Macroprudential Indicators of Financial System Soundness ...

O C C A S I O N A L PA P E R 192

Macroprudential Indicators ofFinancial System Soundness

By a Staff Team led byOwen Evans, Alfredo M. Leone, Mahinder Gill, and Paul Hilbers

and comprisingWinfrid Blaschke, Russell Krueger, Marina Moretti,

Jun Nagayasu, Mark O’Brien, Joy ten Berge,and DeLisle Worrell

INTERNATIONAL MONETARY FUND

Washington DC

April 2000

Preface v

List of Abbreviations vi

I Overview 1

II Indicators for Macroprudential Surveillance 3

Aggregated Microprudential Indicators 4Macroeconomic Factors That Impact the Financial System 10Directions for Further Work 12

III Literature Survey 13

Determinants of Financial System Soundness 13Studies of Macroeconomic Variables 14Studies of Aggregated Microprudential Indicators 16

IV Work Programs of Other Institutions 18

International and Multilateral Institutions 18National Central Banks and Supervisory Agencies 20Indicators Used by Investors and Rating Agencies 24

V IMF Initiatives 26

Reports and Publications 26Surveillance Procedures and Operations 26

VI Measurement Issues 29

Statistical Frameworks for MPIs 29Statistical Issues Affecting MPIs and International Comparability 31Options for Further Development of MPIs 34

VII Macroprudential Indicators and Data Dissemination 37

IMF Initiatives in Data Dissemination Standards 37Conclusions on Dissemination Issues from the Consultative Meeting 38Next Steps 38

Appendices

I Existing Data Collection Frameworks 40

II Special Data Dissemination Standard and General DataDissemination System 43

Contents

iii

CONTENTS

References 45

Box

SectionI 1. Main Conclusions of the Consultative Meeting on Macroprudential

Indicators and Data Dissemination 2

Tables

SectionII 1. Summary of Macroprudential Indicators 4III 2. Macroprudential Indicators in a Selection of Recent Studies 15IV 3. Comparative Listing of Indicators Used by Selected

Country Authorities 21VI 4. Statistical Issues Affecting MPIs 32

iv

The following symbols have been used throughout this paper:

. . . to indicate that data are not available;

— to indicate that the figure is zero or less than half the final digit shown, or that the itemdoes not exist;

– between years or months (for example, 1998–99 or January–June) to indicate the yearsor months covered, including the beginning and ending years or months;

/ between years (for example, 1998/99) to indicate a crop or fiscal (financial) year.

“Billion” means a thousand million.

Minor discrepancies between constituent figures and totals are due to rounding.

The term “country,” as used in this paper, does not in all cases refer to a territorial entity thatis a state as understood by international law and practice; the term also covers some territor-ial entities that are not states, but for which statistical data are maintained and provided in-ternationally on a separate and independent basis.

The international financial turmoil of the second half of the 1990s has provokedmuch reflection and analysis within the international community on ways tostrengthen the international financial system. Together with other international orga-nizations, national authorities, and the private sector, the IMF has been working on aseries of initiatives intended to contribute to a more stable and efficient financial sys-tem, and toward better preparedness to address future systemic problems. Amongthese initiatives are the ongoing efforts to develop and use macroprudential indica-tors—defined broadly as indicators of the health and stability of financial systems.This paper aims to take stock of current knowledge in the area of macroprudential indicators—notably, analytical, identification, and measurement issues—with a viewto providing reference material for national authorities, the private sector, and otherusers of macroprudential indicators. The paper also looks at issues related to the useof macroprudential indicators in IMF surveillance, and possible ways to encouragetheir dissemination through the IMF Special Data Dissemination Standard or in otherways.

The material in this paper was originally prepared for discussion at a September1999 consultative meeting at the IMF with high-level experts from central banks, su-pervisory agencies, international financial institutions, academia, and the private sec-tor. A revised paper reflecting the results of the consultative meeting was used in dis-cussions in the IMF’s Executive Board in January 2000. The final paper has furtherbenefited from comments by Executive Directors and colleagues in the IMF.

The paper was prepared under our direction by a joint staff team led by OwenEvans, Alfredo M. Leone, Mahinder Gill, and Paul Hilbers, and consisting of WinfridBlaschke, Russell Krueger, Marina Moretti, Jun Nagayasu, Mark O’Brien, Joy tenBerge, and DeLisle Worrell. We would like to pay a special tribute to the late OwenEvans, who together with V. Sundararajan, was a major initiator of this project. Wewould like to thank Helen Chin of the External Relations Department for editing andcoordinating production of this Occasional Paper. The views expressed in this paperare those of IMF staff and do not necessarily reflect the views of national authoritiesor of IMF Executive Directors.

Carol S. Carson Stefan IngvesDirector Director

Statistics Department Monetary and Exchange Affairs Department

Preface

v

BCBS Basel Committee on Banking Supervision, BISBIS Bank for International SettlementsBSC Banking Supervision Committee, ECBCGFS Committee on the Global Financial System, BISCPSS Committee on Payment and Settlement Systems, BISDSBB Dissemination Standards Bulletin BoardECB European Central BankEMU Economic and Monetary Union, EUESA95 European System of Accounts, 1995ESCB European System of Central BanksEU European UnionFDIC Federal Deposit Insurance CorporationFIMS Financial Institutions Monitoring SystemFSA Financial Sector AssessmentFSAP Financial Sector Assessment ProgramFSF Financial Stability ForumFSSA Financial System Stability AssessmentG–7 Group of SevenG–10 Group of TenGDDS General Data Dissemination SystemGMS Growth Management SystemIAIS International Association of Insurance SupervisorsIOSCO International Organization of Securities CommissionsMAE Monetary and Exchange Affairs Department, IMFMPI Macroprudential IndicatorMUFA Monetary Union Financial AccountsOCC Office of the Comptroller of the CurrencyOECD Organization for Economic Cooperation and DevelopmentSDDS Special Data Dissemination StandardSNA System of National AccountsSNA93 System of National Accounts, 1993STA Statistics Department, IMFUBSS Uniform Bank Surveillance SystemVaR Value at Risk

List of Abbreviations

vi

Substantial progress has been made during recentyears in forging a consensus on the importance

of strengthening the architecture of the internationalfinancial system. The international community, act-ing through various forums, has identified a numberof priorities in this work, including the need to en-hance its own—and the markets’—ability to assessthe strengths and vulnerabilities of financial sys-tems, and to develop the analytical and proceduraltools needed to perform this task. In particular, theimportance of assessing the soundness of financialsystems as part of the IMF’s surveillance work wasgiven prominence by the Group of Twenty-Two fi-nance ministers and central bank governors in theReport of the Working Group on Strengthening Fi-nancial Systemsin October 1998. The workinggroup recommended that financial sector surveil-lance be anchored to the IMF surveillance process,with expert support from the World Bank and else-where. This process is now well under way as part ofthe joint World Bank-IMF Financial Sector Assess-ment Program (FSAP), and the related FinancialSystem Stability Assessments (FSSAs).1 The devel-opment and possible dissemination of so-calledmacroprudential indicators (MPIs)—defined broadlyas indicators of the health and stability of financialsystems—have been encouraged recently by boththe Group of Seven (G–7) and the IMF InterimCommittee.2 Such indicators will be critical in pro-ducing reliable assessments of the strengths and vul-nerabilities of financial systems as part of IMF sur-veillance, and to enhancing disclosure of keyfinancial information to markets.

This paper aims to take stock of current knowl-edge in the area of MPIs—notably, analytical, iden-tif ication, and measurement issues—with a view toproviding reference material for national authori-ties, the private sector, and other users of MPIs. Thepaper also looks at issues related to the use of MPIsin IMF surveillance, and their dissemination eitherthrough the IMF Special Data Dissemination Stan-dard (SDDS), or in other ways. In particular, thepaper looks at:

• the MPIs that could be used most effectively bythe IMF in its surveillance work under Article IVof the IMF’s charter and within the framework ofthe FSSAs;

• the modalities and options for the compilation ofsuch data; and

• the possible dissemination of MPIs to the public,including through the SDDS.

This paper has benefited from feedback providedduring a consultative meeting on macroprudentialindicators and data dissemination, which was heldat IMF headquarters on September 10–11, 1999.3

The objectives of this outreach meeting were to dis-cuss experiences of member countries and the inter-national community in identifying and using MPIsfor analyzing financial sector soundness, and toconsider possible modes of disseminating these in-dicators to the public. Participants in the consulta-tive meeting included high-level experts from cen-tral banks, supervisory agencies, internationalfinancial institutions, academia, and the private sec-tor (banks, investment funds, rating agencies). IMFmanagement, senior staff, and representatives of theExecutive Board also participated. The main con-clusions of the consultative meeting are summa-rized in Box 1.

One difficulty with identifying MPIs for use insurveillance work is that the research conducted sofar has not produced a consensus on a core set of in-dicators. This is, in part, because different indicatorsmay be relevant in different circumstances. It may

I Overview

1

1The FSAP was launched jointly with the World Bank on apilot basis in May 1999. The program is designed to identify fi-nancial system strengths and vulnerabilities and to help developappropriate policy responses. The FSSA reports, which focus onfinancial system issues of significance for macroeconomic perfor-mance and policies, are prepared on the basis of the FSAP byIMF staff for discussion in the IMF Executive Board, within thecontext of Article IV surveillance. In the World Bank, the FSAPreports provide the foundations for the formulation of financialsector development strategies.

2See United States, Department of the Treasury (1999) and In-ternational Monetary Fund (1999a). 3See Hilbers, Krueger, and Moretti (1999) for details.

I OVERVIEW

also reflect the short time that analytical work in thisarea has been done. In any case, this has meant thatthe initial set of MPIs that the IMF is experimenting

with, in its strengthened surveillance of financialsectors, has been identified as much through past ex-perience in the field as through research.

2

Box 1. Main Conclusions of the Consultative Meeting on Macroprudential Indicators and Data Dissemination

The main conclusions reached by the participants ofthe September 1999 consultative meeting are summa-rized below:

Identif ication, Analysis, and Use of MPIs

• While work on identifying and measuring MPIs hasadvanced substantially in recent years, knowledge inthis area is still limited and more research and analy-sis is needed. In particular, there is no consensus ona model for determining the vulnerability of a finan-cial system or on a set of widely accepted MPIs.

• Prioritization among MPIs and the selection of acore set of indicators is desirable. Use of a singlecomposite indicator, however, would be overlysimplistic and could be misleading.

• Analyses of financial sector vulnerability cannotrely on quantitative indicators alone. Qualitative in-formation on institutional circumstances, com-bined with informed judgment, is also essential.

• There is a need to: (1) improve the quality of ac-counting practices in many countries; (2) assess thehealth of nonbank financial institutions and of thecorporate sector; (3) address the limitations of aggre-gating microprudential information to obtain MPIs;(4) develop benchmarks and norms for the indica-tors; and (5) use stress tests as part of a forward-looking approach to macroprudential analysis.

Measurement and Data Dissemination Issues

• Efforts should be directed toward a greater harmo-nization of MPIs in terms of coverage, periodicity,timeliness, and public access.

• No single set of MPIs is currently being dissemi-nated by a group of countries or seen as superior toother sets.

• National authorities differ in their approaches tothe dissemination of data on the financial system,and no clearly identifiable set of best practices for dissemination of MPI data has emerged. Whilethere is a presumption that disclosure of informa-tion promotes market discipline, there remain in-evitable confidentiality concerns, notably about re-leasing information on individual institutions.

• Given the substantial work ahead in crafting a coreset of MPIs, it would be premature to include MPIswithin the SDDS, though consideration should begiven to how to provide national authorities withincentives to compile and disseminate MPIs.

• It would be useful to conduct a survey of nationalsupervisors, statistical authorities, and users toevaluate prospects for compiling MPIs, in view ofthe complex questions raised about the scope ofmacroprudential work and the technical feasibilityof compiling MPIs.

The ability to monitor financial soundness presup-poses the existence of indicators that can be used

as a basis for analyzing the current health and stabilityof the financial system. These macroprudential indi-cators comprise both aggregated microprudential in-dicators of the health of individual financial institu-tions, and macroeconomic variables associated withfinancial system soundness. Aggregated micropruden-tial indicators are primarily contemporaneous or lag-ging indicators of soundness;4 macroeconomic vari-ables can signal imbalances that affect financialsystems and are, therefore, leading indicators. Finan-cial crises usually occur when both types of indicatorspoint to vulnerabilities, that is, when financial institu-tions are weak and face macroeconomic shocks.

The indicators that are the focus of this paper arequantitativevariables. The availability of these indi-cators alone is not sufficient to make an overall as-sessment of financial system soundness. Such as-sessments also depend on a broad range of elementsthat are not easily quantifiable. In particular, the ade-quacy of the institutional and regulatory frameworksgoverning the financial system significantly affectsthe system’s soundness. Elements include the struc-ture of the financial system and markets; regulationsregarding accounting and other standards, and dis-closure requirements; loan classification, provision-ing and income recognition rules, and other pruden-tial regulations; the quality of supervision offinancial institutions; the legal infrastructure (includ-ing in the areas of bankruptcy and foreclosure); in-centive structures and safety nets; and liberalizationand deregulation processes.

The importance of these qualitativeelements callsfor a high degree of experience in analyzing them, andan ability to couple the analysis of MPIs with in-formed judgment on the adequacy of the institutionaland regulatory frameworks of individual countries.5

Although it may be possible to develop a reasonablyclear picture of these elements in a given countryfairly quickly, a deeper understanding of how well thefinancial system actually works is generally expectedto develop only after careful observation over a periodof time.

Because the relevance of individual indicatorsmay vary from country to country, MPIs cannot beused mechanically.6 Rather, any assessment needsto be based on a comprehensive set of indicators,taking into account the overall structure and eco-nomic situation of a country and its financial sys-tem. In many instances, monitoring of indicatorsover time (an intertemporal comparison) can bemore meaningful than comparisons across coun-tries, due to differing accounting and prudentialstandards as well as differences in the structure offinancial systems. Changes in regulations such asaccounting and provisioning norms can, however,lead to breaks in time series.

Prudential indicators should be monitored notonly for the (narrowly defined) banking system, but,if systemically relevant, for other financial institu-tions as well, including nonbank depository corpora-tions (if they exist) and nondepository financial intermediaries.

A limited set of macroeconomic indicators thatare considered most relevant for a particular countrymay be used for stress tests, to evaluate quantita-tively the impact of large changes in those indicatorson the portfolios of financial institutions, and on theaggregate solvency of the financial system. Usingthe IMF’s macroeconomic forecasts, and observingpast relationships between macroeconomic and pru-dential indicators, it may also be possible, to somedegree, to project likely future developments in pru-dential indicators.

A set of indicators that the IMF has identifiedthrough its financial sector surveillance, technicalassistance, and program work over the years is de-

II Indicators for MacroprudentialSurveillance

3

4Observation lags should be short to allow timely monitoring.Stress testing these indicators could provide an early warning re-garding vulnerabilities.

5The IMF’s FSSAs combine the analysis of MPIs with a com-prehensive review of these qualitative aspects (see Section V).

6For example, whereas in one country an indicator may be con-structed using a narrow monetary aggregate, in another country abroad aggregate may be more meaningful.

II INDICATORS FOR MACROPRUDENTIAL SURVEILLANCE

scribed in Table 1. Background is provided onMPIs that have been used for monitoring thesoundness of financial systems, along with discus-sion of the usefulness of these indicators. The MPIsare divided into two broad categories: (1) aggre-gated microprudential indicators; and (2) indicatorsof macroeconomic developments or exogenousshocks that could affect the financial system. Table1 provides a comprehensive listing of the MPIsidentified thus far.

