Financial Soundness Indicators-Experience with the Coordinated Compilation Exercise and Next Steps1

Overview. This paper reports on the experience with the work program on Financial Soundness Indicators (FSIs) and offers proposals for taking forward the work on FSIs. The work program aimed at (i) increasing member countries’ FSI compilation capacity and supporting their compilation efforts; (ii) expanding reporting and analysis of FSIs in the work of the Fund; and (iii) undertaking further analytical work on FSIs. In this context, a Coordinated Compilation Exercise (CCE) for FSIs was conducted, and experience was gained with the use of FSIs in Article IV surveillance, the Financial Sector Assessment Program (FSAP), and the interdepartmental Vulnerability Exercise (VE). The paper proposes that the Fund aintains an ongoing role in collecting and disseminating FSIs.


Overview. This paper reports on the experience with the work program on Financial Soundness Indicators (FSIs) and offers proposals for taking forward the work on FSIs. The work program aimed at (i) increasing member countries’ FSI compilation capacity and supporting their compilation efforts; (ii) expanding reporting and analysis of FSIs in the work of the Fund; and (iii) undertaking further analytical work on FSIs. In this context, a Coordinated Compilation Exercise (CCE) for FSIs was conducted, and experience was gained with the use of FSIs in Article IV surveillance, the Financial Sector Assessment Program (FSAP), and the interdepartmental Vulnerability Exercise (VE). The paper proposes that the Fund aintains an ongoing role in collecting and disseminating FSIs.

I. Introduction

1. The financial turmoil of the late 1990s highlighted the need for macroprudential analysis. This was seen as important for identifying vulnerabilities in the financial system as a whole, including to international capital flow reversals as well as to shocks to the corporate and household sectors. The recent turbulence in subprime mortgage markets that have hit the balance sheets of banks and other financial corporations and have created a credit and liquidity squeeze have again pointed to the need for improved information on the soundness of financial systems. However, the paucity of data in this area, and a lack of dissemination and cross-country comparability have been recognized as key stumbling blocks.

2. In response, the Fund has worked closely with national agencies and regional and international institutions to develop a set of Financial Soundness Indicators (FSIs).3 The Executive Board endorsed a set of core and encouraged indicators (Box 1) in June 2001 and a work program in June 2003 aimed at (i) increasing member countries’ FSI compilation capacity and supporting their compilation efforts; (ii) expanding reporting and analysis of FSIs in the work of the Fund; and (iii) undertaking further analytical work on FSIs.4

Financial Soundness Indicators: The Core and Encouraged Sets

article image

Or in other markets that are most relevant to bank liquidity, such as foreign exchange markets.

3. The Executive Board last discussed the Fund’s work on FSIs in June 2003. At that time, Executive Directors underscored the importance of FSIs in assessing financial sector soundness, facilitating financial sector surveillance, and increasing the transparency of the international financial system. They encouraged the work on the Financial Soundness Indicators: Compilation Guide (Guide), which they saw as a milestone in establishing a standard reference on the concepts and definitions, data sources, and techniques to be used in the compilation of cross-country comparable FSIs. The Board endorsed the proposal for the Fund to conduct a Coordinated Compilation Exercise (CCE) for FSIs involving around 60 member countries to support country compilation efforts, as well as the Fund’s role in promoting FSIs through the Financial Sector Assessment Program (FSAP), Article IV, and UFR missions, workshops, and technical assistance. They generally supported the further development of guidance and analytical work on the use of FSIs and their expanded reporting and analysis in bilateral and multilateral surveillance, including Article IV reports and the Global Financial Stability Report.5

4. A progress report on the FSI work program was submitted for the Executive Board’s information in June 2005.6 That paper reviewed the steps taken leading to the finalization of the Guide, progress on the analytical work on FSIs and on the CCE, and collaboration on FSIs of Fund staff with the staff of regional and international institutions. The paper envisaged that the staff would report back to the Board in 2007 on the experience with the CCE and on proposals for future work of the Fund on FSIs.

5. The Fund’s work on FSIs was also raised at the sixth review of the Fund’s data standards initiatives, including the possibility of including FSIs in the Special Data Dissemination Standard (SDDS).7 However, a decision on this matter was postponed until the Board discussion of the present report.

6. This paper reports on the experience with the FSI work program and offers proposals for taking forward the work on FSIs. Building on the CCE, the FSIs that are already published as part of the GFSR, as well as the experience with FSIs in the Fund’s work in Article IV surveillance, the Financial Sector Assessment Program (FSAP), and the interdepartmental Vulnerability Exercise (VE), the paper proposes that the Fund maintain an ongoing role in collecting and disseminating FSIs, but it argues that it would be premature to include FSIs in the SDDS.

7. The paper is organized as follows: Section II provides an overview of and discusses the experience with the CCE. Section III discusses developments and experience in the use of FSIs in the Fund. Section IV discusses proposals regarding the way forward with FSIs in the work of the Fund. Section V offers issues for discussion by the Board. A companion Background Paper details the experience with the CCE, and explains the issues that arose regarding the FSI compilation methodology as documented in the Guide and the arguments that have been taken into account in considering specific amendments to be made to the Guide.

