Financial Soundness Indicators-Experience with the Coordinated Compilation Exercise and Next Steps-Background Paper

This is the companion background paper to the staff report on Financial Soundness Indicators (FSIs): Experience with the Coordinated Compilation Exercise and Next Steps. It provides detailed information on the modalities of the Coordinated Compilation Exercise (CCE), the experience with the CCE, the issues that arose in that exercise regarding the compilation methodology in the Financial Soundness Indicators: Compilation Guide (Guide), and the matters that were taken into account in considering the specific amendments to the Guide, presented in the staff report.

Abstract

This is the companion background paper to the staff report on Financial Soundness Indicators (FSIs): Experience with the Coordinated Compilation Exercise and Next Steps. It provides detailed information on the modalities of the Coordinated Compilation Exercise (CCE), the experience with the CCE, the issues that arose in that exercise regarding the compilation methodology in the Financial Soundness Indicators: Compilation Guide (Guide), and the matters that were taken into account in considering the specific amendments to the Guide, presented in the staff report.

I. Introduction1 2

1. This is the companion background paper to the staff report on Financial Soundness Indicators (FSIs): Experience with the Coordinated Compilation Exercise and Next Steps.

It provides detailed information on the modalities of the Coordinated Compilation Exercise (CCE), the experience with the CCE, the issues that arose in that exercise regarding the compilation methodology in the Financial Soundness Indicators: Compilation Guide (Guide), and the matters that were taken into account in considering the specific amendments to the Guide, presented in the staff report.

II. Modalities of the CCE

2. Countries participating in the CCE (Box 1) 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 countries were also encouraged to compile some or all of the 28 encouraged FSIs and provide them along with their metadata.1 The terms of reference (ToR) of the CCE included a timetable of key actions for the Fund and for the countries participating in the CCE (Box 2). It was agreed that all FSI data and metadata provided to the Fund would be disseminated to the public through the Fund's website.

CCE Participating Countries

article image

Timetable of Key Actions in the CCE

article image

3. To ensure full collaboration with other international and regional institutions in the area of FSI work, the Statistics Department (STA) created a CCE/FSI Reference Group.

The Reference Group is composed of 17 international and regional agencies (Box 3). In the course of the CCE, three meetings were held with representatives of the Reference Group, who also joined the plenary meetings of CCE countries in 2004 and 2007. The Reference Group meetings provided a forum for the exchange of information on the work of various institutions in the area of FSIs and in related data areas. The meetings provided opportunities for Fund staff to present to the Reference Group issues about the conduct of the CCE, the methodology for compiling FSIs, and the way forward with FSIs, and to receive feedback on these matters. Finally, the meetings of the Reference Group provided opportunities to identify areas where harmonization of data compilation and reporting methodologies is needed to minimize duplication of effort on the part of countries and institutions.

Members of the CCE/FSI Reference Group

article image

4. To facilitate the implementation of the CCE work program, STA provided technical assistance to countries in the compilation methodology recommended in the Guide. This assistance took the form of (i) provision of technical advice from headquarters, in particular during the staff review of three submissions of countries' draft metadata, and the submission of countries' final data and metadata; (ii) two rounds of (cross-) regional meetings to discuss issues in the implementation of the CCE, address cross-cutting issues in the compilation of FSIs and the preparation of metadata, and review with countries their draft FSI metadata and preliminary FSI data; (iii) bilateral consultations with countries that visited STA for additional technical assistance in their work in the CCE;2 and (iv) a mission for technical assistance and training.3

5. To ensure uniform and consistent collection and reporting of data and metadata, STA in consultation with CCE countries and the Monetary and Capital Markets Department (MCM) developed standard data and metadata report forms. These forms included a data report form (DRF), a metadata template (MT), and a metadata questionnaire (MQ). The DRF facilitated the presentation of FSI ratios compiled and the underlying data series used in their compilation, the review of the reported FSI ratios, and the processing and tabulation of the FSIs for meeting various data needs. The MT and MQ aimed at organizing the information on the national practices that govern the compilation of FSIs. The MT was used to describe succinctly the deviations of the country practice from the Guide. The MQ provided a fuller and more comprehensive description of national practices and deviations from the Guide.

6. STA with the assistance of the Technology and General Services Department (TGS) created a set of web pages on the Fund's external website to disseminate the information collected in the CCE. They include a CCE home page, a FSI data page, and a FSI metadata page.4 The CCE home page provides information on the FSIs and the CCE. The FSI data page provides access to the FSI data and underlying data reported by countries. Information can be accessed by country participating in the CCE. Links take the user to the specific metadata of a given FSI or underlying series. The FSI metadata page provides access to the metadata submitted by countries—the MT and the MQ. Information on metadata can be searched by country, by questions, and by text search within the MQ for one or more countries, as well as by type of information within the MT for one or more countries. Moreover, links take the user from metadata to the data, from the MT to the MQ, and from both the MQ and the MT to the relevant passages in the Guide. These facilities are provided to assist data users to obtain and compare information across countries effectively, and facilitate the retrieval of metadata and their comparison with the recommendations of the Guide.

7. STA with TGS also made available FSI data and metadata to Fund internal users through the Economic Data Sharing System (EDSS). Apart from the provision of the standardized FSI tables for each country, internal Fund users were given the facility to sort the FSI data, as well as the underlying data, by nine analytically significant dimensions (e.g., consolidation basis, presence of consolidation adjustments, use of accrual accounting, etc.), thus promoting the identification of truly comparable data.