Aggregated Microprudential Indicators

Indicators of the current health of the financialsystem are primarily derived by aggregating indica-tors of the health of individual financial institutions.One commonly used framework for analyzing thehealth of individual institutions is the so-calledCAMELS framework, which involves the analysis ofsix groups of indicators reflecting the health of fi-nancial institutions:

4

Table 1. Summary of Macroprudential Indicators

Aggregated Microprudential Indicators Macroeconomic Indicators

Capital adequacyAggregate capital ratiosFrequency distribution of capital ratios

Asset qualityLending institution

Sectoral credit concentrationForeign currency-denominated lendingNonperforming loans and provisionsLoans to loss-making public sector entitiesRisk profile of assetsConnected lendingLeverage ratios

Borrowing entityDebt-equity ratiosCorporate profitabilityOther indicators of corporate conditionsHousehold indebtedness

Management soundnessExpense ratiosEarnings per employeeGrowth in the number of financial institutions

Earnings and profitabilityReturn on assetsReturn on equityIncome and expense ratiosStructural profitability indicators

LiquidityCentral bank credit to financial institutions Segmentation of interbank ratesDeposits in relation to monetary aggregatesLoans-to-deposits ratiosMaturity structure of assets and liabilities (liquid asset ratios)Measures of secondary market liquidity

Sensitivity to market riskForeign exchange riskInterest rate riskEquity price riskCommodity price risk

Market-based indicatorsMarket prices of financial instruments, including equityIndicators of excess yieldsCredit ratingsSovereign yield spreads

Economic growthAggregate growth ratesSectoral slumps

Balance of paymentsCurrent account deficitForeign exchange reserve adequacyExternal debt (including maturity structure)Terms of tradeComposition and maturity of capital flows

InflationVolatility in inflation

Interest and exchange ratesVolatility in interest and exchange ratesLevel of domestic real interest ratesExchange rate sustainabilityExchange rate guarantees

Lending and asset price boomsLending boomsAsset price booms

Contagion effectsTrade spilloversFinancial market correlation

Other factorsDirected lending and investmentGovernment recourse to the banking systemArrears in the economy


• Capital adequacy,• Asset quality,• Management soundness,• Earnings,• Liquidity, and • Sensitivity to market risk.7

Indicators of market perceptions often supplementthese indicators. Because the CAMELS categoriza-tion of indicators is helpful in analyzing the variouspossible areas of vulnerability, the discussion of sys-temwide indicators in this chapter follows the samestructure.

Capital Adequacy Indicators

Capital adequacy and availability ultimately deter-mine the robustness of financial institutions toshocks to their balance sheets. Thus, it is useful totrack capital adequacy ratios that take into accountthe most important financial risks—foreign ex-change, credit, and interest rate risks—includingrisks involved in off-balance sheet operations, suchas derivative positions.8

Aggregate Risk-Based Capital Ratios.The mostcommonly used indicator in this respect is the aggre-gate risk-based capital ratio (the ratio of capital torisk-adjusted assets). A declining trend in this ratiomay signal increased risk exposure and possible cap-ital adequacy problems. It is possible to estimatevulnerability based on average sectorwide capitaladequacy ratios, but these may be misleading undersome circumstances (see Section VI). In addition toadequacy, it may also be useful to monitor indicatorsof capital quality. In many countries, bank capitalconsists of different elements that have varyingavailability and capability to absorb losses, evenwithin the broad categories of Tier 1, Tier 2, and Tier3 capital.9 If these capital elements can be reportedseparately, they can serve as more reliable indicatorsof the ability of banks to withstand losses, and helpin putting overall capital ratios into context.

Frequency Distribution of Capital Ratios.As analternative to the use of aggregate capital ratios, it

may be possible to build an aggregate view based onthe analysis of the capital ratios of individual institu-tions, or groups of selected large institutions, such asthe three largest banks. It may often be useful tofocus on particular subgroups such as state-ownedbanks and previously intervened banks. Another wayof avoiding problems of aggregation is to look at thenumber of banks (and their market share) with risk-based capital ratios below certain thresholds, such asthe minimum required under international or domes-tic standards.10

Asset Quality Indicators

The reliability of capital ratios depends on the reli-ability of asset quality indicators. Risks to the sol-vency of financial institutions often derive from im-pairment of assets, so it is important to monitorindicators of asset quality. First, we deal with indica-tors that directly reflect the current state of creditportfolios; macroeconomic indicators that indirectlyimpact asset quality are outlined below. Indicators ofasset quality need also to take into account credit riskassumed off-balance sheet via guarantees, contingentlending arrangements, and derivatives. In some coun-tries, trust activities and operations of offshore banksalso pose significant contingent risk and the indica-tors should, as much as possible, reflect consolidatedinformation. Indicators of asset quality include indi-cators at the level of the lending institution, and indicators at the level of the borrowing institutions.

Indicators at the Level of the Lending Institution

Sectoral Credit Concentration.A large concentra-tion of aggregate credit in a specific economic sectoror activity, especially commercial property, may sig-nal an important vulnerability of the financial sys-tem to developments in this sector or activity. Manyfinancial crises in the past (including the Asiancrises) have been caused or amplified by downturnsin particular sectors of the economy spilling overinto the financial system via concentrated loanbooks of financial institutions. In practice, this has

5

7On an aggregate basis for the financial system as a whole,however, some of the indicators that are useful for individual in-stitutions may not be applicable and meaningful. Problems of ag-gregation and measurement are discussed further in Section VI.

8Actual (observed) capital adequacy ratios are lagged indica-tors of banking problems—by the time capital adequacy ratiosshow a decline, financial institutions generally have already beenexperiencing serious problems.

9Tier 1 capital consists of permanent shareholders’equity anddisclosed reserves; Tier 2 capital consists of undisclosed reserves,revaluation reserves, general provisions and loan-loss reserves,hybrid debt-equity capital instruments, and subordinated long-term debt (over five years); Tier 3 capital consists of subordinatedshort-term debt (over two years). See BIS (1988, 1996).

10The analysis of financial sector stability may sometimes re-quire information on the condition of individual large banks be-cause of their market power or the possibility of contagion toother firms; see, for example, Downes, Marston, and Ötker(1999). A specific problem for macroprudential analysis is how tointegrate (1) microinformation on specific firms, which is highlyaffected by accounting and supervisory standards and the struc-ture of the firm’s global operations; (2) information on the struc-ture of the industry (e.g., concentration, foreign ownership, publicsector institutions, overconcentrated lending); and (3) nationalmacroeconomic information. This process might involve usingmeasures of dispersion, concentration, large-bank group analysis,or multivariate analysis.


often been the case for concentration in real estate,which can be subject to severe boom and bust pricecycles. Loan concentration can be dangerous in al-most any sector of the economy, however, includingcommodities and certain export industries.

Foreign Currency-Denominated Lending. Severalfinancial crises have been preceded by periods offast growth of foreign currency-denominated creditto domestic firms that frequently lacked a stablesource of foreign exchange revenues.11 These trans-actions shift the foreign exchange risk to final bor-rowers, but often imply a higher credit risk to thelenders.12

Nonperforming Loans.An increasing trend in theratio of nonperforming loans to total loans signals adeterioration in the quality of credit portfolios and,consequently, in financial institutions’cash flows,net income, and solvency.13 It is often helpful to sup-plement this information with information on non-performing loans net of provisions, and on the ratioof provisions plus interest suspension on impairedloans to total loans—particularly if impaired loanshave not yet been classified as nonperforming.14 Al -though these indicators are primarily backwardlooking, reflecting past problems that have alreadybeen recognized, they can be useful indicators of thecurrent health of the financial system, and are oftenused in connection with stress tests of financial insti-tutions. Trends in nonperforming loans should belooked at in conjunction with information on recov-ery rates—for example, using the ratio of cash re-coveries to total nonperforming loans. Such infor-mation points to the level of effort or the ability offinancial institutions to cope with high nonperform-ing loan portfolios.

Loans Outstanding to Loss-Making Public SectorEntities(notably public enterprises or regional gov-

ernments). The presence of such loans, which areoften the result of past directed lending, may alsosignal significant credit risk. Depending on thecountry, loans to loss-making public enterprises or toregional governments may not be classified as non-performing, even though they may not be repaid on atimely basis and/or in full.

Risk Profile of Assets(ratio of risk-weighted as-sets to total assets by weight category). A high ratioof investment in securities with low regulatory riskweights (such as bonds issued by governments ofOECD member countries) to total assets usually in-dicates a conservative investment policy on the partof financial institutions. At the same time, it is oftena reflection of the structure of the economy, and reg-ulatory incentives that favor government financing inparticular. In some instances, however, it might bean indication of trouble at some institutions that in-vest in securities with low risk weights because ofcapital adequacy problems.15

Connected Lending. A high ratio of connectedlending to total loans indicates a concentration ofcredit risk on a small number of borrowers, that is, alack of diversification. Lending to entities that formpart of the same group as the financial institution it-self is common in many countries, and can be in-dicative of deficiencies in credit analysis. Loans toentities of the same group are often easily approved(“pocket banks”), regardless of credit quality, andproblems in these entities can spill over into the fi-nancial institution.16

Leverage Ratios.Financial institutions’lever-age—measured by the ratio of assets to capital—increases when bank assets grow at a faster ratethan capital. For institutions that are primarily in-volved in lending activities, the ratio of loans tocapital roughly approximates the leverage. It is thereverse of the capital adequacy ratio (a simplifiedversion).17

Indicators at the Level of the Borrowing Entity

The quality of financial institutions’loan portfo-lios is directly dependent upon the financial healthand profitability of the institutions’borrowers, espe-cially the nonfinancial enterprise sector. Therefore,any analysis of asset quality needs to take into ac-count indicators of the likelihood of borrowers torepay their loans.

6

11Among other factors, differential reserve requirements insome countries create incentives for foreign currency-denomi-nated intermediation by making it relatively more competitivethan intermediation denominated in domestic currency.

12Foreign currency-denominated lending is often measured as apercentage of total lending. Aggregate figures on foreign currencylending are usually available, but in countries where only a few in-stitutions have access to foreign exchange, the lending patterns ofthese particular institutions may merit individual attention.

13Adequate loan classification and accounting standards are es-sential for the ratio of nonperforming loans to total loans to bemeaningful. The utility of this ratio may also be diminished in anenvironment in which banks tend to roll over loans that otherwisewould become nonperforming—a practice also described as“evergreening.”

14Some countries allow the netting of the collateral valueagainst the impaired loan in calculating the provisions for loanlosses. Different rules in this respect may make cross-countrycomparison of provisioning data difficult. Under some circum-stances, when netting is allowed, provisioning ratios may becomemeaningless due to difficulties with valuing and liquidating collat-eral (e.g., real estate collateral subsequent to a real estate bubble).

15One advantage of these ratios as MPIs is their easy availabil-ity from prudential returns.

16Even though this ratio is usually low on an aggregate basis,risk can be significant in countries with small numbers of largeconglomerates.

17Therefore, it has similar drawbacks. Nevertheless, it can be auseful indicator where loan valuation may be regarded as adequate.


Debt-Equity Ratios.Excessive corporate borrow-ing has often preceded periods of financial systemdistress. Thus it is important to monitor nonfinancialprivate sector leverage.18 Fast growth of corporateindebtedness—for example, at a rate higher thanGDP growth—may be seen as a sign that banks’credit screening procedures have been relaxed. It isimportant to monitor if the increase in corporate in-debtedness is concentrated in sectors that are partic-ularly vulnerable to shifts in economic activity, suchas real estate, or to exogenous economic shocks,such as export industries.19

Corporate Sector Profitability. Sharp declines incorporate sector profitability, for example, as a resultof economic deceleration, may serve as a leading in-dicator of financial system distress.20

Other Indicators of Corporate Conditions.Be-sides debt-equity ratios, several other indicators alsoprovide information on corporate financial vulnera-bility. These include cash flow-based indicatorssuch as the interest coverage ratio (the ratio of oper-ating income to interest expenses), and compositeindicators such as the Altman’s Z-score.21 Alterna-tive indicators that could help assess the conditionsof corporations and the implications for the bankingsystem, include delays in payments, the trend in thecurrentness of loans to the largest borrowers, andfrequency information on application for protectionfrom creditors.22

Household Indebtedness.The quality of bankportfolios also depends on the condition of borrow-ers from the household sector. Information on theoverall level of household indebtedness is useful inthis context.

Management Soundness Indicators

Sound management is key to financial institu-tions’performance. Indicators of the quality of man-agement, however, are primarily applicable to indi-vidual institutions, and cannot be easily aggregatedacross the sector. Although aggregated indicatorscan be used, they are more likely to reflect financialsector structure and the country’s economic situation

than management quality. Although several indica-tors can be used as proxies for the soundness ofmanagement, such evaluation is still primarily aqualitative exercise, particularly when it comes tothe evaluation of the management of operationalrisk, that is, the functioning of internal control sys-tems. This being said, the following indicators aresometimes used.

Expense Ratios.A high or increasing ratio of ex-penses to total revenues can indicate that financialinstitutions may not be operating efficiently. Thiscan be, but is not necessarily due to management de-ficiencies. In any case, it is likely to negatively affectprofitability.

Earnings per Employee. Similarly, low or decreas-ing earnings per employee can reflect inefficienciesas a result of overstaffing, with similar repercussionsin terms of profitability.

Expansion in the Number of Financial Institu-tions.Another possible ratio of management sound-ness is the rate of expansion in the number of banksand other financial institutions. Whereas some ex-pansion may reflect a healthy degree of competition,too rapid a rate of expansion may indicate lax licens-ing requirements, unsound management, and a gapin the supervisory capacity.

Earnings and Profitability Indicators

As chronically unprofitable financial institutionsrisk insolvency, it is important to follow indicators ofprofitability. Declining trends in those indicatorsmay signal problems regarding the profitability of fi-nancial institutions. On the other hand, unusuallyhigh profitability may be a sign of excessive risk tak-ing. The following (aggregate) ratios can serve as in-dicators of current financial sector profitability.

Return on Assets.The ratio of (net) profits to aver-age total assets is one of the most commonly usedmeasures of profitability. The ratio can be calculatedwith various profit measures, for example, before orafter provisions, and before or after tax charges and(net) extraordinary items.23

Return on Equity. The ratio of (net) profits to aver-age capital reflects the average return investors getfrom holding bank capital. The ratio has to be inter-preted with caution, since a high ratio may indicateboth high profitability as well as low capitalization,and a low ratio can mean low profitability as well ashigh capitalization. The usefulness of this ratio canbe enhanced by employing different measures ofcapital, for example Tier 1 capital only versus totalcapital, and different measures of profits.

7

18Corporate debt-equity ratios depend, in part, on countries’legal definitions of debt and equity, and, therefore, are not easilycomparable across economies.

19Few countries have reliable disclosure laws, however, so thatdata on corporate debt to equity ratios may have to be obtainedfrom bank supervisors, if they collect bank information on theirclients’credit quality and on large borrowers or credit concentra-tion, or by observing shifts in lending practices.

20Care should be taken to identify cyclical movements in cor-porate sector profitability.

21See the literature survey in Section III for details.22The latter indicator can be influenced by the quality of bank-

ruptcy and related legislation.

23For comparisons between countries, pretax profits should be used to eliminate the effects of different national taxationpractices.


Income and Expense Ratios. In order to get aclearer picture of the sustainability of profits, and ofthe extent of risk-taking by financial institutions, it isuseful to look at the sources of profitability such as(net) interest income, commissions, trading and for-eign exchange results, and other operating income.Similarly, expense ratios can reveal sources of prof-itability problems. Expense ratios can be calculatedon various kinds of expenses—staff expenses, ad-ministrative expenses, and other expenses.24 Ratioscan be constructed by setting these against measuresof total income and/or average total assets.

Structural Indicators. In addition to indicators ofcurrent profitability, there are a number of forward-looking indicators that are more geared towardmedium- and long-term profitability. A narrowbank customer base, for example, may signal com-petitiveness problems of domestic institutions andtheir inability to foster financial deepening througha wider customer base. These problems have impli-cations for financial system costs, margins, andprofitability. The size of, and changes in interestrate spreads indicate whether institutions are oper-ating in a favorable environment—and may signalthe existence of oligopolistic financial marketstructures.

Liquidity Indicators

Initially solvent financial institutions may be driventoward closure because of poor management of short-term liquidity, so it is also important to monitor liquid-ity indicators. On the liability side, indicators shouldcover funding sources, including interbank and centralbank credits. Liquidity indicators should also be ableto capture large maturity mismatches in the largest fi-nancial institutions or in the overall financial sector.25

Central Bank Credit to Financial Institutions.Alarge increase in central bank credit to banks andother financial institutions—as a proportion of theircapital or their liabilities—often reflects severe liquidity (and frequently also solvency) problems inthe financial system.

Segmentation. A high dispersion in interbank ratesmay signal that some institutions are consideredrisky. Banks may also control their interbank posi-tions by using quantitative controls, and high-risk in-stitutions might be forced to engage in aggressivebidding for deposits. Changes in interbank creditlimits or an unwillingness of some institutions tolend to other ones may indicate serious concerns.Very often, banks themselves first detect problems

as they are exposed, or potentially exposed, to trou-bled institutions in the interbank market.

Deposits as a Share of Monetary Aggregates.Adecline in the ratio of deposits to M2, for example,may signal a loss of confidence and liquidity prob-lems in the banking system. It could also indicatethat nonbank financial institutions are more effi-cient in that they offer an array of other financialproducts, or they are acting as banks in all but inname, or they may have set up pyramid schemes.