II. The Coordinated Compilation Exercise

A. Scope and Modalities

8. Following the finalization of the Guide, in March 2004 the Fund’s Statistics Department (STA) launched a voluntary CCE for FSIs.8 It aimed at (i) developing the capacity of participating member countries to compile FSIs important to the surveillance of their financial systems, (ii) promoting cross-country comparability of FSIs, (iii) coordinating efforts by national authorities to compile FSIs, and (iv) disseminating the FSI data compiled in the CCE, along with metadata,9 to increase transparency and strengthen market discipline. The terms of reference of the CCE were established in consultation with participating countries and international organizations in a plenary meeting for CCE countries in November 2004. Countries participating in the CCE were required to compile the 12 core FSIs as of end-2005 and provide them to the Fund along with underlying data series and related metadata. CCE participating countries (Box 1 of the Background Paper) were also encouraged to compile some or all of the 28 encouraged FSIs and provide them along with their metadata. It was agreed that all FSI data and metadata provided to the Fund in the context of the CCE would be disseminated to the public through the Fund’s website.

9. A CCE/FSI Reference Group was established to foster collaboration with other international and regional institutions in the area of FSI work. The Reference Group, comprising 17 international and regional agencies, provided a forum for the exchange of information on FSI-related work of various institutions, and opportunities to identify areas where harmonization of data compilation and reporting methodologies was needed to minimize duplication of effort on the part of countries and institutions.10

10. To facilitate the implementation of the CCE work program, the Fund provided technical assistance and training to countries, developed standard data and metadata report forms, and created web pages for the dissemination of CCE information.11 STA provided technical assistance and training to countries on the methodology of compilation of FSIs recommended in the Guide and on the preparation of metadata. Standard data and metadata report forms were developed by STA, in consultation with CCE participating countries and the Monetary and Capital Markets Department (MCM), with a view to ensuring uniform and consistent collection and reporting of data and metadata. A set of web pages on the Fund’s external website was developed by STA with the assistance of TGS to disseminate the information collected in the CCE.12 FSI data and metadata were made available to Fund internal users through the Economic Data Sharing System (EDSS).

11. A plenary meeting for CCE participating countries and members of the CCE/FSI Reference Group was convened in May 2007 upon completion of the CCE. The meeting was aimed at drawing conclusions from the CCE and provided a forum for participating countries and collaborating regional and international institutions to exchange views on the CCE, FSIs and their methodology, and the way forward regarding the compilation and reporting of FSIs. Countries’ views on these matters are summarized in this paper and reported in greater detail in the Background Paper.

B. Experience with the CCE

12. A total of 57 out of the 62 CCE participating countries have so far submitted their final FSI data and metadata as envisaged in the FSI work program (Table 1). The CCE revealed the status of countries’ capacity and practices in compiling FSIs, informed about the degree of adherence to the recommendations of the Guide, identified areas where the methodology recommended in the Guide needed to be amended, and contributed to ascertaining countries’ technical assistance needs in the compilation and dissemination of FSIs. The experience with the CCE is described in greater detail in the companion Background Paper, which also presents the FSI data provided by the 57 CCE reporting countries.

Table 1.

FSIs Compiled by Countries in the CCE

article image
Source: CCE.

A given FSI may be compiled more than once by a given country using different consolidation bases.

Or in other markets that are most relevant to bank liquidity, such as foreign exchange markets.

13. The scope of data compiled for the CCE reflects a number of factors. In particular, data availability on deposit takers, which remains central to financial soundness analysis, is better than for other sectors. Virtually all countries, as required, compiled the core FSIs, which focus on deposit takers. Encouraged FSIs for deposit takers were also compiled more regularly than FSIs for other sectors, such as households and nonfinancial corporations (Figure 1). Indicators for the latter were on average the least compiled, partly reflecting limited availability of the underlying data.

Figure 1.
Figure 1.

Country Submissions of FSIs by Sector

(In percent of total countries)

Citation: Policy Papers 2007, 043; 10.5089/9781498333245.007.A001

14. The CCE revealed a diversity of methodologies in compiling FSIs. Explanatory factors include: (i) diversity of supervisory and accounting practices across CCE participating countries; (ii) data availability in the participating countries; (iii) costs involved in collecting additional data to fully implement the recommendations of the Guide; and (iv) diversity of country views on the appropriate methodology of FSI compilation. The main methodological departures from the recommendations in the Guide were on institutional coverage (consolidation basis), the elimination of certain positions and flows among institutions in this coverage (inter-group consolidation adjustments), and on a few accounting issues under the International Financial Reporting Standards (IFRS).13

15. A variety of consolidation bases were used by CCE participating countries in compiling FSIs for deposit takers.14 Only a small proportion of FSIs for deposit takers were compiled using the domestically controlled, cross-border (DCCB) consolidation basis, which is the approach recommended by the Guide for compiling FSIs. The most commonly used bases were the domestic consolidation (DC) basis, the domestically controlled, cross-border and cross-sector (DCCBS) basis, and the cross-border, cross-sector basis for all domestically incorporated deposit takers (CBCSDI). There was also a diversity of consolidation bases used within countries. More than a third of the countries in the CCE used a mix of consolidation bases to compile the core set of FSIs, while about a fifth of countries compiled more than one set of core FSIs using different consolidation bases.