III. Experience with the CCE

8. Following the CCE work program, 57 out of the 62 CCE participating countries have submitted their final FSI data and metadata so far. FSI data for individual countries are included in Tables 2, 3, and 4, arranged in various potential peer groups.5

9. FSIs for deposit takers (DTs) were the most commonly compiled indicators, reflecting a number of factors. All countries compiled the core set of FSIs (Table 1), which all pertain to the sector of DTs, as compilation of the core set was required under the terms of reference of the CCE.6 The DTs sector was also the sector for which encouraged FSIs were most commonly compiled by countries (84 percent), confirming the view of the Guide that this sector remains central to financial soundness analysis, and is favored by a greater availability of data. FSIs for the other financial corporations (OFCs) sector were the second most compiled FSIs, followed by the household sector and the sector of nonfinancial corporations (NFCs). The relatively small number of countries (35 percent) compiling FSIs for NFCs reflected to a large extent the difficulties experienced by many countries in collecting aggregate data for the NFC sector.

10. Regarding the source of data, supervisory sources dominated the compilation of FSIs for DTs with well over 80 percent of FSIs compiled using such data (Table 5.a). This reflected the fact that FSIs are seen (and used) by countries by-and-large as an extension of microprudential ratios compiled for individual deposit takers (groups) for supervisory purposes. For other sectors, the data sources are more diverse, confirming a lower degree of data development for these sectors. The number of submissions for any given FSI may exceed the number of countries reporting FSIs in the CCE (Table 1) because some countries saw the need for different coverages for certain indicators.

11. A wide range of methodologies were used in compiling FSIs for the CCE. This was due to (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 areas of methodological differences from the Guide are on the consolidation basis, inter-group consolidation adjustments, and on a few accounting practices under the International Financial Reporting Standards (IFRS).

12. There was wide diversity of consolidation bases in data reported for deposit takers in the CCE. Domestically controlled, cross-border (DCCB) consolidation basis is the approach recommended by the Guide for compiling FSIs for DTs (Box 4). However, as shown in Table 6.a, DCCB was not the consolidation basis most commonly used to compile the core set of FSIs for DTs—only 9 percent of the reported core FSIs were compiled on this basis. More commonly used than the DCCB were the domestic consolidation (DC) basis (39 percent), the domestically controlled, cross-border and cross-sector (DCCBS) basis (26 percent), the cross-border, cross-sector basis for all domestically incorporated DTs (CBCSDI) (15 percent), and other bases (11 percent). A similar result emerges from an inspection of the consolidation bases used, arranged by country (Table 7). The additional element that emerges from this view is that there is a diversity of consolidation bases used not only across countries but also 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 19 percent of countries compiled more than one set of core FSIs using different consolidation bases.

Consolidation Bases

Consolidation basis refers to institutional coverage, i.e., the totality of entities that are defined to constitute a population of institutions (a sector) for which FSIs are to be compiled. There are different consolidation bases, and thus different populations that can be defined. The data of the institutions that belong to the defined population are the basis on which aggregation and any consolidation adjustments can be applied to produce the population-level (sector-level) data from which FSIs are compiled. It should be noted that the term consolidation basis does not imply that certain consolidation adjustments have necessarily been applied to the aggregated data. The consolidation bases used by countries in the CCE are the following:

Domestically controlled, cross-border (DCCB) consolidation basis includes the data of domestically controlled and incorporated entities with their branches (domestic and foreign) and all their subsidiaries (domestic and foreign) that are classified in the same sector.

Domestic consolidation (DC) basis includes the data of resident entities along with those of their branches and subsidiaries (if any) in the same sector that are resident in the domestic economy.

Domestically controlled, cross-border and cross-sector (DCCBS) consolidation basis is similar to DCCB, but has a broader coverage in that the data also include those of subsidiaries in other sectors (such as securities dealers and investment funds).

Cross-border consolidation basis for all domestically incorporated entities (CBDI) covers both domestically controlled entities and the local subsidiaries of foreign entities in the sector, along with their branches and subsidiaries in the same sector, which can be either domestic or foreign residents.

Cross-border and cross-sector consolidation basis for all domestically incorporated entities (CBCSDI) has a broader coverage than CBDI as the data also include those of subsidiaries of the reporting groups in other sectors.

Other consolidation basis refers to institutional coverages other than those mentioned above.

13. Intra-group consolidation adjustments were the norm in the CCE data. The Guide recommends that both intra-group and inter-group consolidation adjustments are carried out in compiling FSIs (see Box 5 for a brief description of consolidations adjustments). Regarding intra-group consolidation adjustments, there was wide ranging consistency with the Guide, as intra-group consolidation adjustments were fully carried out in 71 percent of core FSIs reported, and were partially carried out in 17 percent of core FSIs (Table 8.b). The reason is that these data are by-and-large from supervisory sources, which usually require that such adjustments be carried out in the data by reporting entities (groups).

Consolidation Adjustments

Intra-group consolidation adjustments involve consolidation of all flows and positions among units within each group in the population. A group may consist of more than one institutional unit and in the Guide it is defined to include a parent, its subsidiaries, and its branches. It should be noted that if a cross-border consolidation basis is adopted (such as DCCB, DCCBS, or CBCSDI), these institutional units may be domestically based or foreign based. Moreover, if a cross-sector consolidation basis is adopted (such as DCCBS or CBCSDI), these institutional units may belong to the same or different economic sectors. Therefore, the composition of groups would vary if the consolidation basis changes, and thus the intra-group consolidation adjustments that are applicable would vary in magnitude.