Loans-to-Deposits Ratios.Viewed over time, theratio of credit to total deposits (excluding interbankdeposits) may give indications of the ability of thebanking system to mobilize deposits to meet creditdemand. A high ratio may indicate stress in thebanking system and a low level of liquidity to re-spond to shocks.26

Maturity Structure of Financial Institutions’As-sets and Liabilities.Indicators that reflect the matu-rity structure of the asset portfolio, such as theshare of liquid assets to total assets (liquid assetratio), can uncover excessive maturity mismatchesand highlight a need for more careful liquiditymanagement. A major shortening in the maturitystructure of financial institutions’liabilities mayimply a higher liquidity risk and could also reflectthe uncertainty of depositors and other creditors onthe long-term viability of the institutions.27

Secondary Market Liquidity. Liquid asset ratiosshould be seen in connection with measures of thebreadth and depth of secondary markets for liquidassets, such as bid-ask spreads and turnover figures.

Sensitivity to Market Risk Indicators

Banks are increasingly involved in diversifiedoperations, all of which involve one or more as-pects of market risk. A high share of investments involatile assets may signal a high vulnerability tofluctuations in the price of those assets. In general,the most relevant components of market risk are in-terest rate and foreign exchange risk, which tend tohave significant impacts on financial institutions’assets and liabilities. Moreover, in some countries,banks are allowed to engage in proprietary tradingin stock markets, so it is also of interest to track eq-uity risk. Similarly, commodity risks derived fromthe volatility of commodity prices can be importantin certain countries.28

8

24They are sometimes also used as indicators of managementproblems.

25Liquidity can change rapidly, however, requiring frequent up-dating of relevant indicators.

26In cases where liquid secondary markets exist, one could alsolook at the ratio of liquid assets to total deposits.

27Indicators of maturity structure should distinguish betweendomestic and foreign liabilities, and indicate the currency denom-ination of the liabilities.

28Most of these indicators can be extracted from prudential re-turns to supervisory authorities, some directly, others via the cap-


Foreign Exchange Risk. Large open foreign ex-change positions (including foreign exchange matu-rity mismatches) and a high reliance on foreign bor-rowing (particularly of short-term maturity) maysignal a high vulnerability of financial institutions toexchange rate swings and capital flow reversals. In-dicators of foreign exchange risk, which is incurredindirectly via foreign currency-denominated creditto local borrowers (without significant foreign cur-rency cash flow), are considered in the section oncredit risk.

Interest Rate Risk.Interest rate risk is one of themost common financial risks, and virtually all financial institutions are subject to it. Even thoughit is considered here as a market risk indicator, interest rate risk arises from both an institution’sbanking book as well as from its trading book.29

Equity Price Risk. Financial institutions can, inmany countries, incur substantial equity price risk,either by trading or investing in the stock market,or via derivatives, which exposes the institutions tothe risk of stock market crashes. Indicators of eq-uity price risk would include the absolute size ofcertain classes of financial institutions’investmentin equities, their size in terms of various balancesheet indicators, or the capital charges allocatedagainst equity price risk.

Commodity Price Risk. The significance of com-modity price risk for financial institutions varies sig-nificantly from country to country. Although the in-vestment of most financial institutions incommodities or commodity derivatives is small,commodity prices are typically more volatile thanexchange or interest rates, and markets are often lessliquid. Indicators can be constructed that are similarto those for interest rate and equity risk, by lookingat the absolute size of the investment in commoditiesor by following a maturity ladder approach.30

Market-Based Indicators

Although not included in the six-group CAMELSframework, market-based assessments of the finan-

cial sector, as implied by the prices (yields) of financial instruments and the creditworthiness rat-ings of financial institutions and large corporations,can also be useful indicators of financial systemvulnerability.

Market Prices of Financial Instruments Issued byFinancial Institutions and Corporations.A declinein the stock prices of financial institutions (relativeto average stock prices) may signal adverse marketperceptions of the health of these institutions.31

Similarly, one could analyze the development ofyield spreads of tradable financial instruments is-sued by financial institutions and large corporateissuers—especially subordinated debt—to detectsigns of a “flight to quality,” notably on the part ofinvestors.

Excess Yields.Yields offered by any institution (orgroup of institutions or market) that are significantlyabove others (excluding interbank deposits) may sig-nal problems in these institutions or the existence of unsustainable schemes that would merit close examination.

Credit Ratings.A downgrade in the ratings of localfinancial institutions elaborated by international ratingagencies may signal negative market perceptions atthe international level. Credit ratings of the corporatesector can also be important, since they inform on thecreditworthiness of the banks’major borrowers. Asthe Asian crisis has shown, ratings have not alwaysbeen good indicators of vulnerability. While they arecertainly helpful in establishing an overall picture ofthe stability of the financial system, a relatively goodrating, by itself, cannot always be taken as a reliableindicator of the robustness of a country’s financialsystem. For IMF purposes, financial strength ratingsare likely to be more useful than ratings that incorpo-rate the likelihood of government support.32

Market-based indicators of a country’s vulnera-bility—such as trends in sovereign yield spreads33

and sovereign ratings—reflect the market’s assess-ment of the credit and foreign exchange risks asso-

9

ital charges allocated against the particular risks. For exchange-traded instruments, indicators may also be obtained from stockand derivatives exchanges, in particular, from position and mar-gin data. See BIS (1996).

29Supervisors often collect information on interest rate riskfrom individual financial institutions. A commonly used report-ing framework is one where a financial institution’s interest-sensitive positions are slotted into time bands, according to thetime until the next repricing. Alternatively, interest sensitivitycan be determined via duration analysis, weighting and aggre-gating the durations of individual financial instruments held bya financial institution. See BIS (1997).

30For a description of the maturity ladder approach to measur-ing commodities risk, see BIS (1996).

31If shareholders have the perception that the government willbail out troubled financial institutions, however, this data may notadequately reflect the underlying institutional risk.

32See Section IV for details. For a recent analysis of ratingagencies’performance, see BIS (1999c), and International Mon-etary Fund (1999c), Chapter V and Annex V. Since rating agen-cies generally have to rely on the published accounts of compa-nies being rated, and do not have access to supervisory data,their judgments can be affected by deficiencies in accountingand provisioning standards. On the positive side, rating agenciestry to take into account such deficiencies in their evaluations.They usually update their analysis more frequently than other in-stitutions, and may be in closer and more frequent contact withmarket participants.

33Most commonly, yield spreads are benchmarked against U.S.Treasury yields, and are subdivided into credit and foreign ex-change risks.


ciated with investing in a particular country. Fol-lowing the Asian crisis, such indicators now in-creasingly include assessments of the risks posedby a weak financial system, although the weight offinancial vulnerability in the composite is difficultto isolate.

Macroeconomic Factors That Impactthe Financial System

The operation of a financial system is dependenton overall economic activity, and financial institu-tions are significantly affected by certain macroeco-nomic developments. Most macroeconomic indica-tors are normally monitored in the broader context ofArticle IV surveillance. Recent empirical analysishas shown that certain macroeconomic developmentshave often predated banking crises, which suggeststhat financial system stability assessments need totake into consideration the broad macroeconomicpicture, particularly factors that affect the economy’svulnerability to capital flow reversals and currencycrises. The following list includes a set of indicatorsof macroeconomic developments or exogenousshocks that could affect the financial system.

Economic Growth

Aggregate Growth Rates.Low or declining aggre-gate growth rates often weaken the debt-servicingcapacity of domestic borrowers and contribute to in-creasing credit risk. Recessions have preceded manyepisodes of systemic financial distress.

Sectoral Slumps.A slump in the sectors where fi-nancial institutions’loans and investments are con-centrated could have an immediate impact on finan-cial system soundness. It deteriorates the quality offinancial institutions’portfolios and profitabilitymargins, and lowers their cash flow and reserves. Intransition economies, these problems may also arisedue to lack of progress in the restructuring of state-owned enterprises.

Balance of Payments

Current Account Deficit. A rise in the ratio of thecurrent account deficit to GDP is generally associ-ated with large external capital inflows that are inter-mediated by the domestic financial system and couldfacilitate asset price and credit booms. A large exter-nal current account deficit could signal vulnerabilityto a currency crisis with negative implications forthe liquidity of the financial system, especially if thedeficit is financed by short-term portfolio capital in-flows. Financial crises that have immediate reper-

cussions for the financial system may happen whenforeign investors consider the current account deficitunsustainably large and, hence, shift their financialinvestments out of the country.

Reserves and External Debt.A low ratio of inter-national reserves (in the central bank and financialsystem as a whole) to short-term liabilities (domesticand foreign, public and private) is seen, particularlyby investors, as a major indicator of vulnerability.Another popular indicator of reserve adequacy isgross official reserves in months of imports of goodsplus services. Total external debt and its maturitystructure are important indicators as well.34

Terms of Trade. Past experience indicates that alarge deterioration in the terms of trade has been acontributing factor to banking difficulties in manycountries. Small countries with high export concentra-tion are the most vulnerable to banking crises inducedby a sudden and large deterioration in the terms oftrade. On the other hand, large improvements in theterms of trade have the potential of causing problemsin the financial system through inflation and assetprice bubbles. These impacts are exacerbated whenthe terms of trade improvement is transitory.

Composition and Maturity of Capital Flows.Thecomposition of capital flows (portfolio versus directforeign investment; official versus private; highlyleveraged institutions and investment banks versuscommercial banks and trade finance) may also be agood indicator of potential vulnerability. Countriesare particularly vulnerable if their current accountdeficits are accompanied by low investment ratios, orby over-investment (low-productivity investments).

Inflation

Volatility in Inflation. Such volatility makes theaccurate assessment of credit and market risks moredifficult. Inflation is often positively correlated withhigher relative price volatility, a factor that raisesportfolio risk and erodes the financial institutions’information base for planning, investment, andcredit appraisal. On the other hand, a significant andrapid reduction in the rate of inflation could lead tolower nominal income and cash flows, thereby ad-versely affecting the liquidity and solvency of finan-cial institutions. In particular, in some cases bankscan profit from the management of assets in a highinflation environment, and the sudden reduction ofinflation exposes the weakness of their more tradi-tional banking practices.35 In addition, collateral

10

34Bussière and Mulder (1999).35Bank income under high inflation is often derived from the

float on payments, the inflation tax collected on nonremunerateddemand deposits, and foreign exchange dealing.

Macroeconomic Factors that Impact the Financial System

value could decline below the loan amount, particu-larly in cases of imprudent lending (including highratios of loan to collateral valuation) prior to theturnaround in inflation.

Interest and Exchange Rates

Volatility in Interest and Exchange Rates.Themore volatile these rates are, the higher the interestrate and foreign exchange risks are for financial in-stitutions. The vulnerability of the financial systemwill be higher given (1) a higher external debt bur-den, and (2) a higher share of foreign portfolio in-vestments in total foreign investment. Volatility inexchange rates could cause difficulties for financialinstitutions because of currency mismatches be-tween bank assets and liabilities.36 Past experiencehas shown that rising international interest rates in-crease the vulnerability of emerging markets (andtheir financial systems) in three ways: through theasset substitution channel (capital outflows),through an adverse impact on the creditworthinessof emerging market borrowers, and through an ex-acerbation of information problems in credit mar-kets (e.g., adverse selection). On the other hand, de-clining international interest rates promote capitalinflows that could contribute to risky lendingbooms. Moreover, volatile domestic and interna-tional interest rates could have damaging effects onthe financial system both directly—if banks cannotavoid taking interest rate risk—and indirectlythrough a deterioration of credit quality—if bankscan shift interest rate risk to their customers.

Level of Domestic Real Interest Rates.Unless theeconomy has high growth rates, financial institu-tions tend to be stressed under high real interestrates. Increasing real interest rates contribute tohigher nonperforming loans. On the other hand,persistent negative real interest rates could signaldistortions in the financial system created by thegovernment fixing of nominal interest rates (i.e., fi-nancial repression).

Exchange Rate Sustainability. A large real appre-ciation could weaken the export sector’s capacity to service debt. On the other hand, a large devalua-tion could improve the capacity of the export sectorto service its debt but, at the same time, it couldweaken the debt-service capacity of non-export-related domestic borrowers. Moreover, largechanges in the exchange rate could put pressure onthe financial system either directly by changing

asset values, or indirectly via possible effects onthe real economy.

Exchange Rate Guarantees.The existence of im-plicit or explicit exchange rate guarantees and incon-sistencies of monetary and exchange rate policiesare major contributors to volatility in capital flowsand excessive foreign currency exposures.

Lending and Asset Price Booms

Lending Booms(rapid growth of the ratio of bankcredit to GDP). Such booms have preceded severe fi-nancial crises. Rapid expansion in lending by finan-cial institutions often occurs because of poor analy-sis of the quality of loan applications. In addition, aweak regulatory environment, including the pres-ence of implicit or explicit public sector guarantees,could encourage excessive risk taking by individualfinancial institutions and contribute to risky creditexpansions. Mortgage and other consumer lendingand foreign currency loans have preceded recentlending booms, particularly in emerging marketeconomies.

Asset Price Booms.Expansionary monetary poli-cies, among other reasons, could contribute to ex-cessive booms in the stock and real estate markets.A subsequent tightening of these policies has oftenled to large reductions in the value of stock and realestate and a downturn in economic activity, creatingconditions for financial distress. Also, a capital mar-ket slump normally reduces financial institutions’income and the value of investment portfolios andcollateral.

Contagion Effects

Since a country’s financial system is linked toother countries’systems through capital marketflows and bilateral trade, the occurrence of financialcrises in other countries could trigger a financial cri-sis or distress at the domestic level.

Trade Spillovers.When a country experiences a fi-nancial crisis marked by a significant depreciation ofits currency, other countries may suffer from tradespillovers owing to the improved price competitive-ness of the crisis country.

Financial Markets Correlation. Contagion risk ishigher for countries that have similar macroeco-nomic characteristics or close financial links (suchas through commercial banks, capital market flows)with the country in crisis. In particular, correlationbetween stock market prices, exchange rates, and in-terest rates in different countries is often seen as anindicator of the risk of contagion.37

11

36Under fixed exchange rate regimes, by definition, volatilitycannot be observed before a devaluation actually occurs, so otherindicators have to be used, for example, the volume of foreign ex-change intervention by the central bank.

37For a summary of financial contagion effects, see Interna-tional Monetary Fund (1999b), p. 66.


Other Factors

Directed Lending and Investment.Portfolio re-strictions channeling credit to specific activities orsectors based on nonmarket criteria often lead to theinefficient allocation of resources and negatively af-fect the solvency of financial institutions.

Government Recourse to the Banking System andOther Quasi-Fiscal Imbalances.For example, a sud-den increase in central bank credit to the governmentcould lead to inflationary pressures and affect the fi-nancial system.

Arrears in the Economy. The buildup of arrearscould signal debt-service difficulties by the govern-ment or by private sector borrowers. These problemsnegatively affect the solvency and liquidity of finan-cial institutions.

Directions for Further Work

The set of indicators identified so far for conduct-ing macroprudential analysis is already large andwill potentially increase as a result of the additionalresearch needed in this area. In particular, the con-clusions of the September 1999 consultative meetingpointed to the need for better indicators of develop-ments in specific sectors and markets that haveproven relevant in assessing financial vulnerabilities,but that have been difficult to gauge in practice.These sectors and markets include real estate, thecorporate and household sectors, nonbank financialinstitutions, and off-balance sheet exposures of fi-nancial institutions, including institutional investors(e.g., mutual funds, pension funds, insurance compa-nies, and hedge funds).

In parallel with the development of more compre-hensive indicators, work should also be done on se-lecting a smaller and more manageable subset ofMPIs, notably for the purposes of periodic monitoringand data dissemination. Indicators included in such asubset, or core set, of MPIs would need to be focusedon core markets and institutions, based on acceptedanalytical relationships, comparable across countriesand relevant in most circumstances (i.e., not country-specific), among other things, to permit cross-countrystudies.38 Participants at the consultative meeting con-cluded that the research conducted so far has not pro-duced a consensus on the composition of such a coreset of indicators. A variety of different indicators ap-pear to be relevant in different countries under differ-ent circumstances. Moreover, potential vulnerabilitiesmay be exacerbated by country- or region-specific cir-cumstances (including inadequate legal and financialinfrastructure to absorb shocks), which a core set ofquantitative indicators may not detect.

Participants at the consultative meeting also dis-cussed the possibility of developing a composite in-dicator of financial system soundness. There was ageneral sense, however, that the complex reality offinancial markets may not lend itself to being cap-tured in such indicators. In particular, composite in-dicators could prove simplistic and potentially mis-leading, as they may conceal or misrepresentproblems by offsetting positive and negative signalsfrom different individual components.

12

38Particularly among academics, investment managers, and an-alysts participating in the consultative meeting, interest was highin indicators that would permit cross-country studies, that is, indi-cators that are suitable for comparative analysis.