16. Intra-group consolidation adjustments were the norm in the CCE data, but inter-group adjustments were carried out to a much smaller extent.15 The reason is that data used for FSI compilation are by-and-large from supervisory sources, which typically require intragroup consolidation. Meanwhile, most inter-group adjustments have to be carried out by data compiling agencies, requiring additional data to be collected from reporting entities. Partial inter-group adjustments for a given FSI were common: countries often either carried out fully only a subset of the recommended inter-group adjustments and/or they had thresholds which would not permit the adjustment of positions (flows) with some members of the population.16 In the case of FSIs for the sector of deposit takers, only one country made full inter-group adjustments to all its reported FSIs as recommended by the Guide.17

17. The global trend toward adoption of IFRS increased overall consistency between national accounting practices and the Guide. However, in a few areas where the Guide differs from the IFRS,18 it also resulted in some divergence of practices in the compilation of FSIs. More than two thirds of participating countries prepared FSIs for deposit takers that had been compiled in full or partial consistency with IFRS. Regarding the specific areas of divergence between the Guide and IFRS (excluding the matter of consolidation basis), a large proportion of countries opted for the IFRS treatment on any given area. The remainder provided data consistent either with the Guide’s recommendation or with local accounting standards, which differ from both the IFRS and the Guide.

18. The use of national practices in compiling supervisory-based data series, which can differ from the approach of the Basel Committee on Banking Supervision (BCBS) under Basel I, introduced an additional source of diversity across CCE participating countries’ FSIs. Key supervisory-based data series are total regulatory capital, Tier 1 capital, and risk-weighted assets. The Guide defers to national practices in the compilation of such data. Almost all CCE participating countries used the Basel framework to compile these series, but with some deviations from Basel guidance in about half of the participating countries.

19. The diversity of methodologies used in compiling FSIs for the CCE emphasizes the relevance of adequate metadata. Although wide-ranging cross-country comparability of FSIs remains a goal still to be achieved, the metadata compiled now allow comparisons to be made in a more informed way. The extensive metadata collected from CCE participating countries permit building up peer groups of countries which allow for more substantive comparisons.

C. Countries’ Assessment of the CCE

20. Most CCE participating countries found that the CCE met its stated goals and that the exercise proved useful and relevant to their countries’ financial stability work. A survey of CCE countries showed that 90 percent of responding countries believe the CCE met its goal of promoting international comparability of FSIs to various degrees.19 About two thirds of countries said that the CCE met its goal of developing country capacity to compile FSIs. Countries noted that the CCE promoted the compilation and availability of FSIs (including in their own websites and publications), and fostered awareness of FSIs among country authorities. Also, the use of the Guide and the consultations carried out with the staff of the Fund and other countries in the context of the CCE improved the understanding of FSI compilation methodology and fostered expertise in this area among country officials.20 Ninety two percent of countries found the metadata produced for the CCE fully or moderately useful in making transparent the mode of compilation of FSIs across countries and allowing informed cross-country comparisons and approaches to benchmarking. Virtually all countries found the FSI/CCE website created to disseminate the FSI data and metadata useful.

D. Costs of the CCE

21. Among the CCE countries that addressed the issue of the benefits and costs of the CCE in the survey conducted by the staff, most stated that the benefits exceeded the costs.21 Several countries noted that the most significant costs were associated with the development of FSI metadata, which they were preparing for the first time, and which involved coordination among several agencies.

22. STA devoted an average of 6.7 staff years per year during fiscal years 2005–07 to conduct the CCE and related FSI work.22 In addition, information technology costs averaged US$185,000 per year in fiscal years 2006 and 2007. As STA was not assigned additional resources to work on FSIs, the reallocation of resources to conduct the CCE resulted in a slowdown in developmental work and in the delivery of technical assistance and training activities in other areas of monetary and financial statistics. Indicatively, the number of technical assistance missions on traditional monetary and financial statistics areas delivered by the STA division undertaking the FSI work declined by over 40 percent during the conduct of the CCE. Moreover, there has been a slower rate of implementation of standardized report forms (SRFs) for monetary and financial statistics and thus for support to the balance sheet approach. The completion of the implementation of the SRFs by member countries, originally envisaged to have taken place by end-2007,23 has been delayed to end-2009.

III. Experience In The Use Of Fsis In The Fund

23. FSIs have become a standard tool in financial stability analysis. FSIs are now routinely monitored by staff as part of the Fund’s enhanced surveillance of financial systems, and are frequently reported in staff reports and in FSAP reports. Member countries also increasingly use and report FSIs, including in financial stability reports. Although most FSIs are seen as summary measures of the health of financial systems that provide a lagging indicator of developments, they are recognized as representing an essential starting point for stability analysis.

