilities to and claims on nonresident banks, both for the bank and nonbank sectors.

22. The BIS has undertaken a considerable amount of work to improve the coverage and classification of the international banking statistics so that they can be used by statisticians for compiling and/or checking certain categories of external claims. STA has on several occasions informed national compilers of the availability of these data and their use for statistical compilation. The BIS also makes available to central banks data from its international securities database, which provide compilers with information on the securities issued abroad by entities resident in their countries. These data can also be used for compiling and/or checking country balance of payments and IIP data.

23. Second, the Task Force undertook to prepare, in close consultation with national compilers of external debt and balance of payments statistics, new international guidance on external debt statistics. In December 2001, a final draft of External Debt Statistics: Guide for Compilers and Users, subject only to editing, was posted on the IMF's external website, along with the other statistical manuals. ^{15 16} The Guide provides guidance on (i) the concepts, definitions, and classifications of external debt data; (ii) the sources and techniques for compiling these data; and (iii) the analytical use of these data. ¹⁷ The Guide is intended to be of use to both compilers and users of external debt statistics.

E. Direct Investment Methodology

24. In 1997, the Fund and the OECD conducted a survey on country data sources, collection methods, and dissemination and methodological practices for direct investment statistics. The survey results were published and additional details were made available electronically to balance of payments compilers and staff of international organizations. ¹⁸ An update was conducted in 2001 and the results-summary report as well as detailed metadata for most of the participating countries-are expected to be published in 2002. This information is an aid to countries to compare their practices for consistency with BPM5 methodology and to introduce and/or revise surveys on foreign direct investment, an important component in the IIPs of most countries.

V. Other Work in Progress to Encourage Compilation of IIP Statistics

25. An increase in the number of IIP-reporting countries and in the comprehensiveness of reported data can be achieved only by a significant effort by the countries themselves, given the complexity of the work and the significant resource costs to national compilers and data providers. In response to Directors' request that additional work on data issues related to IIPs be undertaken, the following are several initiatives in train that have important implications for the development of IIP statistics:

26. First, has been the inclusion of the IIP in the IMF's Special Data Dissemination Standard (SDDS). The transition period for the dissemination of IIP data expired on December 31, 2001, and, therefore, data for this date should have been disseminated no later than June 30, 2002. The SDDS requires annual data to be disseminated with six months timeliness; however, subscribers who have begun disseminating data for the new quarterly external debt data category may take up to nine months to disseminate their IIP, as explained below. ¹⁹

27. By July 31, 37 subscribers had disseminated complete IIP data on their national web sites. Of the remaining thirteen subscribers, one is using a fiscal year ending in March and therefore IIP data are to be disseminated by end-September, five are using available flexibility options for timeliness, and another three have chosen to begin disseminating quarterly external debt data ahead of schedule-the transition period ends in March 2003-to avail themselves of the additional three-months' timeliness for disseminating IIP data. Two countries have disseminated partial annual IIP data and are working with Fund staff in addressing outstanding coverage issues. Two subscribers are working with Fund staff in addressing outstanding coverage issues on quarterly external debt, in order to avail themselves of an additional three-months' timeliness for disseminating IIP data.

28. Second, a guidance paper-International Investment Position: A Guide to Data Sources-on the practical aspects of IIP compilation is being finalized and will be transmitted to countries that do not report IIP data to STA. The paper's major focus is a discussion of data sources that may be used to construct IIP statistics and to provide practical guidance on IIP compilation. The covering letter to national compilers would stress the growing importance of IIP statistics and emphasize that the objective should be to work toward compiling the data with the necessary detail depending on country circumstances and characteristics, and in line with the relevant statistical methodologies. For example, for many countries these components would include foreign direct investment and private sector external debt, while for others it would be for foreign asset holdings.

29. Third, in collaboration with the Policy Development and Review Department, the support of the Inter-Agency Task Force on Finance Statistics has been sought in developing, at the country level, improved data for vulnerability assessment. In the Executive Board paper on data provision to the IMF, discussed on May 10, 2002, balance sheet data on the basis of the IIP was among the identified data requirements for vulnerability assessment. Following a meeting of the Task Force last May, this initiative will identify, inter alia, repositories of information that could be used to make an assessment of data gaps across a broad range of countries. A paper will be prepared that will prioritize data requirements and also examine how to motivate countries further to introduce or expand statistical collections in this area.

VI. Conclusions

30. While much has been achieved in recent years in improving the availability of IIP data, more member countries need to make the necessary commitment to compile IIP statistics and to provide the necessary resources to develop appropriate compilation systems for those data series most relevant to their circumstances. STA will continue to support this work through its training and technical assistance programs, as well as updating, as required, international statistical guidelines. Area Departments can also be encouraged to engage countries to compile IIP statistics when this information is needed for surveillance.

31. The process of establishing comprehensive IIP collections will take time, and it is likely to be many years before a global IIP statement can be compiled along the lines as produced for balance of payments statistics. ²⁰ The more difficult areas include data on foreign direct investment and other nonbank private sector external assets and liabilities. As noted earlier, countries are making good progress in the development of data on portfolio investment assets through their participation in the CPIS.

32. The staff intends to update Directors further on this work in the context of discussions on the Sixth Review of Fund's Data Standards' Initiatives and on Data Provision to the Fund for Surveillance that are expected to come to the Board in the second half of 2003.

Appendix 1 Countries Reporting IIP Statistics to IMF Statistics Department
(As at end-July 2002)
Report comprehensive information 1			Report partial information 2 3
Argentina	France	Portugal	Azerbaijan
Armenia	Germany	Romania	Bangladesh
Australia	Greece	Russia	Cambodia
Austria	Hungary	Rwanda	Chile
Bahrain	Iceland	Senegal	China, Hong Kong SAR
Belarus	Israel	Slovak Republic	Costa Rica
Belgium	Italy	Slovenia	Jordan
Benin	Japan	South Africa	Korea
Bolivia	Kazakhstan	Spain	Lesotho
Botswana	Kyrgyz Republic	Swaziland	Macedonia
Bulgaria	Latvia	Sweden	Malaysia
Canada	Lithuania	Switzerland	Maldives
Colombia	Malta	Tanzania	Mauritius
Croatia	Moldova	Thailand	Myanmar
Cote d' Ivoire	Netherlands	Togo	Namibia
Czech Republic	New Zealand	Tunisia	Norway
Denmark	Panama	United Kingdom	Turkey
El Salvador	Paraguay	United States	Uganda
Estonia	Peru	Uruguay	Venezuela
Finland	Poland	Vanuatu
		Yemen
1 Comprehensiveness was assessed by the availability of data for most of the broad BPM5 functional categories of the IIP and by the provision of data in recent years.
2 Ten of these countries were designated partial reporters because they lacked information on foreign direct investment; four countries did not provide information in respect of current years; and the remainder had gaps in other data sets, e.g., portfolio investment. 3 Since end-June 2002, some of the SDDS-subscribing countries have disseminated more comprehensive IIP data and thus are in compliance with the coverage and timeliness requirements of the SDDS.