This chapter reviews the theoretical and empiri-cal literature, other than work done by the

IMF,39 which would support the selection of a coreset of MPIs. In general, these studies look at the fea-tures of crisis-prone systems, with a view to antici-pating future crisis events. By attempting to identifyleading indicators of crises, rather than contempora-neous indicators of financial soundness, much of theearlier literature did not specifically review the fullrange of potential MPIs. More recently, the focus ofmany studies has shifted toward contemporaneousindicators of financial health. No consensus has yetemerged, however, from this body of work on a setof indicators that is most relevant to assessing finan-cial soundness, or to building effective early warn-ing systems. The statistical significance of individ-ual indicators is often found not to be strong, andsome of the studies have produced conflicting re-sults. This may be due to differences among crises,so that specific indicators may be more or less rele-vant to each case.

We present below the following:

• A brief survey of the theoretical literature on theorigins of financial crises. These theoreticalstudies are not used to derive MPIs directly, butthey underpin the empirical studies discussed insubsequent sections.

• A review of empirical evidence on macroeco-nomic factors that affect the health of the finan-cial system. This literature has focused on lead-ing indicators of financial crises.

• A review of empirical evidence on prudentialfactors used to assess financial soundness. Thesestudies suggest additional variables that can beused as contemporaneous indicators.

The literature provides some empirical justifica-tion for the use of most of the variables that havebeen identified as macroeconomic and prudential in-dicators of financial soundness. The variety of speci-fications, time periods, and objectives of the empiri-cal studies, however, makes it difficult to prioritize

the indicators, or to eliminate any of them on thebasis of empirical evidence. The empirical resultsrepresent work in progress, and serve only to con-firm the potential usefulness of the indicators.

Determinants of Financial System Soundness40

Over the years, researchers have developed a vari-ety of economic theories to explain soundness in fi-nancial markets. While earlier researchers relied onmovements in economic fundamentals as the originof financial distress and crisis, recent studies havehighlighted the role of the information available to,and the expectations of, investors in explaining thebehavior of financial markets. The rest of this sec-tion reviews the literature on banking soundness, be-cause historically, banks have been the most impor-tant financial intermediaries.

The classic explanation for financial fragility isgiven by Irving Fisher (1933). He argues that fragilityis closely correlated with macroeconomic cycles, andhighlights, in particular, debt liquidation. A downturntriggered by over-indebtedness in the real economyrequires, at some point, liquidation of this debt inorder to bring the economy back to equilibrium. Debtliquidation would result in a contraction of monetaryliabilities and a slowdown of velocity. These changeshave several economic implications—reductions inprices, output, and market confidence, and increasesin bankruptcies and unemployment. According toFisher, therefore, financial fragility is largely based ondeterioration in economic fundamentals.

Other theories highlight factors affecting depositorconfidence. Diamond and Dybvig (1983) discuss thepotential existence of multiple equilibria in financialmarkets. Banks offer a mechanism of maturity trans-formation whereby deposits are often lent withlonger maturities. It is possible that the “good” equi-librium prevailing in normal times is not the only

III Literature Survey

13

39Work conducted by the IMF is the subject of Section V.

40Davis (1999) provided a very useful reference in writing thissection.

III LITERATURE SURVEY

equilibrium, and that the banking sector finds itself ina “bank run” equilibrium. Diamond and Dybvig as-sume that these equilibria are a function of randomevents known to all agents. Therefore, a bank run oc-curs when agents have deposited funds into a bank ata time of low probability of a bank run, and then laterobserve negative events that increase their anticipa-tion of a bank run. This study points to the impor-tance of a high level of confidence in banks as asource of banking sector stability.41

Some studies focus on information issues.Mishkin (1996) stresses that information asymme-tries between creditors and borrowers result in an ad-verse selection problem.42 Borrowers often havemore information than banks on the quality of the in-vestment they wish to finance. Creditors insurethemselves against this source of uncertainty bylending only at the average rate between nonriskyand risky investments. It follows that borrowers withhigh-quality investments (i.e., high-return invest-ments with low risk) pay interest rates that are higherthan in the absence of asymmetric information,while those with low-quality investments pay lowerrates. This can lead to a situation where high-qualityinvestments are displaced by low-quality invest-ments, causing deterioration in the overall quality ofbank portfolios.

Guttentag and Herring (1984) extend the argu-ment on asymmetric information to the possiblepractice of credit rationing. In the presence of uncer-tainty about the true return on investment, there maybe a discrepancy between return expectations on thepart of creditor and borrower. When the creditor’sexpected return on a project is less than the return onhis alternative use of funds, the borrower may be ra-tioned. Their argument suggests that credit rationingincreases with the level of uncertainty, and thus of fi-nancial vulnerability. The introduction of a depositinsurance mechanism is often seen as one way tolessen this problem. But Keeley (1990) points outthat the possible existence of moral hazard problemsin a deposit insurance scheme can lead financial in-stitutions to take more risks than they would other-wise do—borrowing at the risk-free rate (i.e., therate on the insured deposits) and investing in riskierassets.

Recent studies point to the existence of asymmet-ric information in financial markets as a source ofcontagion of financial crises from one country to an-other. This is a vital concern because more countrieshave liberalized their markets and are now highly

linked with other countries’markets. Through thischannel, negative external shocks may be directlytransmitted to countries that are healthy. Kodres andPritsker (1998), for example, develop a theoretical,multiple-asset, rational expectations model of thedeterminants of contagion, in which adverse effectsof contagion depend on the sensitivity of the affectedcountry to common macroeconomic risks and to thelevel of asymmetric information prevailing in theeconomy. They also point out that in the presence ofhedging mechanisms, contagion may occur withoutcommon macroeconomic risks in two countries if in-vestors hedge by reducing their overall exposure toemerging markets. This seems to have been the ex-perience of many Asian countries.

Davis (1996) argues that institutional investorsmay contribute to financial fragility because of prin-cipal-agent problems in the relationship betweenfund managers and their clients—that is, fund man-agers may not act to maximize the client’s profitswithout appropriate supervision. He argues that oneway to reduce the principal-agent problem is to in-troduce more frequent monitoring and performanceevaluation systems. From the perspective of the fundmanager, when strict monitoring and evaluation arein place, one way to show the quality of his manage-ment is to imitate others (the so-called herd behav-ior) rather than trust his own judgment, since the ini-tial financial asset information available to him oftencontains elements of uncertainty. In this way then,Scharfstein and Stein (1990) says mimicking otherinvestors is likely to maintain the manager’s reputa-tion by reducing his risk of underperformance rela-tive to the average for the market. Therefore, anevent perceived as adverse by just one investor mayresult in large movements in financial asset prices.

These theoretical studies have been the point ofdeparture for much of the empirical work discussedbelow. Table 2 summarizes the main indicators iden-tif ied in the empirical literature.

Studies of Macroeconomic Variables

Several studies of financial problems appeared inthe wake of the Mexican crisis in 1994, and beforethe emergence of the Asian crisis in 1997.43 These

14

41In addition, when a bank run occurs, the institution tries torapidly liquidate assets to meet demand for deposit withdrawal. Inthese circumstances, assets are likely to be sold at a discount andthe financial position of the bank may deteriorate further.

42For a discussion of adverse selection, see Akerlof (1970).

43The range of approaches is illustrated by Demirgüç-Kunt andDetragiache (1998), an econometric study of banking crises in 65countries; Federal Reserve Bank of Kansas City (1997), confer-ence proceedings covering a wide range of issues beyond quanti-tative ones, but offering comments on the importance of macro-economic variables; Gavin and Hausmann (1996) andRojas-Suárez and Weisbrod (1995), both on Latin America; Gold-stein and Turner (1996), which contains a comprehensive surveyof possible origins of banking crises in emerging countries; andGoodhart (1995), which focuses on asset market volatility.

Studies of Macroeconomic Variables

15

Tabl

e 2.

Mac

ropr

uden

tial

Ind

icat

ors

in a

Sel

ecti

on

of R

ecen

t S

tudi

es

Stud

ies

by a

utho

rs1

C-K

F-R

S-T-

VH

GH

-P-B

B-G

B-P

E-L

E-R

FH

-PK

K-L

-RR

-SD

K-D

Year

of p

ublic

atio

n19

9619

9619

9619

9719

9719

9919

9919

9819

9819

9819

9819

9919

9819

9819

98Fo

cus

of s

tudy

BC

CB

BC

CC

BC

BB

& C

CC

BB

= b

anki

ng c

risi

sC

= c

urre

ncy

cris

is

Agg

rega

ted

mic

ropr

uden

tial i

ndic

ator

sFo

reig

n ex

chan

ge e

xpos

ure

••

••

••

•Se

ctor

al c

redi

t co

ncen

trat

ion

••

Non

perf

orm

ing

loan

s•

Agg

rega

te r

isk-

base

d ca

pita

l rat

io•

Cen

tral

ban

k cr

edits

to

finan

cial

inst

itutio

ns2

••

Segm

enta

tion

••

••

Rat

io o

f dep

osits

to

M2

(or

GD

P)•

••

••

Stoc

k ex

chan

ge p

rice

s•

••

•A

ggre

gate

ave

rage

ret

urns

•

Mac

roec

onom

ic in

dica

tors

Lend

ing

boom

s (e

.g.,

cred

it/G

DP)

••

••

••

••

••

••

Ass

et p

rice

boo

ms

••

••

••

••

Con

tagi

on e

ffect

s•

Exte

rnal

def

icits

••

••

••

••

••

Agg

rega

ted

grow

th r

ate

••

••

••

••

•Vo

latil

ity o

f int

eres

t an

d ex

chan

ge r

ates

••

••

Term

s of

tra

de•

••

••

•Le

vel o

f dom

estic

inte

rest

rat

es•

••

••

••

••

Exch

ange

rat

e m

isal

ignm

ents

••

••

••

••

••

••

Gov

ernm

ent

reco

urse

to

bank

ing

syst

em•

Vola

tility

in in

flatio

n•

•

1 Cap

rio

and

Klin

gebi

el (

C-K

);Fr

anke

l and

Ros

e (F

-R);

Sach

s,To

rnel

l,an

d Ve

lasc

o (S

-T-V

);H

onoh

an (

H);

Gon

zále

z-H

erm

osill

o,Pa

zarb

as ,io

gv lu,a

nd B

illin

gs (

GH

-P-B

);Ba

ig a

nd G

oldf

ajn

(B-G

);Be

rg a

nd P

atill

o (B

-P);

Esqu

ivel

and

Lar

raín

(E-

L);E

iche

ngre

en a

nd R

ose

(E-R

);Fr

atzs

cher

(F)

;Har

dy a

nd P

azar

bas ,io

gv lu (

H-P

);K

amin

sky

(K);

Kam

insk

y,Li

zond

o,an

d R

einh

art

(K-L

-R);

Rad

elet

and

Sac

hs (

R-S

);an

d D

emir

güç-

Kun

tan

d D

etra

giac

he (

DK

-D).

2 As

a pr

opor

tion

of t

heir

cap

ital o

r lia

bilit

ies.

III LITERATURE SURVEY

studies investigate the vulnerability of financial in-stitutions in the face of exogenous shocks. Financialintermediaries are generally highly leveraged, en-gage in maturity transformation, transact in marketswith asymmetric information, and are subject tomoral hazard through explicit or implicit deposit in-surance. Sources of financial fragility explored inthe studies include a falling growth rate, deteriora-tion in the balance of payments, high inflation,volatile exchange rates, surges in stock market activ-ity and prices, credit booms, weakening perfor-mance of export sectors, and deterioration in theterms of trade. In addition, these studies highlightnonquantifiable indicators of financial fragility, suchas deficient banking supervision, inadequate instru-ments of monetary control, overly generous depositinsurance, inadequacies in the operation of the legalsystem, overexposure in international financial mar-kets, lack of adequate accounting standards andpractices, insufficient financial disclosure, and per-verse incentive structures.

The Asian crisis has provoked a new wave of fi-nancial sector studies, which confirm that macroeco-nomic shocks to output, exports, prices and theterms of trade, asset price booms, and inappropriatemonetary and exchange rate policies, all result in fi-nancial pressure and contribute to crises in financialsystems that are inherently fragile.44 In addition, thisresearch points to the destabilizing effects of marketoverreaction, the feedback effects of crises thatweaken corporate balance sheets, and the impact ofunexpected shocks, such as the rapid change in theyen-dollar exchange rate and the swift emergence ofnew competition from Mexico (in the wake of thedeep devaluation in 1994–95) and from China.These are important factors that must be evaluated incases of economic instability. Except for the impactof third-party exchange rate changes on the domesticeconomy, however, these developments generally donot appear in advance of financial weakening and,therefore, do not offer additional early warning indi-cators of financial health.

The contagion of financial crises from one coun-try to another has been the focus of several empiricalstudies.45 The factors that appear to expose a coun-try’s financial system to contagion include close cor-relation in the past behavior of currency and equitymarkets, export and import ties (or competition intrade), cross-market banking links, low levels of for-eign reserves, the extent of exchange rate overvalua-

tion, and the inherent weakness of the financial sys-tem. In addition, Kaminsky and Reinhart (1998) findevidence that sharing a common creditor with a cri-sis country creates a high risk of contagion.

Given that currency and financial crises oftenoccur simultaneously, the factors underlying cur-rency crises have the potential to contribute to an as-sessment of the health of financial institutions. Cau-sation between exchange rates and financialvariables, however, may go in either or both direc-tions. This relationship has been the subject of sev-eral studies, including Dornbusch, Goldfajn, andValdés (1995); Kaminsky and Reinhart (1999);Kaminsky (1999); and Kaminsky, Lizondo, andReinhart (1998). Their results suggest that exchangerate crises provoke financial crises when the bankingsector is vulnerable, that is, when the impact of a de-valuation on the quality of bank assets is largeenough to wipe out the banks’net worth. Therefore,simulations (stress tests) of the impact of a devalua-tion of various magnitudes on banks’capital ade-quacy can be useful as an additional indicator of fi-nancial robustness. Kaminsky and Reinhart (1999),however, point out that in about half of the crises inthe 1980s and early 1990s that they examined, finan-cial crises preceded currency crises.

Studies of Aggregated Microprudential Indicators

Much of the earlier literature on aggregated mi-croprudential indicators follows the categorizationof the CAMELS rating—see Altman (1968), Sinkey(1978), and Thomson (1991). This portfolio-basedassessment method is broadly consistent with the listof MPIs identified in Section II. These variables areused in empirical research less frequently thanmacroeconomic indicators, due to the availability ofhigher frequency data for the latter. A classic studyby Altman (1968) uses the so-called Z-score model,which is based on several financial ratios capturingasset quality, earnings performance, and liquidity,but this analysis is at the level of the individualfirm.46

More recent literature—including Frankel andRose (1996); Sachs, Tornell, and Velasco (1996); andHonohan (1997)—emphasizes the important role offoreign borrowing, particularly short-term liabilitiesdenominated in foreign currency, to measure the de-gree of exposure to currency and inflation risks. Re-

16

44These studies include Berg and Patillo (1999), Bussière andMulder (1999), Corsetti, Pesenti, and Roubini (1998), Furmanand Stiglitz (1998), International Monetary Fund (1998), Kawai(1998), Kwack (1998), and Radelet and Sachs (1998).

45Baig and Goldfajn (1999), Fratzscher (1998), Glick and Rose(1998), and Sachs, Tornell, and Velasco (1996).

46The Z-score model uses the linear discriminant analysismethod to identify healthy and unhealthy firms, and “Z” repre-sents the composite score used to distinguish between these twogroups of firms.

Studies of Aggregated Microprudential Indicators

cent literature also focuses on the level of nonper-forming loans—such as González-Hermosillo,Pazarbas,iogvlu and Billings (1997). González-Hermosillo (1999) shows empirical evidence that theCAMELS-type assessment is statistically significantonly if nonperforming loans and capital adequacyare simultaneously considered.47 This is consistentwith theoretical explanations for the eruption of theAsian financial crisis, which posit financial institu-tions’weaknesses as a major cause of the crisis.

Other indicators to capture financial vulnerabilityinclude a measure of segmentation (often proxied byan interbank interest rate differential), the deposits-to-M2 ratio, and aggregate stock indices. In survey-ing literature on these indicators, Demirgüç-Kuntand Detragiache (1999) point to criticisms on the useof CAMELS-based criteria to measure bankstrength. A comprehensive study by Kaminsky, Li-zondo, and Reinhart (1998) concludes that these in-dicators are less able to explain currency crises thanis exchange rate misalignment.

Many of these studies use logit/probit models tocapture banking fragility or to differentiate healthybanks from unhealthy banks. But their ability to de-tect future events in the out-of-sample forecastingcontext is limited. Lane, Looney, and Wansley (1986)and Whalen (1991) use the Cox proportional hazardsmodel, which is capable of providing information onthe expected time of failure, but the overall conclu-sion on the poor performance of the CAMELS-typemodel remains unchanged. Consequently, González-Hermosillo (1999) combines both micro- and macro-factors in explaining banking fragility, and concludesthat the introduction of macroeconomic variables sig-

nificantly improves the explanatory power of modelsbased on microprudential indicators only.