24. At present, FSIs at the Fund are collected from a variety of sources and compiled under different methodologies. In addition to the FSIs compiled as part of the CCE, in their surveillance work Fund staff collect and/or compile a range of FSIs from various external data sources, including national authorities, banks, and commercial data providers. Given the current variety of sources, lack of a uniform methodology, and uneven efforts at maintaining metadata, there have been no assurances of comparability of these FSI data over time and across countries.

A. The Role of FSIs in Financial Sector Surveillance

25. Financial sector surveillance uses a wide variety of analytical tools. Given the complexity of linkages between the financial sector and the real economy, there is no single, widely accepted, detailed method for assessing financial sector stability. In practice, financial stability analysis must rely on a range of available tools and indicators. They include stress testing and sensitivity analysis, various balance sheet-type methods of risk analysis, inferential methods for extracting information from market-based assets and derivative prices, and a variety of indicators and sources of information. Along with quantitative tools, qualitative tools are also typically used—notably the assessments of observance of relevant financial sector standards and codes.

26. FSIs are an essential starting point for financial surveillance.24 They provide a basic measure of a financial system’s exposure to different types of risk and can help gauge the system’s capacity to handle shocks should they occur (its resilience)—notably when used in the context of stress tests.25 However, FSIs are typically viewed as current or lagging indicators of soundness (especially bank-related FSIs), and they are often complemented by an analysis of market-based data, which can provide more forward looking information on expectations and volatility, and at a high frequency (e.g., daily). FSIs for nonfinancial corporations and households have empirically been found to be better leading indicators of vulnerability.

27. Analytical work on FSIs has become part of the broader efforts by the Fund to strengthen its financial surveillance toolkit. Work continues on methodologies for financial surveillance, including on the relative strengths and weaknesses of the available tools and indicators in detecting vulnerabilities to risks, and on how best the tools can be used depending on data availability and country circumstances. In areas related to FSIs, work is ongoing on refining specific aspects of stress testing techniques, including modeling of specific risks, correlations among risks, and channels of contagion among financial institutions, both domestically and cross-border. Another strand of active work focuses on ways to incorporate the balance sheet information underlying FSIs with finance and risk management data, including through applications to financial stability analysis of the Contingent Claims Approach.26

28. As with most indicators, FSIs need to be interpreted with caution. This is because the legal and regulatory systems that produce them—including accounting and prudential definitions, and the broader legal, judicial, information, and governance infrastructure—can vary significantly across countries and can affect both the interpretation and cross-country comparability of the indicators. The CCE has been helpful in bridging these gaps by promoting collection of FSIs and their metadata as well as appropriate methodologies of compilation, but more efforts in this direction are needed.

B. Uses of FSIs

29. The analysis of FSIs has become an integral part of Fund surveillance and MCM and other Fund departments fully support further work by the Fund in FSIs and have explicitly emphasized the importance of the FSI work program to go forward. FSIs are collected and analyzed in the context of FSAP and are increasingly monitored as part of the Article IV process. In addition, FSIs are used as inputs in the interdepartmental VE.

  • FSAP reports routinely include tables on FSIs. These typically cover FSIs for deposit takers from the core set and, to the extent available, the encouraged set; additional indicators for deposit takers may be included when relevant. Coverage of FSIs for non-deposit takers is more limited. As noted earlier, definitions of these indicators vary substantially across countries. FSI data is typically analyzed using peer groups and in the context of stress tests.

  • Inclusion of FSIs in Article IV staff reports has also become more widespread. At least some FSIs are included in staff reports (in a dedicated table and/or as part of the table on financial and external vulnerability). Coverage of FSIs has been generally commensurate with the coverage of financial sector issues more broadly, and has increased in line with the enhanced focus on financial sector issues in Fund surveillance. More regular and broader provision of FSIs in Article IV consultations should be considered, taking into account the resource constraints of member countries and Article IV missions.

  • FSIs are monitored as part of the assessment of underlying vulnerabilities in the VE. The VE methodology combines cross-country analysis of vulnerability indicators and judgment-based assessments for individual countries. For each of the 48 emerging market countries covered, a vulnerability index summarizing information from 20 external, public, financial, and corporate sector indicators allows an examination of cross-country differences and trends in vulnerabilities. FSIs used include three indicators for deposit takers and three indicators for the nonfinancial corporations sector.27

30. In recent years, the Global Financial Stability Report (GFSR) has served as a valuable means to disseminate cross-country FSIs. Since 2003, the GFSR has included a statistical appendix that presents data on key financial indicators, including selected FSIs. The FSI tables cover about 90 countries, with annual time series spanning from 1998 to 2006. As the most extensive published source of cross-country time-series information on FSIs, the GFSR database has received positive feedback from within and outside the Fund, including market participants. However, shortcomings in data accuracy and comparability remain given the variety of sources and methodological approaches (indeed, one concern is that the FSIs in the GFSR do not match those reported as part of the CCE, owing to methodological differences). Efforts to address these shortcomings—including by STA—could allow the GFSR to become an important vehicle for disseminating FSIs.