Inter-group consolidation adjustments involve the elimination of some (but not all) flows and positions among groups. This amounts to the elimination of positions and flows between a parent and its associates (i.e., institutions on which the parent has some control, and which are not subsidiaries) as well as between one group and other (less-related or unrelated) entities in the population. However, inter-group interest flows are not eliminated. Moreover, inter-group positions exempted from consolidation are positions among groups in debt and financial derivatives. This is recommended in order to enhance monitoring of risk exposures and the potential for contagion.

For the income and expense statement, there are seven inter-group adjustments, which relate to the elimination of (1) provisions for accrued interest on inter-group loans, (2) inter-group fees and commissions, (3) inter-group dividends receivable and payable, (4) an investing group's pro-rated share of the earnings of an associate from the same reporting population, (5) other inter-group income, (6) gains and losses on ownership of shares and equity of other groups in the reporting population, and (7) specific provisions on loans to other groups in the reporting population.

For the balance sheet, there are three inter-group adjustments, which relate to the elimination of (1) the investing group's pro-rated share of capital and reserves of an associate in the reporting population, (2) the market value of the shares and other equity investments in other groups in the reporting population, and (3) specific provisions on loans to other groups in the reporting population.

14. Inter-group consolidation adjustments were carried out to a much smaller extent and most often in a partial manner. About 34 percent of the submitted core FSIs were compiled using full (10 percent) or partial inter-group adjustments, while 42 percent were compiled without any inter-group adjustments; inter-group adjustments were not applicable to 24 percent of the reported core FSIs (Table 8.b). The incidence of inter-group consolidation adjustments was significantly increased by the fact that they were de facto carried out (in part or in full) on account of supervisory deductions from the FSI-underlying data series for total regulatory capital, Tier 1 capital, and risk-weighted assets. However, this often also led to differential application, across different FSIs of a given country, of the adjustments related to equity positions in the population. If one were to abstract from the effect of the supervisory-based series, any given inter-group adjustment of balance sheet data would be found to have taken place in 23-25 percent of countries, and for income and expense statement data in 9-14 percent of countries (Table 9.a). Partial inter-group adjustments for a given FSI were common: countries often either carried out fully only a subset of the recommended adjustments (Table 9.b) and/or they applied thresholds which would not permit the adjustment of positions (flows) with some members of the population.7 In the event, only three countries made in full all recommended inter-group adjustments to balance sheet and income and expense statement data used in their reported FSIs.

15. The ongoing convergence towards adoption of IFRS increased overall consistency between national accounting practices and the Guide. However, in a few areas in which the Guide departs from IFRS it also resulted in the divergence of practices of the majority of CCE participating countries from the Guide. The Guide's recommendations that differ from the IFRS are identified in Box 6. As of the end-2005 reference date of the CCE, 75 percent of CCE participating countries prepared FSIs for DTs that had been compiled in full (32 percent) or partial consistency with IFRS (Table 10). Regarding the specific areas of divergence between the Guide and IFRS (abstracting from the matter of consolidation basis), 44-70 percent of countries opted for the IFRS treatment on any given area (Table 10). The remainder provided data consistent either with the Guide's recommendations on any given area (5-40 percent of countries) or with local accounting standards, which differ from both the IFRS and the Guide.

Main Differences between the Guide and IFRS

Consolidation basis: IFRS, according to International Accounting Standards (IAS) 27, require that financial statements be prepared for a group that includes the parent company and all of its subsidiaries irrespective of the sector they belong to. Subsidiaries are entities controlled by a parent company. The Guide recommends that a domestically controlled cross-border (DCCB) consolidation basis be used, where the consolidation includes only deposit takers (see Box 4).

Principle for recognizing assets and liabilities: The International Accounting Standards Board (IASB) framework determines whether assets and liabilities exist and are outstanding on the basis of criteria regarding control and the probability of future economic benefits that can be measured reliably. Moreover, IAS 39 notes that an entity shall recognize a financial asset or liability on its balance sheet when the entity becomes a party to the contractual provisions of the instrument. IAS 39 also states that a regular way purchase or sale of financial assets shall be recognized and derecognized, as applicable, using trade date accounting or settlement date accounting. The Guide states instead that whether assets and liabilities exist and are outstanding is determined at any moment in time by the principle of ownership.

Fees and commissions receivable/payable: The Guide differs from IAS 18 in that it does not adjust the effective yield of a financial instrument for fees that are an integral part of the effective yield of an instrument, but records them under fees and commissions receivable and payable.

Gains and losses on available-for-sale assets: Under IAS 39, they are recognized in equity except in the case of impairment losses, foreign exchange gains and losses, and when the asset is derecognized. The Guide recommends instead that gains and losses on available-for-sale assets always be included in the income statement.

Gains and losses on cash flow hedges, hedges of a net investment in a foreign operation, and (potentially) hedges for the foreign currency risk of a firm commitment: According to IAS 39, the ineffective portion of gains and losses on the above hedging instruments is recorded in the income statement. Meanwhile, the effective portion of the gains or losses on a cash flow hedge or hedge of a net investment in a foreign operation, and potentially on a hedge for the foreign currency risk of a firm commitment, is accounted directly in equity and is recycled to the income statement only when the hedged transaction affects profit or loss. This differs from the Guide, which recommends that all gains and losses on hedging instruments always be included in the income statement.