Indicators of capital adequacy provide importantinformation about financial fragility. Minimum stan-dards for risk-weighted capital adequacy have beenagreed to by the Basel Committee on Banking Su-pervision, but there remain reservations about thesestandards, which are currently under review.48 Theliterature points to some of these limitations. For ex-ample, the Basel Committee on Banking Supervi-sion (BIS (1999b)) shows that the improvement inthe ratio for Group of Ten (G–10) countries from 9.3in 1988 to 11.2 percent in 1996 did not reflect a sig-nificant improvement in the overall health of the sys-tem.49 Proposals currently under discussion at theBasel Committee would supplement capital ade-quacy measures with supervisory reviews that couldrequire higher levels of capitalization and use differ-ent measures of risk exposure, such as the increas-ingly popular Value-at-Risk (VaR) models.50

17

48The Basel Committee’s recent recommendations on capitaladequacy (still in the form of a discussion draft) can be found inBIS (1999a).

49BIS (1999b) also points out that the way in which undercapi-talized banks meet the minimum capital requirement depends onindividual cases. Ediz, Michael, and Perraudin (1998) show thatbanks in the United Kingdom tend to raise their Tier 2 capital first,followed by their Tier 1 capital.

50Value-at-Risk is an estimate of the maximum loss on a port-folio with a given (small) probability over a preset horizon. TheVaR methodology uses a standard statistical technique usuallybased on the historical volatility and correlation of portfolio re-turns to measure market risk (not credit risk)—see Hendricks(1996), and Dimson and Marsh (1997). While the incorporationof VaR models into capital adequacy regulation could permit amore accurate estimation of risk, it should be noted that there aredrawbacks to VaR models. In particular, these models are unableto account for shocks that depart considerably from past experi-ence (e.g., a large devaluation).

47This study includes a recent survey of empirical studies onbanking failures.

In response to the recent crises, many institutionshave initiated or intensified work on developing

macroprudential indicators and macroprudentialanalysis capabilities. A selection of these efforts issummarized in this section. The statistical frame-works that are in place in some of the institutions arediscussed in Appendix I.

International and MultilateralInstitutions

European Central Bank

Upon the request of the European Central Bank(ECB) Banking Supervision Committee (BSC), theECB’s Working Group on Financial Fragility has car-ried out preliminary work on MPIs. This workinggroup separated potential indicators into three cate-gories: (1) systemic indicators of the health of thebanking system, (2) macroeconomic factors that in-fluence the banking system, and (3) contagion factors.With regard to the first category, it proposed the mon-itoring of the following variables: lending behavior,competitive conditions, liquidity situation, exposureconcentrations, asset quality, profitability, capitalbuffers, and market assessments. The macroeconomicindicators suggested as factors that influence thebanking system were income development, leverage(financial fragility), debt burden, asset prices, mone-tary conditions, and the external position.51

The BSC has recently established the WorkingGroup on Macroprudential Analysis. The mandate ofthe working group is to develop a framework formacroprudential analysis following the approaches ofseveral Scandinavian countries (explained below).

Based on this framework, the group is to draft a reporton European Union (EU)-wide MPIs that would serveas a basis for BSC discussion of the soundness andcharacteristics of the EU banking systems. The frame-work would draw on economic statistics as well as su-pervisory insights. Currently there are no plans tomake the results of the exercise public. The BSC alsooperates the Cooperative Forum on Early WarningSystems, with voluntary participation by interestedEU member states. Even though these projects are fa-cilitated by the relatively high degree of harmoniza-tion of standards within the EU, comparability of in-dicators among member states is complicated by,among other factors, remaining differences in ac-counting and provisioning norms.

World Bank

The World Bank has been conducting FinancialSector Assessments (FSAs) for a number of coun-tries where it is involved in financial sector work andwhere financial sector issues are considered particu-larly relevant to the World Bank’s lending decisions.These assessments are focused on both stability anddevelopmental aspects of the financial sector andcontain an analysis of a range of MPIs, both macro-economic indicators as well as banking indicators.The latter are drawn primarily from public data(published accounts). In order to avoid overlap andduplication of work in this area, the IMF and theWorld Bank have recently established a joint frame-work for comprehensive assessments of financialsectors, the FSAP.52

In addition to the more qualitative FSAs, theCountry Credit Risk Department of the World Bankuses various risk-rating models in order to determinethe likelihood of default of its borrowers. Thesemodels include a checklist model that is more com-plex but otherwise very similar to the ones used byrating agencies and investment banks. The mainareas covered by this model are (1) structural andmacroeconomic indicators of economic performance

IV Work Programs of Other Institutions

18

51These indicators are to be analyzed in their relation to the fol-lowing intermediate risk targets: aggregate credit risk, interestrate risk, equity price risk, real estate risk, foreign exchange risk,and liquidity risk. The ultimate target variable is the aggregatesolvency of the banking system. The analysis under the first cate-gory would draw on confidential supervisory data, whereas theanalysis under the second would be based largely on public infor-mation as well as on available macroeconomic forecasts. 52See Section V for details.

International and Multilateral Institutions

and external vulnerability, (2) external debt and itssustainability, (3) political risk and policy perfor-mance, and (4) World Bank exposure and history ofdebt service to the Bank.

The Country Credit Risk Department is also thesecretariat of the Short-Term Risk MonitoringGroup of the World Bank. This group is responsiblefor assessing and monitoring countries that are vul-nerable to political, economic, or financial crises inthe near term. The group reports to the senior man-agement of the World Bank on both vulnerablecountries and trends in the global economy and fi-nancial markets on a monthly basis. A country’s vulnerability to domestic or exogenous shocks is as-sessed using macroeconomic and policy perfor-mance indicators (including indices developed bysome of the World Bank’s regional departments), fi-nancial market indicators, and qualitative assess-ments provided by operational and central units, in-cluding financial sector specialists. As the secretariatof the group, the Country Credit Risk Department isalso responsible for assisting operational units inpreparing contingency plans for countries that areconsidered highly vulnerable. These contingencyplans are intended to ensure a broad understandingof the situation in the country concerned and to facil-itate a discussion by senior management of theWorld Bank’s response.

G–7 and G–10 Initiatives

The G–7 endorsed, in February 1999, a proposalby Hans Tietmeyer, former president of the Bundes-bank, to establish the Financial Stability Forum(FSF). The forum comprises representatives fromthe G–7 countries plus Australia, Hong Kong SAR,the Netherlands, Singapore, the IMF, the WorldBank, the Bank for International Settlements (BIS),the OECD, the Basel Committee on Banking Super-vision, the International Organization of SecuritiesCommissions (IOSCO), the International Associa-tion of Insurance Supervisors (IAIS), the Committeeon Payment and Settlement Systems (CPSS), and theCommittee on the Global Financial System (CGFS).Part of the mandate of the FSF is to strengthen themonitoring and assessment of systemic vulnerabili-ties. The FSF has established three working groupsdealing with highly leveraged institutions, capitalflows, and offshore financial centers, respectively. Inaddition, a task force on implementation of stan-dards and study groups on deposit insurance and in-surance issues have been set up. A study paper on In-ternet and electronic trading in financial markets hasalso been commissioned. It appears that the FSF willprimarily focus on specific areas of systemic vulner-ability of the financial system and not on the contin-uous monitoring of large sets of indicators.

The CGFS has a mandate from the governors ofthe central banks of the G–10 member countries. Itacts as a central bank forum for the monitoring andexamination of broad issues relating to financialmarkets and systems with a view to elaborating ap-propriate policy recommendations to support thecentral banks in the fulfillment of their responsibili-ties for monetary and financial stability. The tasksperformed by the CGFS fall into three categories:systematic short-term monitoring of global financialsystem conditions, so as to identify potential sourcesof stress; in-depth, longer-term analysis of the func-tioning of financial markets; and the articulation ofpolicy recommendations aimed at improving marketfunctioning and promoting stability. The mandaterecognizes that the causes of financial instability canarise from both the behavior of markets and the com-plex interrelationships that exist between institu-tions, markets, infrastructures, and macroeconomicpolicy.

Non-G–10 central banks now routinely attendmeetings of the CGFS and participate in its workinggroups. In recent public reports, the committee hasfocused on the nature and use of information avail-able for banks’country risk assessments; it has iden-tif ied general principles and more specific policyrecommendations for the promotion of liquid gov-ernment securities markets; and it has examined theevents surrounding the financial turbulence in manyinternational markets in autumn 1998.53 One of itsworking groups has formulated a set of public dis-closure guidelines to provide a meaningful basis forcomparing levels and types of risk across institutionsand across countries. This work, in turn, has been in-corporated in a multidisciplinary effort comprisingrepresentatives from the banking, securities, and in-surance regulators. A pilot effort involving privatesector firms is to be conducted.

The G–10 Working Party on Financial Stability inEmerging Market Economies published a report onfinancial stability in emerging markets in April1997. This report identifies both macroeconomicsources of vulnerability (instability, inflation, liber-alization, and failures in the design of macroeco-nomic policy instruments), as well as sector-specificsources of vulnerability (corporate governance andmanagement, market infrastructure and discipline,and supervision and regulation). The report also con-tains a list of indicators of robust financial systems.These indicators are categorized under the followingsix groupings: (1) legal and judicial framework; (2)accounting, disclosure, and transparency; (3) stake-holder oversight and institutional governance; (4) market structure; (5) supervisory and regulatory

19

53BIS, Committee on the Global Financial System (1999).

IV WORK PROGRAMS OF OTHER INSTITUTIONS

authority; and (6) design of a safety net. Under thesecategories, the report lists areas of importance, butnot specific indicators. The areas mentioned arebroadly equivalent to those identified by other insti-tutions, including the IMF.

National Central Banks andSupervisory Agencies

Until recently, relatively few countries paid in-depth attention to macroprudential analysis at the na-tional level, although national central banks and su-pervisory agencies in many countries have longmonitored and reported on issues relating to financialsystem stability. While much of this work has amacroprudential aspect, it is not necessarily carriedout with a formal framework for using MPIs to assessfinancial system soundness.54 The complex nature ofthe analysis and the need for high-quality data on in-dividual banks, collected and stored in a manner con-ducive to aggregation and analysis, has meant thatmuch of the existing MPI-related work has been car-ried out relatively recently and mainly, but not exclu-sively, in the industrial countries. Within this group,the most developed approaches tend to be those bycountries that have had a major financial sector crisisin recent years. Work is, however, currently underway in a number of countries to develop macropru-dential data collection and analysis frameworks.

Following is a selective country-by-country sum-mary of work that has been carried out, both with aspecific focus on identifying MPIs, and with otherfocuses that may, nevertheless, provide a guide topossible MPIs. Table 3 gives a comparative listing ofindicators used in a few selected countries. The list-ing is certainly not comprehensive—neither with re-spect to the countries nor the indicators being used.Rather, our intent is to provide a sketch of the differ-ent types of approaches and key indicators.

As a general observation, to the extent that spe-cific MPIs are identified for a given country, theytend to be aggregated microprudential indicatorsrather than macroeconomic indicators. This may re-flect the fact that while it is generally possible toidentify the ex post macroeconomic variables thathave contributed to systemic problems, it is difficult,

ex ante, to predict the macroeconomic developmentsthat may trigger problems, or their precise timing.Nonetheless, it is clear that macroeconomic indica-tors play a central role in almost all macroprudentialanalysis frameworks. As a further observation, whilethe indicators monitored in different countries arebroadly similar, where differences do arise theyoften reflect country-specific characteristics of pastfinancial sector problems.

China

As with many other countries, the Chinese author-ities are in the early stages of building a frameworkfor macroprudential analysis. A range of indicatorsto be collected and monitored has been identified.Macroeconomic variables include GDP, inflation,monetary aggregates, the current account balance,external debt, international reserves, and the ex-change rate. Aggregated microprudential indicators,building on the off-site supervision of commercialbanks, include indicators of general performance(trends in total assets, loans, and deposits); safety in-dicators (capital adequacy, asset quality, and creditconcentration); liquidity indicators (liquidity of as-sets, excess reserves, and liquidity of domestic andforeign currency liabilities); earnings indicators (re-turn on equity and assets); and overall control indi-cators (loan to deposit ratios, proportion of interbankfinancing, and proportion of offshore financing).

The People’s Bank of China is currently workingto further develop a framework for collecting and an-alyzing MPIs. The ongoing transition to a marketeconomy means that parts of the financial infrastruc-ture—such as accounting and auditing practices—are still being developed. In addition, prudential dataquality may vary across different regions of thecountry. Improvements are under way in the area ofregulation and supervision of financial institutions,including upgrading loan classification standards,and strengthening the supervisory capabilities of thecentral bank, which will enhance the quality ofmacroprudential analyses.

Finland

The Bank of Finland is one of the few centralbanks to have a framework for forecasting bankingsector developments. In addition, the Bank of Fin-land is actively participating in the work being un-dertaken by the ECB in developing a system ofMPIs. The banking forecast framework was devel-oped in the wake of Finland’s banking sector crisisin the early 1990s, with a view to assisting in its res-olution by analyzing likely developments in thebanking sector. The framework produces a forecastof trends in banking system profitability over about a

20

54As an example, the Reserve Bank of India prepares an annualreport on banking trends in India in terms of a statutory require-ment under the Banking Regulation Act of 1949. The report cov-ers developments in banking policy, cooperative banking, banksand nonbanking institutions, and provides some information onMPIs, including financial ratios, off-balance sheet exposures,nonperforming loans, and profitability. The Reserve Bank ofIndia, however, does not report the use of these MPIs in any for-mal framework covering systemic soundness.

National Central Banks and Supervisory Agencies

21

Tabl

e 3.

Co

mpa

rati

ve L

isti

ng o

f Ind

icat

ors

Use

d by

Sel

ecte

d C

oun

try

Aut

hori

ties

Uni

ted

Stat

es__

____

____

____

____

____

____

____

____

____

____

____

____

____

____

____

__Fe

dera

l Res

erve

Fede

ral D

epos

it In

sura

nce

Fina

ncia

l Ins

titut

ions

Cor

pora

tion

Bank

of F

inla

nd1

Nor

ges

Bank

2Sv

erig

es R

iksb

ank

Mon

itori

ng S

yste

mG

row

th M

anag

emen

t Sy

stem

Bank

ing

sect

orA

sses

smen

ts o

f ban

k C

apita

l/ass

ets

ratio

sPr

ofita

bilit

y tr

ends

Tang

ible

cap

ital/a

sset

s ra

tioEq

uity

/ass

ets

ratio

vari

able

sex

pens

esPr

ofita

bilit

y tr

ends

Ret

urn

on a

sset

sLo

ans

past

due

,30–

89 d

ays

Ass

et g

row

thA

sses

smen

ts o

f loa

n R

etur

n on

ass

ets

Gro

ss in

tere

st m

argi

nsLo

ans

past

due

,90

days

or

mor

eLo

an g

row

thlo

sses

and

oth

er

Gro

ss in

tere

st m

argi

nsSp

read

bet

wee

n sh

ort-

N

onac

crua

l loa

nsVo

latil

e lia

bilit

ies/

asse

tsw

rite

-offs

Spre

ad b

etw

een

shor

t- a

nd

and

long

-ter

m in

tere

st

Fore

clos

ed r

eal e

stat

eLo

ans/

asse

ts r

atio

long

-ter

m in

tere

st r

ates

rate

sN

et in

com

e/as

sets

rat

ioPo

rtfo

lio c

once

ntra

tion4

Ope

ratin

g co

sts

tren

dsO

pera

ting

cost

s tr

ends

Res

erve

sG

row

th in

por

tfolio

con

cent

ratio

n4D

epos

its/lo

ans

ratio

sBa

nks’

sha

re p

rice

sIn

vest

men

t se

curi

ties/

asse

ts r

atio

Equi

ty/a

sset

s gr

owth

4

Dep

osit,

loan

gro

wth

rat

esBa

nkru

ptcy

tre

nds

UBS

S as

set

grow

th p

erce

ntile

sco

re3

Loan

s/as

sets

gro

wth

4

Tren

ds in

ris

kine

ss o

f Tr

ends

in p

ast

due

loan

s U

BSS

com

posi

te p

erce

ntile

sco

re3

Vola

tile

liabi

litie

s/as

sets

gro

wth

4,5

corp

orat

e de

btby

sec

tor

Prio

r m

anag

emen

t ra

ting

Sect

oral

deb

t an

d de

bt-

Sect

oral

deb

t an

d de

bt-

Prio

r co

mpo

site

CA

MEL

rat

ing

serv

icin

g ra

tios

serv

icin

g ra

tios

Fore

cast

pas

t-du

e lo

ans

Cou

nter

part

ies’

ris

k Tr

ends

in b

ank

finan

cing

ex

posu

refr

om o

ther

sou

rces

Mac

roec

onom

ic

Mar

ket

inte

rest

and

Im

pact

of i

nter

est

rate

Le

vel o

f rea

l int

eres

t ra

tes

vari

able

sex

chan

ge r

ates

chan

ges

Gro

wth

rat

e of

lend

ing

Ass

et p

rice

sG

row

th r

ate

of le

ndin

gTr

ends

in in

flatio

n an

dO

utpu

t an

d in

com

eA

sset

pri

ces

expe

ctat

ions

Savi

ngs

and

inve

stm

ent

GD

P gr

owth

Tren

ds in

inte

rmed

iatio

nM

onet

ary

aggr

egat

esC

orpo

rate

deb

t le

vels

and

com

petit

ion

Bala

nce

of p

aym

ents

1 Var

iabl

es u

sed

to fo

reca

st b

ank

prof

itabi

lity.