31. A more concerted effort at FSI data collection and dissemination would also support the increased focus of the Fund membership on financial stability analysis. One of the most visible signs of this increased focus on financial stability analysis has been the number of Financial Stability Reports published by central banks.28 These reports typically use FSIs along with a variety of other tools and indicators—reflecting efforts by member countries to refine the financial analysis toolkit—that are not dissimilar to those at the Fund. Greater availability through the Fund of internationally comparable FSIs would facilitate cross-country analysis by member countries.

32. Internal training supports the analysis and proper interpretation of FSIs by Fund staff. The Fund’s Institute (INS), jointly with MCM, organized three interdepartmental seminars on the use of FSIs in 2006 as part of its internal economics training program. In addition, a number of MCM seminars have focused on particular aspects of FSI analysis. Complementing training provided by MCM on the use of FSIs in surveillance, STA has also been providing training to Fund staff on the compilation of FSIs. In 2006-2007, INS organized jointly with STA three interdepartmental seminars on the compilation of FSIs.

IV. Proposals Going Forward

33. This section discusses options on the way forward with FSIs in the work of the Fund and provides the staff’s proposals on these matters. The options considered and proposals advanced in this section have been formulated taking into consideration the views of CCE participating countries and collaborating international agencies, as well as those of MCM and other Fund departments. In concept, the options are:29

  • (1)    No further work on FSIs by the Fund,

  • (2)    No further work on FSIs by the Fund, but countries encouraged to continue compiling FSIs,

  • (3)    Add FSIs as a data category under the GDDS,30

  • (4)    Add FSIs as a required data category under the SDDS and leave FSI data work to countries, or

  • (5)    Regular FSI data collection and dissemination by the Fund.

34. The widely shared views by the CCE participating countries that the Fund’s participation in FSI data collection would be a catalytic force in facilitating countries’ work on FSIs, the Fund departments’ views on the importance of FSI data for the Fund’s work in financial sector surveillance, and the limitation of countries’ institutional capacity for compiling cross-country comparable FSIs argued against options (1) to (3). Therefore, we expand on options (4) and (5).

35. Regarding inclusion of FSIs as a data requirement under the SDDS, most CCE participating countries did not support this step at this stage. If FSIs were to be included in the SDDS, not all countries that are currently SDDS subscribers would necessarily be able to report FSIs. In addition, if the inclusion of FSIs in the SDDS were to require that countries closely adhere to the Guide’s compilation methodology, many SDDS subscribers that participated in the CCE may not be able to report FSI data given the envisaged persistence of diversity in compilation practices. Even if these issues were addressed, the rate at which new countries would begin to produce data and metadata on FSIs within the framework promoted by the Fund would likely be limited to new SDDS subscribers. If FSIs were deemed essential for Fund surveillance activities, including FSIs in the SDDS would still require creation of a FSI database in the Fund. Inclusion of FSIs in the SDDS would also involve costs related to the preparation of non-CCE countries (that are already subscribing or will in the future subscribe to SDDS),31 monitoring and Data Standards Bulletin Board (DSBB)-maintenance of a new SDDS category, as well as costs related to recurring training and technical assistance.

A. Collection and Dissemination of FSIs via the IMF

36. There is clear value in regular collection and dissemination of FSIs by the Fund. This was indicated by CCE participating countries (Figure 2), which noted that this would be catalytic to the production and use of FSIs (by both CCE participants and others).

Figure 2:
Figure 2:

Countries’ Views on the Way Forward with FSI Work at the Fund

Citation: Policy Papers 2007, 043; 10.5089/9781498333245.007.A001

  • Regular collection and dissemination of FSIs by the Fund through STA would support Fund surveillance. The main benefits for the Fund would be consistency of FSI data used across Fund departments; minimization of errors with a systematic review of the data by Fund staff; and transparency on the construction of data leading to greater clarity on cross-country and cross-time comparability. The database could be used in the context of FSAPs, Article IV consultations, the GFSR, and the VE. There would be advantages to the Fund since centralizing the reporting of data would reduce the burden on MCM and other Fund departments,32 and would yield quality improvements in terms of cross-country comparability, as well as enhanced access to metadata and time series data.33

  • Continuing collection and dissemination of FSIs by the Fund with the Guide as a reference would also lead to greater cross-country comparability of indicators over time, benefiting analytical work as well as bilateral and multilateral surveillance. The goal would be for the database to eventually cover most Fund membership, as new (nonCCE) countries’ FSIs and their metadata are gradually brought into the database.

  • There would be benefits to members of a centralized Fund database, since this could reduce the reporting burden.

  • Finally, the database would be available to member countries, international institutions, and markets, providing them with well-documented data for their cross-country financial sector analyses, leading to greater transparency and market discipline.

37. Regular collection and dissemination of FSIs by the Fund would require setting up and maintaining a system of regular reporting by countries. The FSI database and dissemination system created for the CCE would need to be adapted to collect, store, and disseminate data and metadata on a regular basis. A system of effective collection, validation, updating, and uploading of the reported information would also need to be put in place. STA would work with countries that participated in the CCE to initiate their regular reporting of FSI data and metadata, and would provide technical assistance and training to new (non-CCE) countries with a view to preparing them for regular reporting. The work would also involve amendment of the Guide and development of new FSIs, as discussed below.