Valuation of held-to-maturity instruments: Under IAS 39, held-to-maturity instruments (that are otherwise tradable) are to be valued at amortized cost. In the Guide it is recommended that they are market or fair valued if they are tradable.

Valuation of loans and receivables: Under IAS 39, loans and receivables are to be valued at amortized cost. In the Guide it is recommended that nominal value is used instead, as for all nontradables.

16. Besides country differences from the Guide on consolidation basis, consolidation adjustments, and IFRS-related accounting, there were also other differences. These differences were either related to the definition of a specific FSI and its underlying series or to some aspect of accounting. Regarding differences from the Guide in the definition of data series, some of the most common were those related to the series underlying liquidity indicators (I10 and in the core set: 51 percent of countries defined liquid assets to include inter-bank claims and 58 percent of countries defined short-term liabilities to include inter-bank liabilities. Differences also arose often regarding the definition of loans—59 percent of countries deviated from the Guide by including tradable loans and trade credits and/or excluding repos/financial leasing and inter-bank loans. The definition of the encouraged FSI on large exposures (I14) was also an area where there were significant differences from the Guide, and several countries opted to redefine the indicator as total value of large exposures to capital in place of the available options offered in the Guide. Other differences with the Guide, included the exchange rate used for conversion into the unit of account, the definition of foreign currency instruments, the treatment of potential costs, and the specification of the asset boundary (see Tables 11a, 11b, 11c, and 11d).

17. The use of national practices in compiling supervisory-based series, which can differ from the Basel Committee on Banking Supervision (BCBS) approach under Basel I,8introduced another source of variation across country practices in compiling FSIs. The Guide defers to national practices in compiling such data series, which comprise total regulatory capital, Tier 1 capital, and risk-weighted assets.9 Almost all CCE countries (98 percent) used the Basel framework to compile these series, but only 54 percent followed BCBS guidance in all areas (Table 12). Almost all of the remaining countries (44 percent) differed from the framework in one or more of the following areas: use of Tier 3 capital, treatment of goodwill, and treatment of certain innovative instruments.10 The remaining two percent of participating countries followed national standards.

18. The diversity of methodologies used in compiling FSIs for the CCE indicates that wide-ranging cross-country comparability of FSIs remains a goal still to be achieved, but comparisons among countries can now be made more reliably. The extensive metadata on CCE countries' data provide crucial information on the establishment of suitable peer groups within which comparisons could be made. Different peer groups can be constructed by specifying an increasing number of methodological areas where countries have the same practice. Given the current level of diversity of compilation practices across countries, specifying a larger number of methodological areas in which countries follow the same approach would lead to peer groups with very few members. The analyst therefore has to exercise judgment in striking an appropriate balance between homogeneity of peer groups and adequate number of countries in these groups. An important input into such an assessment is the potential effect of a methodological variation on the value of a FSI.11 Examples of possible peer-groups constructed from the CCE data (Tables 2, 3, and 4) show how a continuum of peer groups can be created by using countries' FSIs whose compilation is the same in an increasing number of methodological dimensions.

IV. Methodological Issues in Compiling FSIs

19. Countries found the Guide helpful during the CCE,1214 but in the light of the experience gained in this exercise and in the Fund's surveillance and analytical work with FSIs in recent years, amendments to the Guide in specific areas are deemed necessary. The issues and perspectives that were taken into consideration to ascertain the needed amendments, which are summarized in section IV.B of the staff report, are described in detail below.

20. The views of CCE participating countries and collaborating international institutions were major inputs in considering possible amendments in FSI methodology.

Countries, along with collaborating international and regional institutions, expressed their views during a meeting held at IMF headquarters during May 30-31, 2007 on the occasion of the completion of the CCE. The staff also sent out a survey to methodically collect the views of all CCE participating countries so as to reflect their views in this paper. In the event, 51 out of the 62 countries participating in the CCE returned the survey to STA with their answers.1315 The results of the survey are provided in Table 13.

A. Consolidation Basis

21. A major issue that arose in the CCE was the appropriate consolidation basis in compiling FSIs for deposit takers.16 In the view of most of the CCE participating countries (86 percent) that responded to the survey, the Guide's recommendation for a domestically controlled cross-border consolidation basis should be amended (Table 13). Most responding countries (86 percent) also expressed a preference for the Guide not to provide a singular recommendation on consolidation basis; virtually all of those countries preferred a limited set of options to be offered by the Guide. From the combination of options offered by countries, no clearly preferred combination emerged. However, certain options for consolidation basis were clearly more often included than others in the combinations proposed by countries (Table 14); these were the DCCBS, DC, and CBCSDI bases, in order of prevalence. When countries were asked which single consolidation basis was best suited for assessing the health of a country's financial system, CBCSDI, DCCBS, and DC were noted in that order of preference (Table 13, section II, question 1).

22. The issue of whether the Guide should amend its current recommendation on consolidation basis ought to be informed by countries' needs in financial soundness analysis. Currently, there is clearly a diversity of needs across countries in their financial soundness analyses. These needs depend on country circumstances regarding (i) the degree of cross-sectoral ownership between deposit takers and other financial corporations, (ii) the prevalence of foreign control of deposit takers operating domestically, (iii) the modalities of the presence of foreign controlled institutions in the country (i.e., subsidiaries as opposed to branches), and (iv) the extent of cross-border operations of domestically incorporated deposit takers. For example, countries would find most useful DCCBS as a consolidation basis when they have a high degree of cross-sectoral ownership, foreign controlled institutions are limited in importance, and cross-border operations are significant. If, in addition, there is also a significant degree of foreign control of domestically incorporated institutions, then CBCSDI would be an appropriate choice (potentially with DCCBS as a supplementary option to provide a view of the part of the sector primarily supervised by national supervisory authorities). Finally, in case there is little cross-sectoral ownership, significant foreign presence—including in the form of branches of foreign deposit takers—and limited cross-border operations, then the DC basis would meet the needs of financial soundness analysis.