2 Inc

lude

s se

para

te a

naly

sis

of s

avin

gs b

anks

and

sec

uriti

es m

arke

ts.

3 The

out

put

from

the

Uni

form

Ban

k Su

rvei

llanc

e Sy

stem

(U

BSS)

is in

putt

ed t

o th

e FI

MS.

The

UBS

S ra

nks

inst

itutio

ns’ f

inan

cial

rat

ios

rela

tive

to t

hose

of t

heir

pee

r gr

oups

,so

as t

o id

entif

y w

hich

inst

itutio

nsar

e pe

rfor

min

g re

lativ

ely

poor

ly.4 T

hese

var

iabl

es h

ave

been

tes

ted

to s

ee if

the

y im

prov

e m

odel

per

form

ance

.5 V

olat

ile li

abili

ties

are

defin

ed a

s th

e su

m o

f:tim

e de

posi

ts o

f $10

0,00

0 or

mor

e;de

posi

ts in

fore

ign

offic

es;f

eder

al fu

nds

purc

hase

and

rep

urch

ase

agre

emen

ts;d

eman

d no

tes

issu

ed t

o th

e U

.S.T

reas

ury;

and

othe

r lia

bilit

ies

for

borr

owed

mon

ey.


two-year time span. While the framework can high-light areas of vulnerability in the banking systemthrough the impact on profits, it is not focused onidentifying specific vulnerabilities. Potential vulner-abilities, however, are taken into consideration in thesensitivity analyses.

The framework is formally linked to the Bank ofFinland’s macro forecasting model of the Finnisheconomy. As such, it incorporates a number ofmacroeconomic variables (for example, GDP growthand interest rates) as inputs to forecast the compo-nents of gross income, while the outputs of theframework (bank profitability, and bank lending anddeposit interest rates) are inputs to the model. Thereis thus a feedback relationship between the frame-work and the macro model, requiring iterations be-tween the two forecast procedures. Data inputs to theframework are used to develop forecasts of the vari-ous components of the banking system’s gross in-come and expenses, which are then aggregated toproduce an estimated trend in banking sector prof-itability over the forecast time frame.

Norway

Norges Bank has produced reports on the situationand outlook for the financial sector since 1995.55 Thework includes both analyses of developments in fi-nancial institutions, primarily in the banking sector,and the relationship between macroeconomic and fi-nancial sector developments. Analyses of the financialposition of households and enterprises are importantelements of this framework. The reports are for the in-ternal use of the financial sector authorities and arenot published. Norges Bank has published excerpts,however, in the second and fourth editions of eachyear’s quarterly Economic Bulletinsince 1997. Inkeeping with the summary approach being taken, thepublished reports are relatively qualitative, with over-all assessments of financial health rather than a focuson critical values of specific indicators.

The approach taken is to generate an initial assess-ment of the trends in macroeconomic variables thatare of relevance to the financial sector and, in partic-ular, to the earnings of financial institutions. Thesevariables include economic growth, interest rates,credit growth, and sectoral debt levels. Followingthis analysis, a range of individual indicators of thefinancial health of the banking system are incorpo-rated in the assessment (e.g., capital adequacy ratios,credit growth rates, trends in overdue loans, interestrate trends, and trends in operating costs). Specificattention is paid to the banks’exposures to the realestate market. Given that past experience has shown

that enterprise sector loans have been a major sourceof bank losses, attention is also paid to the exposureof banks to the enterprise sector, as well as to theability of firms in that sector to cope with an unex-pected deterioration in their financial condition andto stay current with their debt servicing. A similarsectoral analysis is conducted of the financial condi-tion of the household sector. The analysis also cov-ers recent trends in aggregate bank profits, includingthe components of earnings and expenses, and otherbalance sheet items. The objective is to identifytrends, as well as to explain them. Finally, attentionis paid to the risks arising from financial institutions’exposure to the securities markets.

Sweden

The Swedish approach to assessing banking sys-tem health is similar to the one followed in Norway.It is somewhat more formalized but does not follow amodel approach as in the case of Finland. The assess-ments are carried out by the Sveriges Riksbank in thecontext of its responsibility to promote a safe and ef-ficient payment system, as set out in the RiksbankAct. Given the integration of the payment systemwith the financial system as a whole, the Riksbank’ssurveillance of payment system developments en-compasses systemic issues relating to banking sys-tem stability. The Riksbank’s surveillance is directedtoward systems and markets and, therefore, comple-ments the supervision of the banking system by theSwedish Financial Supervision Authority, which isprimarily aimed at individual institutions.

Since 1997, the Riksbank has been publishing re-views of the banking system on a semiannualbasis.56 The approach clearly starts with the paymentsystem. It focuses on aggregated risks, rather thanbank-specific issues. A major objective of the reportsis to raise the financial sector’s awareness of vulner-ability issues. The method is to assess risks to aggre-gate banking sector profits based on informationfrom the markets, on a sector-by-sector basis. Theassessments are carried out by looking at three cate-gories of risk that affect banks’abilities to generateprofits: (1) strategic risks, or factors affecting profitgeneration over the longer term; (2) credit risks, orrisks to profits over the medium term; and (3) coun-terparty and settlement risks, or risks that affectprofits over short and very short terms.

A list of the variables that are examined in theSwedish approach is presented in Table 3. The list is

22

55Norges Bank (1998).

56The first three reports in this series—called Financial MarketReports—focused on an in-depth presentation of several key as-pects of the analysis. Subsequent reports—renamed FinancialStability Reports—have provided updates of the analysis. SeeSveriges Riksbank (1999).

National Central Banks and Supervisory Agencies

indicative, partly because the approach incorporatesa large number of variables, and partly because someaspects of the analysis are qualitative—for instance,the speed of deregulation, the degree of competitionin the banking sector, and assessments of key riskmanagement features—and, therefore, do not lendthemselves to inclusion in a listing of quantitative in-dicators. The macroeconomic variables that are re-viewed include the growth rate of aggregate lending,the rate of change in inflation, changes in inflationexpectations, and the level of real interest rates. Sev-eral banking sector variables are also reviewed, in-cluding profits, the degree of disintermediation,bankruptcies, loan performance by sector, and debt-servicing capabilities by sector.

United Kingdom

Under a 1997 Memorandum of Understanding be-tween the United Kingdom Treasury, the Bank ofEngland, and the Financial Services Authority, theBank of England is responsible for the stability ofthe financial system as a whole.57 A Standing Com-mittee of the Treasury, the Bank of England, and theFinancial Services Authority meets monthly to dis-cuss developments relevant to financial stability.One of the tasks that the Bank of England under-takes to discharge its responsibility is the surveil-lance of financial stability conditions, including theassessment of actual or potential shocks, and of thesystem’s capacity to absorb shocks. The FinancialStability Area of the Bank of England undertakes amonthly assessment of financial stability and pro-duces a variety of more narrowly focused notes. Amore thorough review of the financial stability con-juncture and outlook is undertaken every six monthsand a version is published in the Bank of England’sFinancial Stability Review.58 The review also in-cludes articles relating to the assessment of risks tofinancial stability. While the monthly assessmentsare used to inform some other public documents, theassessments themselves are not published.

Rather than use complex models, the Bank of England’s approach is to review a range of infor-mation from the United Kingdom, industrialeconomies, and emerging market economies and tryto identify key developments, vulnerabilities, andrisks that could affect financial stability. While theBank of England is exploring the use of aggregatedmicroprudential data covering groups of institutionsin this work, and financial market data naturally playa key role, macroeconomic indicators are also re-

garded as important for financial-stability analysis(e.g., saving-investment balances and external bal-ance sheets).

There are some problems with compiling aggre-gated microprudential data, as the existing data reporting systems were designed to support the supervision of individual institutions rather than sur-veillance of the system as a whole. Work is being un-dertaken by the Bank of England, together with theFinancial Services Authority, to address this issue sothat, for example, better analysis of peer groups ofbanks can be carried out. In addition to its work inthis area, the Bank of England is seeking to developits use of MPIs. This work includes identifying MPIsthat might be useful in general, rather than specifi-cally for the United Kingdom, and reviewing generalissues relating to macroprudential surveillance.59

United States

The three institutions that have responsibility fordifferent aspects of banking supervision—the FederalDeposit Insurance Corporation (FDIC), the FederalReserve, and the Office of the Comptroller of the Cur-rency (OCC)—have, over time, developed similarmodels and indicators aimed at assessing the overallhealth of individual banks based on summary datasubmitted by the banks as part of their off-site super-vision exercises. Aggregating the information for in-dividual banks can provide assessments of the healthof significant components of the financial system.Work has also been carried out, for example, by theFederal Reserve, to identify macroeconomic variablesthat could be incorporated in predictive frameworks.While individual variables have been found to be sig-nificant within sample, none have significantly im-proved out-of-sample predictive power.

In general, the variables used in the assessmentsof the future health of individual banks by the super-visory institutions in the United States are proxiesfor the various factors taken into account when per-forming a full ex post CAMELS rating. These vari-ables for an individual bank may be useful as MPIswhen one or more banks individually are sufficientlylarge to have systemic implications. At an aggre-gated level for the banking system, the variables thathave been found to be significant in assessing thecurrent health of a bank may be used as MPIs. As anexample, the variables used by the Federal Reservein its Financial Institutions Monitoring System(FIMS) exercise to assess the current health of abank are included in Table 3 (see page 21).60

23

57Responsibility for the authorization and supervision of indi-vidual financial institutions and providers of financial infrastruc-ture rests with the Financial Services Authority.

58Bank of England (1999).

59Davis (1999).60For a description of the FIMS, see Cole, Cornyn, and Gunther

(1995).


As an extension of the assessments of currenthealth of individual banks, the U.S. supervisoryagencies have also worked on models for assessingthe current riskiness of banks that can generate prob-abilities of future failure. To the extent that thesemodels perform successfully, they can be used tofocus scarce on-site supervision resources on thosebanks judged to be at the highest risk, and to takeearly steps to reduce those risks. Again, the variablesused could be useful as MPIs. As an example of theMPIs that might be derived from this approach, thevariables used by the FDIC in its Growth Manage-ment System (GMS) to assess probabilities of futurebank failures are also provided in Table 3.61 Some ofthe variables that are being examined for possible fu-ture inclusion in the FDIC model are included in thistable as well.

The computerized statistical system that supportsthe work of the three agencies permits joint collec-tion of income, operating activity, and balance sheetdata for individual banks, which are roughly disag-gregated into national and worldwide components.The system generates statistics for a variety of pur-poses by each agency, such as supporting the supervi-sion of individual institutions, econometric research,and aggregation into macroeconomic statistics. Thedata are regularly used for multivariate cross-sectionor time-series analyses of income or balance sheetitems. Also, the data are aggregated by the FederalReserve into the sectoral balance sheets used in theflow of funds and the national accounts. Thus, it is afully integrated system that permits supervisory, sta-tistical, and econometric analysis of microdata, sec-toral structural data, and macroeconomic data.

Indicators Used by Investors andRating Agencies

Private investors such as banks, securities firms,and investment funds use various indicators to evalu-ate the vulnerability of financial systems. The pri-mary purpose is to determine the creditworthiness ofborrowers and issuers that have significant exposureto a particular financial system.62 Such analysis isgenerally based on three kinds of indicators:

• Indicators published by official sources—thegovernment and central bank, as well as interna-

tional financial institutions, such as the IMF, theWorld Bank, and the BIS.

• In-house analyses of vulnerability of financialsystems. This type of analysis, which some fi-nancial institutions conduct, is often based onwell-established indicators such as CAMELSratings, rather than on comprehensive macropru-dential models. Main sources for the indicatorsare IMF or World Bank publications. Indicatorsof the vulnerability of individual borrowers andissuers—an area where financial institutions dohave a comparative advantage through their business relationship—are considered propri-etary information and are usually not publiclyavailable.

• Creditworthiness ratings by credit rating agen-cies.Most investors still regard ratings as the bestavailable indicators of vulnerability, even thoughthe limitations of ratings are well recognized.

For most investors, credit rating agencies are theprimary source of information on the creditworthinessof individual financial institutions (issuer and debt rat-ings), and on the creditworthiness of the government(sovereign risk ratings). Both ratings taken togethercan serve as an indicator of market perceptions of thevulnerability of a country’s financial system.

For sovereign ratings, the most commonly usedindicators are recent economic performance, thequality of economic and financial management, thedepth and sophistication of markets, the stability ofeconomic policy, the stability and effectiveness ofthe political system, and long-term trends and ex-pected future performance. Particular emphasis isoften given to the quality of economic management,the stability of policy, and the depth and sophistica-tion of local markets. Some agencies use method-ologies geared more toward macroeconomic indica-tors such as income and economic structure,economic growth prospects, fiscal flexibility , publicdebt burden, price stability, balance of paymentsflexibility , and external debt and liquidity, as well aspolitical risk. Recent studies have found that credit-worthiness ratings appear to be determined primar-ily by economic events, rather than political vari-ables. Moreover, following the Asian crisis, ratingagencies are placing greater emphasis on factorssuch as external debt and liquidity, banking sound-ness, and corporate leverage.63

For bank ratings, indicators include quantitativefactors such as asset quality, capital adequacy, prof-itability and liquidity, as well as qualitative factorssuch as environment, business franchise values,

24

61For a description of the GMS, see Federal Deposit InsuranceCorporation (1997), Vol. I, pp. 496–507. The OCC has been usinga variety of computer applications to monitor financial institu-tions’ risks; see FDIC (1997), Vol. I, p. 512.

62The indicators used by private investors need to be differenti-ated from the so-called market-based indicators, such as stockmarket and bond indices.

63See Haque, Mark, and Mathieson (1998), and InternationalMonetary Fund (1999c).

Indicators Used by Investors and Rating Agencies

management quality, hidden strengths and reserves,and hidden weaknesses and overvalued assets.64 Inaddition to publishing issuer ratings, rating agen-cies also compile ratings of particular debt issues.In the case of bank debt issues, their usefulness asan MPI may be limited by the fact that these ratingsusually incorporate an evaluation of the likelihoodof government support.65 Therefore, in addition tothose traditional ratings, some agencies have devel-oped ratings that are designed to indicate financial

strength on a stand-alone basis.66 Such ratings re-flect the probability that outside assistance will beneeded, but not the probability that it will be pro-vided. In practice, financial strength ratings pri-marily look at bank-specific elements such as fi-nancial fundamentals, franchise value, and businessand asset diversification, but also take into accountthe bank’s operating environment, including thestrength and prospective performance of the econ-omy, as well as the structure and relative fragilityof the financial system, and the quality of bankingregulation and supervision.

25

64Standard and Poor’s (1999), Fitch IBCA (1998), and Thom-son Financial Bankwatch (1999).

65Government support is often assumed in the presence ofgovernment guarantees, government or quasi-government own-ership or control, high concentration in the banking system, orby precedent.

66Moody’s Investors Service (1999). See also the discussion ofmarket-based indicators in Section II.

IMF surveillance of member countries’economiesunder Article IV of its charter has always in-

cluded, to some degree, the surveillance of financialsystems, primarily with the aim of ensuring the ef-fective functioning of monetary and exchange pol-icy. Article IV staff reports have, on occasion, con-tained special annexes dealing with financial sectordevelopments, but in-depth surveillance by the IMFof financial systems was generally limited, with thefocus being primarily on the IMF’s provision oftechnical assistance in specific areas identified bymember countries or previous IMF missions.

Reports and Publications

Within the framework of multilateral surveillance,the IMF Research Department has published the In-ternational Capital Markets report annually since1980. This report summarizes and analyzes develop-ments in international financial markets, including fi-nancial market indicators that may signal vulnerabili-ties in the global financial system. The report draws,in part, on a series of informal discussions with com-mercial and investment banks, securities firms, stockand futures exchanges, regulatory and monetary authorities, and the staffs of international organiza-tions such as the BIS, the European Commission, theInternational Swaps and Derivatives Association, andthe OECD.