38. Collection and dissemination of FSIs on a quarterly or semi-annual basis beginning in mid-2008 would require resources to address recurring as well as one-off costs. One-off costs would be incurred in FY 2008 and are estimated to be around 0.75 staff years.34 These resources would be allocated mainly toward the adaptation of STA’s FSI database as well as the adaptation of the GFSR FSI database and its linking to the STA database. Recurring costs would amount to 5–6.25 staff years initially and would be allocated toward the following:

  • (1) collection, validation, updating, and uploading of FSI data and metadata into the STA database for countries that participated in the CCE: 2.5 staff years initially, with resource needs increasing somewhat over time as new countries initiate reporting to the Fund;

  • (2) maintenance of the GFSR FSI database: 0.25 staff years;

  • (3) preparation of new (non-CCE) countries to produce the FSIs and their metadata: 1.25–2.5 staff years for 10–20 new countries per year;35 and

  • (4) training and technical assistance for CCE and new countries, training of Fund staff, and methodological work on FSIs: 1 staff year

39. On the assumption that STA’s resource envelope going forward is broadly maintained, STA can continue to reallocate from other activities36 the 5 staff years that are currently devoted to FSI work,37 in which case the only additional resources needed would be those pertaining to the one-off adaptation of the FSI database.38 This would be consistent with the initiation of FSI reporting to the Fund by 10 new countries a year, implying that the membership would be reporting to the Fund within about 12 years. If the aim were to achieve the more desirable scenario of initiating FSI reporting by 20 new countries a year, implying membership coverage within about 6 years, an additional 1.25 staff years would be required.

40. If the Fund were to undertake regular collection and dissemination of FSIs, countries should be encouraged, but not required, to report FSIs to the Fund with quarterly periodicity and a one quarter lag. Semiannual and annual periodicities of compilation would be accepted, and the lag in reporting could be extended beyond a quarter.39 Countries would also be encouraged, but not required, to provide to the Fund for dissemination the sectoral financial statements from which the FSIs they compile are sourced, as well as measures of dispersion.40 While countries would continue to be responsible for the accuracy of their data and metadata, the Fund staff (STA) would play a role in their validation and checking, as well as in assessing the comparability of FSI data.

41. STA should provide training and technical assistance to member countries to promote the compilation of FSIs using the recommendations of the Guide. Continued provision of training and technical assistance was supported by virtually all CCE participating countries in the survey conducted. Consistent with the allocation of resources in this area discussed in paragraph 38 (1 staff year), this could include the provision of about 3 training courses (regionally or at Fund headquarters) per year with the same modalities as training in other areas of statistics. Technical assistance could be available to countries either through missions to countries (2–3 per year) or (cross) regional meetings (1–2 per year); visits of country authorities for individual or group consultations with staff at IMF headquarters (3 per year) would be another alternative. Some of these training and technical assistance activities would also serve as fora for exchanging views on FSI compilation and use, but separate seminars/meetings could usefully be organized for such purposes too, if there was the appropriate (re)allocation of resources. The Fund would take the lead in such training and technical assistance, but would also cooperate with other international agencies, as well as with countries, in the provision of training/technical assistance, and specifically regarding facilities, experts, and funding for these purposes. Training of Fund staff on the compilation of FSIs and available databases would be provided on a regular basis.

B. Compilation Methodology

42. Countries found the Guide helpful during the CCE, but in the light of the experience gained in this exercise and in the Fund’s work with FSIs, amendments to the Guide in specific areas were deemed necessary. Main amendments refer to consolidation basis, consolidation adjustments, relation to supervisory standards and accounting standards, and definitions of FSIs. These and other issues related to the compilation methodology are explained in detail in the Background Paper.

43. Consolidation basis.41 The Guide’s recommendation for a domestically controlled cross-border consolidation basis will be amended. The new recommendation would be (i) the cross-border cross-sector consolidation basis for all domestically incorporated deposit takers (CBCSDI) and/or (ii) the domestically controlled cross-border and cross-sector (DCCBS) consolidation basis.42 Both these consolidation bases have cross-border and cross-sector elements to capture the health and soundness of the financial system in the environment of increased globalization and integration of the financial systems. The Guide would recommend that FSIs and their underlying data be provided for both these consolidation bases if the countries consider that there is significant foreign control of domestically incorporated deposit takers.43 The Guide would continue to include a discussion on alternative consolidation bases (such as DCCB, CBDI, DC) for the reference of countries that may still prefer to use such approaches as best suiting their circumstances. The Guide will be amended to encourage countries to compile and disseminate FSIs and underlying data series for the branches of foreign deposit takers operating in their territories as supplementary information, if their presence is deemed by the country as significant.

44. Consolidation adjustments. Intra-group consolidation adjustments will continue to be fully recommended by the Guide. However, consolidation adjustments of inter-group positions and flows will cease to be recommended by the Guide. The Guide will offer a limited discussion of inter-group consolidation adjustments in financial statements for the reference of countries that may still prefer to use them.