23. In amending the recommendation of the Guide on consolidation basis it is important to consider the need for cross-country comparable data. CCE countries have emphasized during the CCE plenary meeting and elsewhere the importance of generating cross-country comparable indicators. Cross-country comparable data would help enrich financial soundness analysis in a given country as well as across countries, by providing large numbers of countries as true comparators.17 At the same time, however, such country-comparable data should be able to capture emerging trends in the financial system, such as increasing cross-sector ownership, foreign control of institutions operating domestically, and cross-border operations. If the Guide were to recommend a very limited number of consolidation bases that addressed those issues, it would serve to move towards clustering of FSI compilation approaches of countries around those few options over the medium term. In these circumstances, even if not all countries' FSIs would be comparable, there would be comparability among the FSIs of significant numbers of countries that form a given cluster.

Implications for FSIs from Varying the Consolidation Basis

The FSI data collected for the CCE provide a rough indication of the implications of different consolidation bases on FSI values because several countries compiled two or more sets of FSIs using a different consolidation basis for each set. The data presented below show the effect on FSI ratios from switching from a given consolidation basis to another, averaged over the—admittedly small—number of countries that provided data on each pair of consolidation bases. A caveat is that while in some of these country cases the consolidation basis is the only variable that is changing, in others there are some other variables (such as accounting standards) that may be changing too.

Core Set: Quantitative Implications of the Use of Different Consolidation Bases

(in percent)

article image

Average of the absolute value of differences.

The data show that the changes caused to FSI ratios by switching consolidation bases can be up to several percentage points. Therefore, consolidation basis appears to be an important methodological dimension that has to be taken into account in identifying which FSIs (within or across countries) are comparable.

24. Given the considerations above, there are two main options in amending the recommendation of the Guide on consolidation basis:

  • One recommendation would be to choose CBCSDI and/or DCCBS as consolidation bases as these two bases appear suitable to capture most potential risks within the group. The Guide would recommend that FSIs and their underlying data be provided for both these consolidation bases if the country authorities assess that there is significant but not nearly complete foreign control of domestically incorporated deposit takers. The Guide would continue to include a discussion on alternative consolidation bases (such as DC, DCCB, CBDI) for the reference of countries that may still prefer to use such approaches as best suiting their circumstances.18

  • An alternative recommendation would be to offer CBCSDI and DCCBS as choices, but also explicitly add DC as a fully recommended alternative consolidation basis; other elements of the first approach would be maintained.

25. Regarding the two alternative recommendations above, a number of issues could be highlighted:

  • The first recommendation would be consistent with the views of those CCE participating countries that have argued that the Guide should recommend only consolidation bases that a priori satisfy the cross-sector and cross-border institutional coverages of the Basel capital adequacy framework. The second recommendation would be consistent with the views of countries that argued that the set of options should be determined by the frequency with which consolidation bases were actually used in the CCE (Table 6).

  • If the second recommendation were to be adopted, there would likely be greater diversity of compilation methodologies over the long-term. This would decrease somewhat the comparability of FSIs across countries.

  • The DC basis in the second recommendation could offer a more tailored view of some countries' financial sectors, at least for a while.

  • The second recommendation would lead in the short-run to a lower reporting cost of FSIs for countries currently using the DC basis, as data availability considerations were the single most important factor for countries in choosing their consolidation basis for reporting FSIs in the CCE (Figure 1).19 However, even under the first recommendation, countries would not be required to abandon the DC basis, but would be able to compile FSIs using any basis they wished (with supporting metadata).

Figure 1:
Figure 1:

CCE-Rationale for Countries in Choosing Consolidation Bases for Reporting FSIs

Citation: Policy Papers 2007, 042; 10.5089/9781498333252.007.A001

  • Under the second recommendation, there would probably be more switches in country practices from DC to CBCSDI over the medium- and long-term (instead of in the short-term) as developments in countries regarding cross-sector ownership and cross-border operations eventually render the DC basis less useful. This would lead to breaks in the FSI time series (along with deferred increased reporting costs) for these countries and, consequently, difficulties in analysis.

26. Taking into consideration the matters above, the following amendments to the Guide were deemed appropriate: The Guide's current recommendation for a domestically controlled cross-border consolidation basis will be amended. The new recommendation will 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. The Guide will 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.20 The Guide will 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. Metadata and data report forms for reporting to the IMF will be amended accordingly. 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 the presence of such branches is deemed by the country as significant.

B. Inter-Group Consolidation Adjustments

27. Another major issue that arose in the CCE was the relevance of inter-group consolidation adjustments for data used to compile FSIs. As discussed in paragraph 14, there was some use of inter-group consolidation adjustments in compiling data for the CCE, mostly for certain inter-group positions (and to a much lesser extent for inter-group flows) (Tables 8 and 9). Supervisory-based concepts of capital were the data series that showed the highest incidence of adjustments,21 as the latter were de facto carried out when supervisory deductions were made. About 62 percent of the countries responding to the survey expressed the view that inter-group adjustments should not be recommended by the Guide, citing analytical, data availability, quantitative significance, and cost considerations (Table 13). The remaining 38 percent of responding countries were of the view to maintain the recommendation of the Guide on the matter, and expressed a preference for fully inter-group consolidated data as opposed to data that had been consolidated just at the level of reporting groups.