Similarly, the IMF Research Department regularlyanalyzes developments of market-based indicators,including international bond issuance, internationalloans and loan facilities, stock market and bond in-dices (including spreads), and ratings by internationalrating agencies. These data, which reflect market sen-timent toward a country’s economy, and particularlyits financial system, may serve as useful indicators offinancial system vulnerability.67

In 1996 the IMF published Bank Soundness andMacroeconomic Policy, its first major analysis of the

interaction between these two topics.68 The analysishighlights the issues posed by current or potentialbanking system unsoundness in four policy areas: thedesign and implementation of stabilization programs,the use of monetary instruments, the implications forfiscal policy, and the management of internationalcapital flows. It outlines key structural policy issuesrelevant to maintaining a sound banking system andexamines how the IMF might better incorporate bank-ing sector considerations into its surveillance, pro-gram design, and technical assistance work. The bookalso contains a survey of indicators for predictingbank unsoundness, primarily those based on individ-ual bank supervisory information, as used by supervi-sory authorities and central banks.

In January 1998, the IMF published a survey enti-tled Toward a Framework for Financial Stability.69

The survey sets out, among other things, guidelineson the quality of information indicating financial sys-tem vulnerabilities to be used for supervisory report-ing and public disclosure. The guidelines are based oninternationally accepted standards, where they exist,and refer to both qualitative and quantitative informa-tion. Issues covered are, for example, accounting andvaluation rules, loan portfolio review and classifica-tion, treatment of collateral, and loan loss provision-ing. Adherence to internationally accepted minimumstandards in these areas is considered an essential pre-condition for the use of macroprudential data as use-ful indicators of vulnerability.

Surveillance Procedures and OperationsFinancial System Surveillance

In 1998, the Monetary and Exchange Affairs De-partment (MAE) issued an internal guidance notedesigned to facilitate discussions on financial systemissues between the IMF staff and the national author-ities in the context of Article IV surveillance. It sug-gests specific areas for discussion including the fol-lowing topics:

V IMF Initiatives

26

68Lindgren, Garcia, and Saal (1996).69Folkerts-Landau and Lindgren (1998).

67The Research Department also runs an ongoing project to an-alyze, on an experimental basis, the results of early warning sys-tem models.

Surveillance Procedures and Operations

• Sectoral indicators of the health of the bankingand financial system.Indicators of the health ofthe financial system identified as high priorityinclude the foreign exchange exposure of finan-cial institutions, sectoral credit concentration,exposure to large holdings of securities, the ag-gregate ratio of nonperforming loans to totalloans, the aggregate risk-based capital ratio, andcentral bank credits to banks and other financialinstitutions as a proportion of their capital ortheir liabilities.

• Macroeconomic factors that impact the financialsystem.This is a set of indicators concerningmacroeconomic developments that could affectthe financial system, with the following indica-tors considered as high priority: lending booms,asset price booms, high corporate leverage ra-tios, contagion effects, rises in the ratio of theexternal account deficit to GDP, low or decliningaggregate growth rates, and volatility in ex-change and interest rates.

• Elements for the assessment of the institutionaland regulatory frameworks.The focus here is onassessing the adequacy of a broad range of pub-lic policies and frameworks affecting the finan-cial system and the incentive structure, and thelikelihood of the authorities to adhere and en-force best principles and practices. In this con-text, Article IV missions are asked to look at thestructure of the financial system, public disclo-sure and the accounting and legal frameworks,incentive structures and safety nets, prudentialregulations and supervision, and liberalizationand deregulation processes.70

• Main effects of financial system distress.Theguidance note highlights the main macroeco-nomic effects of financial system problems interms of direct monetary effects, direct fiscaleffects, quasi-fiscal effects, and other macro-economic impacts. The guidance note also rec-ommends that these effects be assessed, incor-porated in macroeconomic estimates andprojections, and discussed with the authorities.In case contingent liabilities are identified, it issuggested that IMF staff prepare alternativescenarios of possible additional monetary andfiscal effects.

Complementing the guidance note, MAE trans-mitted to area and functional departments, a set of

related tables and questionnaires. These tables andquestionnaires focus on the structure and perfor-mance of the financial sector and the legal and regu-latory framework for banking supervision. In addi-tion, there is a summary table that may be includedin Article IV reports. The tables and questionnairesare designed to be either sent to the country in ad-vance, and then discussed during the mission, orcompleted during the Article IV mission itself. MAEthen collects the information obtained from the ta-bles and questionnaires and enters it into a databankfor future reference (the databank now containsabout 20 countries). Both the guidance note and theaccompanying tables and questionnaires are beingused by all missions involved in monitoring financialsectors under IMF surveillance.

Financial Sector Assessment Program andFinancial System Stability Assessments

Based on the guidance note and building on workalready conducted in various countries in the contextof Article IV surveillance and the use of IMF re-sources missions, MAE has recently established anenhanced monitoring mechanism for financial sys-tems through in-depth FSSAs. FSSAs are conductedwithin the framework of the joint World Bank-IMFFSAP.71 These assessments are designed to providean instrument to highlight strengths, risks, and vul-nerabilities in the financial sector, as well as the link-ages between financial system developments andmacroeconomic outcomes in the context of IMF sur-veillance, program design, and related technical as-sistance. They also involve an assessment of obser-vance of standards, core principles, and goodpractices in the financial sector, as needed.

The key structural and institutional componentsthat contribute to financial system stability aregrouped into four categories: official oversight andregulations, systemic liquidity developments andpolicy, arrangements for crisis management and re-structuring, and major risk exposures, including sys-temic risks in the payment, clearing, and settlementsystems. These components are to be reviewed andassessed in a comprehensive manner, taking into ac-count the macroeconomic environment and thebroader structural reforms that are under way. Thisprocess is aimed at:

• identifying potential vulnerabilities of financialinstitutions and markets to macroeconomicshocks;

• evaluating the macroeconomic consequences offinancial system vulnerabilities and reform; and

27

70These processes increase opportunities for financial institu-tions and markets to further develop, but may also expose finan-cial institutions to new and more significant risks, while at the same time putting pressure on margins through increasedcompetition. 71For background on the FSAP and FSSAs, see footnote 1.

V IMF INITIATIVES

• developing and sequencing key structural re-forms and restructuring actions to promote fi-nancial system stability.

The appropriate sequencing of these reform compo-nents is to take into account the technical and opera-tional linkages among them and their macroeco-nomic impact.

Both recent IMF initiatives—the guidance noteand the FSSAs—have at their core an analysis ofMPIs. Therefore, the analysis of MPIs forms an inte-gral part of the IMF’s financial system surveillance.The set of indicators that the IMF has identified so

far, through its work on financial systems over manyyears, primarily encompasses the indicators dis-cussed in Section II. The IMF Executive Board hasrecently endorsed additional research and analysiswithin the IMF to identify additional indicators thatcan be useful either generally or in the context ofparticular country circumstances. This work alsoaims at selecting a more limited set of MPIs thatcould be monitored by the IMF on an ongoing basisas part of its surveillance activities. It is expectedthat the experience with FSSAs will contribute tofurther progress in the analysis of MPIs.

28

This section reviews statistical issues pertaining tothe two major types of statistical information

used in macroprudential analysis: macroeconomic in-dicators pertaining to the financial sector (financialmacrostatistics) and aggregated microprudential data.Financial macrostatistics—such as monetary statistics,financial accounts of the System of National Accounts(SNA), and sectoral balance sheets—are frameworksfor organizing data into comprehensive overviews ofthe condition and transactions of the financial sectorand its key components, and thus can provide indica-tors of the activity and operation of the financial sys-tem. Aggregated microprudential data are summationsof (mostly) supervisory information on the conditionof individual banks that may provide indications of theoverall condition of the financial sector.

We examine key issues affecting the statistical ac-curacy, usefulness, and international comparability ofMPIs, and consider how the IMF could integrate workon MPIs into its statistical programs and support na-tional authorities in the compilation of timely and reli-able statistics needed to assess the condition of the fi-nancial system. Appendix I reviews the statisticalframeworks for compilation of macroprudential datathat are in place at the IMF, other international organi-zations, and selected central banks and supervisoryagencies, and reviews the suitability of these frame-works for compilation of macroprudential data.72

The importance of reliable statistics in the assess-ment of the condition of the financial sector is wellestablished. Unfortunately, in a significant numberof problem cases, available statistics have not beenof sufficient timeliness and/or quality to provideearly and clear warning of emerging difficulties. Inthis connection, the importance and quality of mone-tary, balance of payments, and financial system data,as well as the need for comprehensiveness in the col-lection, methodological soundness of the compila-

tion, accuracy of compilation, and timely and infor-mative public disclosure have often been empha-sized. Moreover, comparability of MPIs acrosscountries contributes strongly to their usefulness, apoint emphasized at the September 1999 consulta-tive meeting by private sector users of MPIs. Suchcomparability can be achieved through adherence ofMPIs to internationally agreed supervisory, account-ing, and statistical standards that provide clear rulesfor both the compilation and interpretation of MPIs.

Financial macrostatistics and aggregated micro-prudential data, which are both used in macropru-dential analysis, interrelate in numerous ways be-cause both are derived from individual banks’balance sheets and other detailed financial informa-tion. The two types of data could be brought intocloser correspondence by applying standard statisti-cal concepts (such as definitions of residency, sec-tors, and financial instruments) when compiling ag-gregate microprudential data, and by enhancingfinancial macrostatistics with additional detailneeded for macroprudential analysis (such as infor-mation on nonperforming loans).

Statistical Frameworks for MPIsFinancial Macrostatistics

Nearly all countries compile financial sectormacroeconomic statistics, primarily in the form ofmonetary statistics. However, monetary statistics gen-erally do not provide the specific types of data usedfor macroprudential analysis or may lack needed de-tail. Other financial statistics frameworks, such asflow of funds accounts or sectoral balance sheets,73

VI Measurement Issues

29

72National financial systems are subject to threats from internalconditions and external shocks. This section does not cover statis-tical issues and MPIs related to external shocks because they havealready been discussed extensively in the work leading up to thedevelopment of the data template on international reserves andforeign currency liquidity. See IMF (1999d).

73See Inter-Secretariat Working Group on National Accounts(1993). Financial accounts within the System of National Ac-counts, 1993 (SNA93) framework include detailed flow of fundsaccounts (Tables 11.3a, 11.3b), balance sheets and accumulationaccounts (Table 13.2), and stocks of financial assets and liabilitiesanalyzed by debtor and creditor (Tables 13.3a, 13.3b). Althoughfew countries will compile these accounts at the level of detailpresented in SNA93, the accounts have the flexibility to be fo-cused on analytical or policy questions important to each countrywhile still retaining consistency with the overall framework andinternational comparability.

VI MEASUREMENT ISSUES

can provide the detailed financial information for thefinancial sector and other sectors of the economy thatcan be used for macroprudential analysis. Among thenumerous MPIs that can be constructed directly frommonetary statistics or other financial macrostatisticalframeworks are: central bank credit to banks, the ratioof deposits to M2, the ratio of loans to capital, theratio of loans to total deposits, lending to nonresi-dents, the ratio of foreign currency loans to totalloans, the ratio of foreign currency liabilities to totalcapital, and the distribution of credit by sector.

International standards exist for the constructionof these macrostatistics frameworks, which con-tribute to their comparability across countries. Animportant attribute of these frameworks is that theypresent specific sectors within the context of theoverall economy and can be used to analyze the dy-namics of the financial sector and the transmissionof financial stress across sectors. Also, these frame-works are flexible and can be enhanced with addi-tional detail needed for macroprudential analysis.These frameworks are highly developed in only afew countries.

The IMF is moving to promote compilation of fi-nancial sector macroeconomic statistics harmonizedwith international standards through the forthcomingMonetary and Financial Statistics Manual. Finan-cial statistics compiled in accordance with the man-ual can be further augmented to provide moremacroprudential information, such as on impairmentof claims, credit concentration, maturity of liabili-ties, subordinated debt, capital adequacy, connectedlending, and relations with foreign affiliates.74 Workis currently ongoing also at the ECB to augment themonetary statistics program with macroprudentialinformation (see Appendix I).

Aggregations of Microprudential Data

The second major type of information used formacroprudential analysis consists of summations ofinformation used by supervisors to assess the condi-tion of individual banks. In addition to the use ofthese data in specific MPIs, a recent report by theBank of England called for national supervisory au-thorities to design a template with minimum re-quirements for key indicators of bank quality fordisclosure of aggregated microprudential data to thepublic:

We recommend that national supervisory agencies takeupon themselves the responsibility for the collection,compilation and dissemination of data on banks to meetthe needs of users. These data would be at least at thepeer group and aggregate level; both on solo and con-solidated basis; and include key indicators of capital,asset quality, earnings and liquidity, such as capital ade-quacy ratios, non-performing loans as a percentage oftotal assets, return on assets and equity, and a break-down of assets and liabilities by maturity. Data shouldbe published on a quarterly frequency. The above list isonly a suggested bare minimum and not a comprehen-sive list of indicators. A common disclosure template inthe form of a minimum requirement could be agreed onby the Basel Committee on Banking Supervision andcould be so designed to meet the needs of macropruden-tial surveillance. This would require implementation ofgreater disclosure requirements than those currently ap-plicable in many countries, and possibly even legislativechanges to augment the authority of supervisors to askfor and to publish these data. We recommend that coun-tries take up this task with the priority it deserves.75

Some microprudential information can be mean-ingfully aggregated to provide a useful depiction ofthe condition of the financial sector. Some other mi-croprudential information, however, may reflectspecific information needs of supervisors on thecondition of individual banks that might prove diffi-cult to aggregate or unsuitable for aggregation. Forexample, VaR analysis is only valid for the analysisof specific portfolios. Other potential MPIs are af-fected in a similar way. Also, simple aggregation ofprudential information of individual banks can dis-guise important structural information, and it isoften necessary to supplement the aggregate datawith information on dispersion, peer-group analy-sis, and the interrelationships between systemicallylarge banks.

It is instructive to review how the most com-monly used indicator, the risk-based capital ratio,could be aggregated into a statistic to describe thecondition of the banking sector. The ratios for indi-vidual banks cannot be directly aggregated—dataon the numerator (capital) and the denominator(risk-adjusted assets) must be collected from eachbank and separately aggregated. The supervisorydefinition of capital used as the numerator isunique so that data cannot be extracted directlyfrom either accounting records or statisticalsources, and there are analytical needs to compileseparate information on the three tiers of capitalrecognized by supervisors. Likewise, data on risk-

30

74A number of MPIs can be drawn directly from the financialbalance sheet data used in the forthcoming Monetary and Finan-cial Statistics Manual. An advantage of collecting MPIs throughuse of a standard framework is that macroprudential informationwill apply common statistical standards, such as a standard statis-tical definition of residency, which helps integrate the macropru-dential information into an economywide statistical setting.

75See Davis, Hamilton, Heath, Mackie, and Narain (1999), p. 83.

Statistical Issues Affecting MPIs and International Comparability

weighted assets used in the denominator are alsobased on supervisory concepts not used in account-ing or statistical work, and thus are not comparableacross countries because they are affected by na-tional accounting practices for valuation of assets,accrual of income, and recognition of impairment.The aggregate ratio is calculated by simple divisionof the aggregate numerator by the aggregate de-nominator. A low ratio is a clear sign of vulnerabil-ity, and a declining trend may signal increased riskexposure and possible capital adequacy problems.A relatively high ratio, however, does not guaranteethat there are not serious difficulties in financial in-stitutions that account for a significant share of thesystem’s assets.

There have been numerous calls for compilationand dissemination of information on the aggregaterisk-based capital ratio but, as described above, anumber of practical and conceptual issues, and deci-sions about ancillary information, need to be consid-ered in creating a statistical measure of the ratio.


Table 4 summarizes some of the major statisticalissues affecting MPIs.76 This table cross-classifiesselected MPIs by major types of issues that couldimpede their construction, affect their usefulnessfor analysis or disclosure, or affect internationalcomparability. The focus is on issues related tocompilation of MPIs constructed from aggregatedindividual bank prudential data, which—in contrastto financial macrostatistics, for which there are rec-ognized international standards—are often affectedby a range of statistical problems that might impairtheir comparability across countries and reliabilityas indicators. Even where ample individual bankprudential data exist, there might be practical diffi-culties or conceptual problems in compiling theminto statistical aggregates. The most important sta-tistical issues are discussed in the following sub-sections.

Absence or Diversity of Standards

The usefulness of MPIs for surveillance and pub-lic disclosure is hindered by incomparability across

countries because of a lack of international stan-dards, highly diverse national standards, failure ofstandards to keep up with rapid innovation in finan-cial markets, or failure to adhere to applicable pru-dential or accounting standards. In the cases of su-pervisory and accounting standards, there may be noapplicable international standards, or highly diversenational standards may exist. Also, existing account-ing standards in many countries often apply histori-cal valuations to claims and liabilities, which candisguise changes in corporations’financial condi-tions. Little or no work has been done to date to de-velop statistical formulas and definitions for most ofthe proposed MPIs.