45. Relation of FSI methodology with supervisory standards. The Guide will be clarified to explicitly defer to Basel I and Basel II as the standards for compiling supervisory-based underlying data series, acknowledging that there could be national variations in implementation. The Guide will provide a description of Basel II, in addition to the one on Basel I already included.

46. Relation of FSI methodology with accounting standards. The Guide will be amended to defer to the IFRS as the standard on accounting issues.44

47. Contents of the FSI list and definitions of FSIs. For the time being, the FSI list will remain unchanged. Definitions of FSIs and their underlying series will also remain unchanged with a small number of technical modifications that are described in the Background Paper. Over the medium term, work would continue on assessing the need and feasibility of refining the list of FSIs and adding new FSIs.45 Moreover, it will be explored whether reporting countries should be asked to give priority in compiling certain encouraged FSIs, given their importance in Fund surveillance. Any proposal for additions to the FSI list would be based on broad-based consultations with Fund departments, international experts, and would be brought for consideration by the Board.

C. Coordination with Other International Agencies

48. Coordination between the Fund and other international agencies in the areas of harmonizing the methodology of data compilation and reporting should continue.46 In this context, the Reference Group that was created in the context of the CCE should be maintained. The changes in FSI compilation methodology proposed above aim to address issues of harmonization of methodological issues among institutions, and new or remaining issues should be kept under discussion and review. Other international agencies should be encouraged to consider the proposed revised methodology of the Guide and data/metadata reporting modalities in reviewing the reporting of FSIs and their underlying data to them.

V. Issues For Discussion

49. Do Directors agree that the Fund, within the context of its budgetary environment going forward, should accord priority to the collection and dissemination of FSIs, along with underlying data series and related metadata, on a regular basis (paragraphs 36-39)?

50. If so, do Directors have a view on the pace at which additional countries would be brought to reporting FSIs to the Fund (paragraph 39)?

51. Do Directors agree that FSIs should not be included as a required data category in the SDDS at this time (paragraph 35)?

52. If the Fund were to continue to collect and disseminate FSIs, do Directors agree that countries should be encouraged, but not be required, to report FSIs to the Fund with quarterly periodicity and with a one quarter lag (paragraph 40)?


This paper was prepared by a team led by Andreas Georgiou and comprising Xiuzhen Zhao, Agus Firmansyah, and Phousnith Khay under the guidance of Alfredo Leone and Armida San Jose (all STA), in close collaboration with a team led by Marina Moretti and comprising Philip Bartholomew, Nicolas Blancher, Olivier Frecaut, and Michael Moore under the guidance of Christopher Towe (all MCM).


“Country” in this paper also covers nonsovereign entities, for which statistical data are provided internationally on a separate basis.


The work of the Fund in the area of FSIs took into consideration the work of other international and regional institutions in the area of financial data collection and dissemination.


See Concluding Remarks by the Acting Chair, Financial Macroprudential Indicators, Executive Board Meeting 01/64, June 25, 2001 (; and Financial Soundness Indicators (


See IMF Executive Board Discusses Financial Soundness Indicators ( pn/2003/pn0371.htm).


Progress on the Financial Soundness Indicators Work Program ( external/np/sta/fsi/ eng/ 2005/061405.htm).


Sixth Review of the Fund’s Data Standards Initiatives(


The Guide was finalized in early 2004 and was made available on the Fund’s external website. EXR published the Guide in English in March 2006, and subsequently in French, Spanish, and Russian. A translation of the Guide in Chinese is under preparation, and an Arabic language version is envisaged.


Metadata is the supporting information on the data.


More details on the CCE/FSI Reference Group can be found in the Background Paper.


More information on these actions can be found in the Background Paper.


Besides country differences from the Guide in these main areas, there were also other differences, which were either related to the definition of a specific FSI and its underlying series or to some aspect of accounting (not related to IFRS), as explained in the Background Paper.


Consolidation basis refers to institutional coverage—the set of institutions that are included in the population for which a FSI is compiled (see Box 4 of the Background Paper for more explanations).


Both types of adjustments are recommended by the Guide. Intra-group positions and flows are those between a parent and its branches and subsidiaries that are members of the population specified by the consolidation basis chosen. Inter-group consolidation adjustments extend the elimination of positions and flows to associates and unrelated deposit-takers in the population (see Box 5 of the Background Paper for more explanations).


For example, thresholds are often specified in national supervisory regulations as a percent of capital of the investor. Therefore, in carrying out inter-group adjustments, an equity position would not be eliminated between two members of the population if the position represents a smaller share of the capital of the investor than the specified threshold.


See Table 8a. in the Background Paper.


The Guide’s recommendations that currently differ from the IFRS are identified in Box 6 of the Background Paper.


Fifty one out of the 62 CCE participating countries responded to the survey.


Ninety-eight percent of countries, which responded to the survey, viewed the Guide as fully or somewhat useful in developing FSIs during the CCE.