28. Regarding whether the Guide should amend its current recommendation on inter-group consolidation adjustments, the following issues could be highlighted:

  • Inter-group adjustments were recommended by the Guide to eliminate the double-counting of income and capital in the sector. They are essential in measuring the flows of resources between sectors as well as the equity positions among sectors, but they may not always be useful for the analysis of financial soundness for reasons indicated below.

  • Inter-group consolidation adjustments render the derived FSIs different from simple weighted averages of reporting group data, and thus FSIs cannot be simply disaggregated to the indicators of individual reporting groups. Therefore, interpretation of FSIs with inter-group consolidation adjustments may become more complex.

  • Certain inter-group adjustments could be justified for the same reasons that supervisory deductions for total regulatory capital or Tier 1 capital can include eliminations of positions between institutions where there is a greater-than-specified degree of control: the resources of an institution could potentially be used to buttress those of another one if it was a related entity.22 If such a rationale of inter-group adjustments were to be accepted, adjustments would be called for positions or flows among institutions for which supervisors feel there is a high enough degree of relatedness, and not for all institutions in the population (sector) as currently recommended in the Guide.

  • The quantitative effect of inter-group adjustments on the values of FSIs is difficult to determine a priori and is bound to be country specific. Most CCE countries that responded to the survey viewed the effect as not significant (43 percent on flows and 41 percent on positions), but 24 percent of countries rated the effect as significant or somewhat significant (Table 13). About one third of responding countries indicated the significance of the effect was unknown. There are indications that in the countries that actually carried out the adjustments in the CCE, the effect was relatively limited but not necessarily negligible (see Box 8).

Quantitative Effect of Inter-Group Consolidation Adjustments on FSIs

The FSI data collected for the CCE provide an indication of the effect of certain inter-group consolidation adjustments on the values of FSIs. This is possible because several countries reported, as part of the memorandum items in their sectoral financial statements, the magnitudes of adjustments on account of deposit takers' shares and other equity investments in associates and other deposit takers in the population. These figures correspond to two of the three adjustments recommended by the Guide for stock data such as capital and total assets. With data on these adjustments, it is possible to generate two sets of FSI values (with adjustments and without adjustments, respectively) and to compare the magnitudes of these two sets of FSIs. The results are presented in the summary table below:

Core Set: Quantitative Implications of the Inter-Group Consolidation Adjustments for Equity Investments in Associates and Other Deposit Takers Resident in the Economy

(in percent)

article image

Average of the absolute value of differences.

The data show that the changes caused to FSI ratios by carrying out these specific adjustments have amounted by-and-large to fractions of percentage points for the countries for which data were available. For some countries, therefore, inter-group adjustments may be of limited quantitative significance. However, this does not preclude the possibility that for other countries the quantitative effect of the adjustments could be significant.

  • The data requirements for carrying out inter-group adjustments could be burdensome and costly, especially for countries with many institutions, whose data would be subject to such consolidations. The Guide currently notes that judgment would need to be exercised on the benefits and costs of collecting new data to carry out inter-group adjustments.

  • If the Guide were to recommend against inter-group adjustments, there would still be inter-group adjustments de facto being carried out for many countries' supervisory-based series used to compile several FSIs for deposit takers. Reversing such inter-group adjustments would also be costly and burdensome.

  • Cross country comparability of FSIs could decrease if quantitatively significant inter-group adjustments are made in the data by some countries and not by others. Comparability may be further compromised if countries do not carry out all consolidation adjustments that may be significant, but only some of them.

29. Given the considerations above, there are two main options in addressing the recommendation of the Guide on consolidation adjustments:

  • One recommendation would advise against all inter-group consolidation adjustments. Countries whose supervisory-based series are de facto inter-group consolidated would still be able to use these series to compile FSIs, but the presence and degree of the existing adjustments would be documented in the metadata as they could affect cross-country comparability of FSIs.

  • An alternative recommendation would have the Guide cease to recommend inter-group adjustments. The Guide would offer a limited discussion of inter-group consolidation adjustments in financial statements for the reference of countries that may still prefer to use them. Moreover, de facto inter-group consolidation adjustments on account of implementation of supervisory deductions would not be discouraged by the Guide. The presence and degree of any adjustments (explicitly or otherwise carried out) in a country's data would be documented in the metadata as they could affect cross-country comparability of FSIs.

30. Taking into consideration the matters above, the following amendments to the Guide were deemed appropriate: 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. Moreover, de facto inter-group consolidation adjustments on account of implementation of supervisory deductions will not be discouraged by the Guide. Metadata forms for potential reporting to the IMF would be amended accordingly.

C. Relation of the Guide with BCBS Guidance

31. An issue that arose in the CCE is that the Guide relies on national practice in compiling supervisory-based series and does not explicitly defer to BCBS standards. The large majority (85 percent) of CCE participating counties expressed the view that the Guide should be amended to explicitly recommend that the approaches set out in Basel I and Basel II be followed in compiling supervisory-based underlying data series used to compile FSIs

(Table 13).23 Moreover, about 63 percent of countries responding to the survey preferred that the Guide also explicitly follow BCBS on the matter of consolidation basis, which would favor the DCCBS and CBCSDI consolidation bases. The issues related to this latter matter were addressed above.