From the twenty countries that responded to the question, 18 felt that the benefits of the CCE exceeded the costs. For those that noted that the costs exceeded the benefits, an important issue was that their national definitions of FSIs often differed in some aspects from the definitions in the Guide.


Of these staff resources, about 0.9 staff years were in the form of staff overtime. Moreover, the staff allocated to CCE work included a special appointee for two of the three fiscal years and a staff member seconded from the European Central Bank for ⅔ of a fiscal year. These staff were not part of the permanent STA staff resources.


As of October 1, 2007, 84 member countries had implemented the SRFs.


For a detailed discussion of how FSIs are used within a framework of financial stability analysis, see Financial Soundness Indicators (


FSIs typically serve as output to stress tests, as the effects of shocks are illustrated as changes in FSIs.


See Gapen and others, 2004, The Contingent Claims Approach to Corporate Vulnerability Analysis: Estimating Default Risk and Economy-wide Risk Transfer, IMF Working Paper No. 04/121. For a recent application to the banking industry, see Indonesia: Selected Issues (


Of the nonfinancial corporations indicators used in the VE, one is included in the encouraged set of FSIs, and two have somewhat different definitions from those in the encouraged set. The data on the nonfinancial corporations in the VE comes from the Corporate Vulnerability Utility (CVU) of RES, which uses Worldscope data.


As of last year, about fifty central banks were publishing such reports, and others were considering it. See M. Čihák, 2006, “How Do Central Banks Write on Financial Stability?,” IMF Working WP/06/163.


Options 3–5 are in principle consistent with the Fund’s continuing work on FSI methodology, including on amendments to the Guide and development of new FSIs.


The GDDS does not involve any data dissemination.


Fifty two out of the 64 SDDS countries have completed FSI data/metadata submission under the CCE.


The STA FSI database would be linked to all other FSI databases across the Fund, being the sole source of data and metadata for indicators in the core and encouraged FSI lists.


In the transitional period until the point when the STA FSI database is operational and has adequate time series data for surveillance or research purposes on a sufficiently broad set of countries, the existing databases would continue to be maintained. The transitional period could be minimized with a concerted effort of Fund departments and member countries.


Additional US$ 30,000 would need to be allocated toward related IT work.


At the rate of 20 new countries per year, member countries would be brought to reporting regularly to the Fund within about six years.


The opportunity cost will be the continuing slow-down in the delivery of technical assistance in traditional monetary and financial statistics areas and in the implementation of the standardized report forms (SRFs) across the Fund membership noted in paragraph 22 above. These effects could be alleviated somewhat through greater use of experts to deliver technical assistance if external funding were available, but would entail backstopping costs on staff. Backstopping costs per year to alleviate the compression on technical assistance are estimated at about 0.3 staff years.


The estimate assumes that the resources corresponding to the existing special appointee devoted to STA’s FSI work program will continue to be made available to STA.


A smaller than indicated allocation of financial resources to IT work would delay the implementation of regular collection and dissemination of FSIs beyond mid-2008.


According to the survey conducted, about 62 percent of CCE participating countries would be able and willing to submit quarterly FSIs to the Fund. However, only 32 percent of the countries felt that all FSIs should be compiled quarterly, while 58 percent said that some FSIs should be compiled with different periodicity. Sixty four percent of the countries felt that the periodicity of FSIs submitted to the Fund should be flexible. Regarding timeliness, 66 percent of countries felt they will be able and willing to submit FSIs to the IMF with a one quarter lag, while 54 percent felt that all FSIs should be disseminated with the same timeliness.


The vast majority of CCE participating countries (89 percent) indicated that it should be required or encouraged to also report to the Fund sectoral financial statements (from which FSIs are sourced) for dissemination. A majority of these countries (62 percent) also said that measures of dispersion (such as measures of variability, skewness, and kurtosis) could be usefully encouraged to be reported for some of the FSIs to make them more useful for financial soundness analysis.


See Box 4 of the Background Paper for explanations of the consolidation bases mentioned in this paragraph.


Cross-sector coverage would be limited to exclude subsidiaries of deposit takers that are insurance companies and nonfinancial corporations.


FSIs would be usefully compiled using both consolidation bases when a country deems that the share of foreign controlled institutions in the sector is not so large so as to render compilation of FSIs for the domestically controlled part of the sector of little value.


With regard to the consolidation basis, IFRS guidance would be followed at the level of the financial sector (excluding insurance companies).


New FSIs could potentially be added for major other financial corporations subsectors, for nonfinancial corporations and households in light of the need to provide leading indicators of financial soundness, as well as for deposit takers and financial markets. In addition, further work could involve assessing qualitative indicators (such as disclosure and transparency of the financial and nonfinancial corporations subsectors) as possible additional measures of soundness of the financial system.


The Islamic Financial Services Board has recently adopted the Compilation Guide on Prudential and Structural Islamic Finance Indicators (Islamic Guide), which drew to a large extent from the IMF’s Guide. A pilot exercise to compile prudential and structural Islamic finance indicators is also underway.

Financial Soundness Indicators-Experience with the Coordinated Compilation Exercise and Next Steps
Author: International Monetary Fund