32. If the Guide were to be clarified to explicitly accept BCBS guidance in compiling supervisory-based series, the following issues would need to be taken into consideration:

  • The existence of national variations in the implementation of BCBS guidance could be acknowledged in the Guide and these variations would usefully be documented in the FSI metadata so as to ensure transparency in determining the comparability of data.

  • With the gradual adoption of Basel II across and within counties, there may be issues of comparability of data both across countries as well as over time and across deposit takers within a given country (Box 9). Half of the CCE countries that responded to the survey expressed the view that with the introduction of Basel II, FSI data will become less comparable within a country, while 45 percent said that there would be less comparability across countries.24 In this context, the FSI metadata would need to document how BCBS guidance is implemented in each country regarding the supervisory-based series used to compile FSIs.

  • The Guide may need to be amended to include a discussion on Basel II, in addition to the existing discussion on Basel I.

Key Differences Between Basel I and Basel II

Basel I and Basel II both employ a common set of definitions for capital (i.e., total regulatory capital and Tier 1 capital) for calculation of FSIs that use these as underlying data. The key difference pertains to the method of calculation of another FSI-underlying series—risk-weighted assets.

Under Basel I, risk-weighted assets are based on a uniform set of risk weights that are assigned to common asset categories. For example, all deposit takers must assign the same risk-weight to loans to OECD governments with no difference in the risk weighting.

Basel II reflects the evolution of the Basel capital adequacy framework, as the assignment of risk-weights now more closely reflects loss experience for asset types. Basel II provides a further option that allows countries to use more than one approach for the assignment of the risk weights. These approaches range from the simpler standardized approach (which is very similar to Basel I), to more sophisticated internal ratings based (IRB) approaches that allow deposit takers to use their own risk models to determine the minimum capital.

33. Taking into consideration the matters above, the following amendments to the Guide were deemed appropriate: 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. Metadata forms for potential reporting to the IMF will be amended accordingly.

D. Relation of the Guide with IFRS

34. An issue that arose in the CCE was whether the few differences between the Guide and IFRS should be maintained or the Guide should defer to IFRS on accounting issues.

The vast majority of CCE participating countries (above 90 percent) responding to the survey expressed the view that the Guide should defer to IFRS on all accounting issues, except on the matter of the consolidation basis (Table 13).25 Apart from the latter, the divergences between the Guide and IFRS concern only five areas, and responding countries (91 to 95 percent of them, depending on the area of divergence) felt that these differences should be eliminated by following IFRS guidance. The availability of data and cost of making the adjustments to IFRS-consistent data to conform to the Guide's recommendations, as well as the harmonization of standards were noted as important considerations. Moreover, there do not appear to be overwhelming analytical arguments in favor of maintaining the current accounting recommendations of the Guide in the areas of divergence with IFRS.

35. Taking into consideration the matters above, the following amendments to the Guide were deemed appropriate: The Guide will be amended to defer to IFRS as the standard on accounting issues.26 Metadata forms for potential reporting to the IMF will be updated accordingly.

E. Contents of FSI List

36. The issue is whether there should be additions or deletions of FSIs regarding the list endorsed by the Executive Board and used in the CCE. There have been proposals advanced by about 60 percent of CCE participating countries that responded to the survey to eliminate some FSIs from the current list, while the remaining countries expressed the view that no FSI should be deleted from the current list (Table 13). While the majority of responding countries did favor some deletion from the list, there was significant diversity in their views as to which FSI should be eliminated. In all cases of FSIs being proposed for elimination, the view was supported by a couple of countries only, except in the case of encouraged FSI I14 (Large exposures to capital), the elimination of which was supported by about a quarter of responding countries. There were also numerous proposals for additions to the FSI list. However, there was little convergence on specific FSI formulations in the proposals for new FSIs offered by countries. The need to develop FSIs for OFC subsectors was however noted by several countries. Overall, 35 percent of countries proposed to add to the FSI list, while 30 percent of countries indicated that no new FSIs should be added to the list, and another 35 percent had no view on the matter.

37. A specific area for adding new FSIs that attracted significant interest throughout the CCE, has been the oFC sector. In the CCE plenary meeting in May 2007, there were calls for adding FSIs for major OFC subsectors (such as insurance, pension funds, other financial intermediaries, and financial auxiliaries) to the current list of FSIs. This was also the area for addition of FSIs pointed out most frequently by countries in the survey, although the number of countries that mentioned the matter in their survey responses was still quite limited. While some countries have offered some concrete proposals for such FSIs, in the CCE meeting the general view was that there is a need to reflect further on the appropriate FSIs for OFCs to include in the list, taking into consideration the new international supervisory standards that are being proposed as well as ongoing analytical research on the matter.

38. Taking into consideration the matters above, the following approach was deemed appropriate: 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 below. Over the medium term, work would continue on assessing the need and feasibility of refining the list of FSIs and adding new FSIs.27 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.

F. Definitions of Specific FSIs

39. There were proposals advanced by about 74 percent of responding countries to change the Guide's definitions of certain data series used to compile FSIs. However, the survey results show little convergence on specific reformulations of FSIs in the proposals offered by countries. Nevertheless, from the discussions in the CCE plenary meeting, certain FSIs/underlying data series did appear to emerge as potential candidates for redefinition. They include the encouraged FSIs I14 (large exposures to capital), I32 (Number of applications for protection from creditors), I12 (net open position in foreign exchange to capital), and I25 (net open position in equities to capital), as well as the following underlying series for various FSIs: liquid assets, short-term liabilities, nonperforming loans, and net income. It should be noted, however, that in each of these cases, there were also participants to the CCE meeting who expressed support for retaining the current definitions of the Guide.

40. In considering amending the definitions of the identified FSIs and underlying series, the following issues were raised:

  • Liquid assets: A view expressed by some CCE participating countries is that customer loans with up to three months remaining maturity could be excluded from liquid assets, either because of their perceived low liquidity or on prudential grounds. In support of the continued inclusion of such customer loans in liquid assets, there is some evidence that in periods of uncertainty banks reduce the maturity of their customer loans.28 Moreover, the Guide currently recognizes that determining which assets are liquid involves judgment and allows flexibility in defining the series in each country (which may affect crosscountry comparability of data). Only one country compiled data for the CCE excluding customer loans. Some countries have also argued that liquid assets should not exclude interbank positions, on the basis of their liquidity as well as for consistency with the definition of total assets. Moreover, the inclusion of interbank positions would render the resulting FSI comparable to the ratios reported by individual DTs. At the same time, the Guide argues that interbank positions do not constitute an "external" source of liquidity for the sector.

  • Short-term liabilities: Some counties expressed the view that short-term liabilities should exclude customer deposits, given their perceived stability. Only one country compiled data for the CCE on this basis. Other countries have noted that such deposits could be volatile and therefore should be retained in the definition, as currently recommended in the Guide. Another matter concerns interbank positions: if the definition of liquid assets were to be modified to include them, then the definition of short-term liabilities would arguably need to be modified in that respect also.

  • Nonperforming loans: There were views among CCE participating countries to modify the definition by having a uniform criterion of 90 days for declaring a loan as nonperforming (i.e., by excluding stricter approaches) so as to improve cross-country comparability of data. This approach would also make the definition consistent with the draft of the ongoing System of National Accounts (SNA) revision and the forthcoming sixth edition of the Balance of Payments Manual.

  • Net Income: Some countries have argued for keeping only one of the two options currently offered in the Guide—the pre-tax (and extraordinary items) measure of net income or the post-tax measure—to promote cross-country comparability. To avoid the effects of different tax regimes, the pre-tax measure would be preferable. This measure was followed by 82 percent of countries in their CCE submissions.

  • Large exposures to capital: This indicator was identified by a number of countries responding to the survey either for elimination (25 percent) or redefinition (7.6 percent). The current three versions of the FSI are seen by many countries as either unsuitable for measuring large exposures (in the case of the version measuring total exposure of the 5 largest DTs to the 5 largest resident entities to capital and the version assessing total exposure of DTs to affiliated entities and connected parties to capital) or leading to low cross-country comparability of data (such as in the version measuring the number of large exposures). Several countries used in their CCE submissions an alternative definition of the indicator—total value of large exposures to capital.

  • Number of applications for protection from creditors: The indicator may not be applicable to different countries' legal frameworks, leading to low cross-country comparability of data. To address these issues, some countries proposed that the indicator be modified to measure the cases where bankruptcy proceedings are actually initiated.

  • Net open position in foreign exchange to capital, large exposures to capital, and net open position in equities to capital: The view was expressed that for these ratios, which are assessing solvency, only supervisory-based series of capital should be used, and not the balance sheet series for capital, as currently offered as an option in the Guide.

41. Taking into consideration the matters above, the following approach was deemed appropriate: All FSIs and their underlying series will retain the definitions currently recommended in the Guide. A limited number of exceptions follow, for which data and metadata report forms will be amended accordingly.

  • a. The definition of the data series "liquid assets," underlying the FSIs "liquid assets to total assets" and "liquid assets to short-term liabilities," will be amended to include interbank positions.

  • b. The definition of the data series "short-term liabilities" underlying the FSI "liquid assets to short-term liabilities" will be amended to include interbank positions.

  • c. The series "nonperforming loans" used to compile the FSIs "nonperforming loans net of provisions to capital" and "nonperforming loans to total gross loans" will be redefined on the basis of a uniform criterion of "principal or interest payments 90 days overdue" for declaring a loan as nonperforming (i.e., the definition would not allow, as it currently does, stricter approaches).

  • d. The Guide will be amended to restrict the definition of net income to net income before taxes and extraordinary items.

  • e. The FSI on large exposures will be redefined as the ratio of the value of large exposures to capital, replacing the current versions of the indicator offered as options in the Guide.

  • f. The FSIs "net open position in foreign exchange to capital," "large exposures to capital," and "net open position in equities to capital" will be redefined to be compiled using total regulatory capital or Tier 1 capital.

  • g. The FSI "number of applications for protection from creditors" will be modified to measure the cases where bankruptcy proceedings are actually initiated.

G. Conclusion

42. As discussed above in this paper, there is a need to amend the Guide. It is intended that some aspects of the Guide be amended, as noted above, to deal with the methodological issues that arose in the conduct of the CCE. The current Guide together with an annotation on the amendments would provide the technical underpinning for future Fund work on FSIs as proposed in the Financial Soundness Indicators: Experience with the Coordinated Compilation Exercise and Next Steps.

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.

Table 2.a.

Core FSIs for Countries Using the DCCB Consolidation Basis

(In percent)

article image
Source: CCE.

In parenthesis is the indicator number.

article image
Table 2.b.

Core FSIs for Countries Using the DC Consolidation Basis

(In percent)

article image
article image
article image
article image
Source: CCE.

In parenthesis is the indicator number.