Poor Data on Asset Quality

Poor information on asset quality and on the hold-ers of weak credits impairs the analysis of risks fac-ing the financial sector by reducing the usefulness ofbalance sheet data for making assessments of theconditions of financial institutions. These data limi-tations often can hide the buildup of systemic finan-cial sector problems. Specific data limitations in-clude lack of complete or realistic information onthe full recoverable value of loans and securities,country risk, foreign exchange risk, exposures bycounterparties, and the nettability of claims.77

Use of National Versus Global Consolidations

Much supervisory data is collected using a globalconsolidation that incorporates the worldwide activ-ity of a bank into a single financial statement, whichguarantees that all of its relevant activity is cap-tured. Such data, however, might relate only looselyto financial conditions within any specific countryin which a multinational firm operates, and much ofthe reported data may refer to activity or financialpositions outside national authorities’jurisdictionsand policy control. In contrast, standard macroeco-nomic statistics use a national consolidation, andtherefore exclude affiliated units in other coun-tries.78 National financial statistics can be related tothe other national macroeconomic statistics, such asGDP or national interest rates, and cover national fi-nancial activity that will be under the influence ofnational policy officials.

31

76Table 4 covers MPIs closely related to the banking sector. Sta-tistical needs for MPIs extend over nonbank financial institutions,securities markets, and nonfinancial corporations, but data outsidethe banking sector are often less available. Furthermore, the tableprovides only a first cut at identifying specific statistical problems.The survey of country practices will help identify more preciselythe types of problems that exist and their severity.

77Netting refers to legal and supervisory procedures that permitgross claims and liabilities between two institutions to be nettedinto a single asset or liability position.

78For example, data based on a national consolidation excludethe foreign currency exposures of a bank’s subsidiaries located inother countries. In contrast, such information is captured withinthe global consolidation used by supervisors in order to cover theresources and risks to the entire bank.


32

Tabl

e 4.

Sta

tist

ical

Iss

ues

Aff

ecti

ng M

PIs

Div

erse

No

Prud

entia

lN

o St

atis

tical

Acc

ount

ing

Con

solid

atio

n Po

or D

ata

onBa

nk-S

peci

fic

Der

ivat

ives

St

anda

rds

Stan

dard

sSt

anda

rds

Issu

esA

sset

Qua

lity

Info

rmat

ion

Issu

es

Cap

ital a

dequ

acy

indi

cato

rsA

ggre

gate

cap

ital a

dequ

acy

ratio

s•

••

••

Dis

trib

utio

n of

the

cap

ital a

dequ

acy

ratio

s•

n.a.

•n.

a.•

Ass

et q

ualit

y in

dica

tors

Lend

ing

inst

itutio

nSe

ctor

al c

redi

t co

ncen

trat

ion

••

••

••

Rat

io o

f for

eign

cur

renc

y lo

ans

to t

otal

loan

s•

••

••

•R

atio

of n

onpe

rfor

min

g lo

ans

and

prov

isio

ns t

o to

tal l

oans

••

••

•Lo

ans

to u

npro

fitab

le p

ublic

sec

tor

entit

ies

••

••

Prov

isio

ns fo

r no

nper

form

ing

loan

s•

••

••

n.a.

Ris

k pr

ofile

of a

sset

s•

••

••

Rat

io o

f con

nect

ed le

ndin

g to

tot

al le

ndin

g•

••

••

•n.

a.R

atio

of l

oans

to

capi

tal (

leve

rage

rat

io)

••

••

Del

ays

in p

aym

ents

••

••

Borr

owin

g in

stitu

tion

Deb

t-eq

uity

rat

ios

••

•n.

a.•

Cor

pora

te p

rofit

abili

ty•

••

•n.

a.•

Oth

er in

dica

tors

of c

orpo

rate

con

ditio

nsn.

a.n.

a.n.

a.n.

a.n.

a.n.

a.n.

a.H

ouse

hold

inde

bted

ness

••

••

n.a.

n.a.

Man

agem

ent

indi

cato

rsR

atio

of e

xpen

ses

to t

otal

rev

enue

••

••

••

Earn

ings

per

em

ploy

ee•

••

•N

umbe

r of

new

ly li

cens

ed in

stitu

tions

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

Prof

itabi

lity

indi

cato

rsR

atio

of n

et p

rofit

s to

ass

ets

••

••

•R

atio

of n

et p

rofit

s to

equ

ity•

••

••

Rat

io o

f net

inte

rest

inco

me

to in

com

e/as

sets

••

••

•R

atio

of o

pera

ting

expe

nditu

re t

o in

com

e/as

sets

••

••

•N

arro

w c

usto

mer

bas

en.

a.n.

a.n.

a.n.

a.•

n.a.

Inte

rest

rat

e sp

read

sn.

a.•

n.a.

•n.

a.•

n.a.

Liqu

idity

indi

cato

rsC

entr

al b

ank

cred

it to

fina

ncia

l ins

titut

ions

••

•n.

a.D

epos

its r

elat

ive

to m

onet

ary

aggr

egat

esn.

a.•

n.a.

n.a.

n.a.

n.a.

n.a.

Segm

enta

tion

of in

terb

ank

rate

s•

•n.

a.n.

a.n.

a.•

n.a.

Rat

io o

f loa

ns t

o no

nint

erba

nk d

epos

its•

••

•R

atio

of l

iqui

d as

sets

to

tota

l ass

ets

(liqu

idity

rat

ios)

••

••

•M

atur

ity s

truc

ture

of a

sset

s an

d lia

bilit

ies

••

n.a.

••

•Se

cond

ary

mar

ket

liqui

dity

n.a.

•n.

a.n.

a.n.

a.n.

a.n.

a.


33

Sens

itivi

ty t

o m

arke

t ri

sk in

dica

tors

Rat

io o

f net

fore

ign

exch

ange

exp

osur

e to

ca

pita

l•

••

••

•A

vera

ge in

tere

st r

epri

cing

per

iods

,ass

ets

and

liabi

litie

s•

•n.

a.•

••

Ave

rage

dur

atio

n fo

r as

sets

and

liab

ilitie

s•

n.a.

••

Rat

io o

f equ

ity e

xpos

ure

to c

apita

l•

••

•n.

a.•

Rat

io o

f com

mod

ity p

rice

exp

osur

e to

cap

ital

••

•n.

a.n.

a.•

Mar

ket-

base

d in

dica

tors

Stoc

k m

arke

t pr

ices

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

n.a.

Exce

ss y

ield

sn.

a.•

••

••

•C

redi

t ra

tings

••

•So

vere

ign

yiel

d sp

read

s•

•n.

a.n.

a.n.

a.n.

a.n.

a.

No

tes:

Type

s of

sta

tistic

al is

sues

iden

tifie

d:N

o p

rude

ntia

l sta

ndar

ds.I

ndic

ates

tha

t in

tern

atio

nal s

uper

viso

ry o

r re

gula

tory

sta

ndar

ds d

o no

t ex

ist

for

thes

e M

PIs

or fo

r ke

y co

mpo

nent

s of

the

MPI

s.Fo

r ex

ampl

e,th

e ab

senc

e of

uni

form

su-

perv

isor

y st

anda

rds

for

prov

isio

ns o

r ac

crua

ls o

f inc

ome

on im

pair

ed lo

ans

coul

d re

sult

in w

ide

vari

atio

ns in

the

mea

ning

of M

PIs

empl

oyin

g in

form

atio

n on

the

val

ue o

f loa

ns o

r ba

nk p

rofit

abili

ty.T

here

fore

,da

ta r

epor

ted

by n

atio

nal a

utho

ritie

s m

ay e

mpl

oy d

iffer

ent

conc

epts

or

com

pila

tion

met

hods

and

mig

ht b

e in

com

para

ble

acro

ss c

ount

ries

.N

o s

tati

stic

al s

tand

ards

.Ind

icat

es t

hat

inte

rnat

iona

l sta

tistic

al s

tand

ards

hav

e no

t be

en p

rom

ulga

ted

for

thes

e M

PIs

or fo

r ke

y co

mpo

nent

s of

the

MPI

s,or

tha

t st

atis

tical

equ

ival

ents

to

supe

rvi-

sory

con

cept

s ha

ve n

ot y

et b

een

deve

lope

d.Fo

r ex

ampl

e,in

the

firs

t ca

se,t

here

are

no

stat

istic

al s

tand

ards

on

accr

uals

of i

ncom

e on

impa

ired

loan

s.In

the

sec

ond

case

,litt

le w

ork

has

been

don

e to

dat

e to

deve

lop

stan

dard

sta

tistic

al m

easu

res

to c

aptu

re s

uper

viso

ry in

form

atio

n,ev

en fo

r st

raig

htfo

rwar

d m

easu

res

such

as

sect

oral

con

cent

ratio

n of

lend

ing.

Div

erse

acc

oun

ting

sta

ndar

ds.I

ndic

ates

tha

t an

MPI

is u

nlik

ely

to b

e co

mpa

rabl

e ac

ross

cou

ntri

es b

ecau

se o

f div

erse

acc

ount

ing

prac

tices

in d

iffer

ent

coun

trie

s.C

ons

olid

atio

n is

sues

.Ind

icat

es t

hat

the

cons

olid

atio

n us

ed b

y a

bank

can

affe

ct t

he m

eani

ng o

f the

MPI

.Mos

t im

port

ant,

supe

rvis

ory

info

rmat

ion

colle

cted

from

a m

ultin

atio

nal b

ank

usin

g a

wor

ld-

wid

e co

nsol

idat

ion

that

enc

ompa

sses

act

ivity

thr

ough

out

the

wor

ld m

ight

hav

e lit

tle r

elev

ance

for

the

anal

ysis

of t

he c

ondi

tion

of t

he fi

nanc

ial s

ecto

r in

spe

cific

cou

ntri

es in

whi

ch t

he b

ank

oper

ates

.Con

-so

lidat

ion

issu

es m

ay a

lso

aris

e be

caus

e of

var

iatio

ns b

etw

een

coun

trie

s in

the

uni

ts c

onso

lidat

ed w

ithin

the

rep

orts

(ho

ldin

g co

mpa

nies

,for

eign

tra

de s

ubsi

diar

ies)

.Po

or

data

on

asse

t qu

alit

y.R

efer

s to

poo

r or

mis

sing

info

rmat

ion

on t

he q

ualit

y of

ass

ets,

such

as

impa

irm

ent

of c

laim

s,m

isre

port

ing

of t

he e

ffect

ive

valu

e of

cla

ims,

exce

ssiv

e vo

latil

ity o

f ass

ets,

coun

try

risk

,or

risk

s fr

om o

verc

once

ntra

tion

of in

vest

men

ts.

Ban

k-sp

ecifi

c in

form

atio

n.R

efer

s to

pru

dent

ial i

nfor

mat

ion

spec

ific

to in

divi

dual

ban

ks t

hat

cann

ot p

laus

ibly

be

aggr

egat

ed t

o pr

ovid

e in

form

atio

n on

the

ove

rall

finan

cial

sec

tor.

Qua

litat

ive

in-

form

atio

n,su

ch a

s in

form

atio

n on

the

ski

lls a

nd b

ackg

roun

d of

man

agem

ent,

whi

ch is

info

rmat

ion

typi

cally

sou

ght

by s

uper

viso

rs,c

anno

t be

agg

rega

ted.

Mor

eove

r,in

form

atio

n re

late

d to

spe

cific

por

tfolio

she

ld b

y ba

nks

(VaR

,net

fore

ign

exch

ange

exp

osur

e) o

ften

can

not

be a

ggre

gate

d,or

mus

t be

use

d in

con

junc

tion

with

dis

pers

ion

indi

ces

or m

easu

res

of c

once

ntra

tion.

Der

ivat

ives

issu

es.R

efer

s to

issu

es r

elat

ed t

o th

e re

cogn

ition

,val

uatio

n,or

acc

ount

ing

trea

tmen

t of

der

ivat

ives

and

off-

bala

nce

shee

t in

stru

men

ts.


The use of the two different consolidations canhave important implications for the construction ofMPIs. For example, a global risk-based capitalratio is relevant for the supervision of a bank oper-ating in multiple countries, but it is not possible toaggregate meaningfully global ratios for all banksoperating in a country. This implies that there is aneed to collect separate data for institutions’do-mestic activity and their global activity. Such aseparation is straightforward for some assets andliabilities, such as loans and deposits, but for otheritems there may be difficulties such as uncertaintyover the allocation to individual national branchesof capital items registered at the level of the globalcorporation.79

The scope of MPIs in different countries can alsodiffer significantly depending on the precise collec-tion of units drawn within the consolidated reports.This scope, in turn, depends on factors such as na-tional legal definitions, the scope of activities per-mitted by banks, and rules on consolidation of sub-sidiaries and branches. Moreover, a relatedstatistical coverage problem is that rapid change infinancial markets can result in growth of new finan-cial industries that might not be captured within ex-isting supervisory or statistical reporting systems. Aparticular concern is that supervisory or statisticalsystems may fail to encompass all financial activi-ties that might involve significant systemic risks(e.g., hedge funds and other mutual funds, con-sumer finance companies, trust funds, securitiesclearing systems).

Derivatives and Off-Balance Sheet Positions

Financial derivatives and off-balance sheet posi-tions present special problems in evaluating the con-dition of financial institutions, because of the lack ofreporting of positions, high volatility, and potentiallylarge positions. Such concerns have led the account-ing profession to move toward explicit recognition ofvirtually all derivatives on balance sheets using amarket value or equivalent measure of value (fairvalue). International statistical standards for recogni-tion and valuation of derivatives have also been de-veloped, largely based on work at the IMF. Thesestandards are now just beginning to be implemented,mostly in the context of the Economic and MonetaryUnion (EMU) monetary statistics and the interna-tional reserves template. The Basel Committee on

Banking Supervision, of BIS, and IOSCO have alsoproposed new standards for the recognition, valua-tion, and disclosure of information on derivatives.80

Increased recognition of most derivatives on balancesheets at fair value, which is in line with most newregulatory proposals, will affect many of the pro-posed MPIs.

Options for Further Development of MPIs

A precondition for further work on aggregation ofprudential information for individual banks is ascer-taining through surveys or other means the feasibil-ity (given national legal and supervisory practicesand statistical operations) of collecting data for thevarious types of MPIs that have been proposed. Be-cause of the diversity in national supervisory prac-tices and philosophies of supervision, the types ofprudential data collected by national central banksand national supervisory offices are not well known.The IMF is therefore in the process of carrying out asurvey of national authorities and users of MPIs toascertain what types of MPIs they need, whetherprudential statistics are compiled systematically forindividual banks or are available as aggregates, thetypes of data covered, gaps in coverage, and the ac-counting, legal, and institutional standards that af-fect compilation of the data. National practices andregulations related to public disclosure are alsobeing assessed. An important aspect of the survey isto gather information and ascertain the feasibility ofconstructing a core set of indicators or whether dif-ferent sets of MPIs are required for different typesof economies—such as financial centers, other in-dustrial economies, emerging market economies,and developing economies.

The survey and technical reviews are aimed atgaining a clear understanding of what is involved incompiling or disseminating MPIs. For example, itmight be found that a significant number of MPIsare inherently microeconomic in nature and cannotbe meaningfully aggregated. Moreover, new MPIsmight be proposed and the priorities in the formula-tion of international standards might change.81

34

79In general, it might be difficult to assess the condition of thecapital account of national branches of global enterprises becauseof difficulty in allocating the strengths or weaknesses of theglobal capital account to individual branches. There might also bea lack of transparency on the allocation of income or expenses oncollaborative work between branches in different countries.

80In September 1998, the Basel Committee on Banking Super-vision and the IOSCO Technical Committee issued a joint reportthat covers minimum information standards on credit, liquidity, market, and earnings risk that require marked-to-market and no-tional value data on derivatives by counterparty, maturity, andtype of underlying risk; see BIS and IOSCO (1998). The reportalso suggests that supervisors have access to institutions’internalVaR estimates.

81For example, the Basel Committee on Banking Supervisionhas proposed a substantial revision of the risk-based capital ratio.

Options for Further Develpment of MPIs

Another important element in developing MPIs isto consider them in the context of the rapid changesin perspectives and standards of supervisors, ac-countants, and the public. Many initiatives are under way to develop standards that might bringabout greater coherence and enhance the quality ofMPIs.82 Important changes in standards are now tak-ing place and others are forthcoming in a processthat may take considerable time to approach com-pletion. As standards are developed, national prac-tices will gradually come into line, which should en-hance reporting within each country and improvethe international comparability of data. Moreover, tothe extent that standard-setting organizations cometo agreement among themselves (including on theadoption of applicable international statistical stan-dards), the results will be greater coherence in com-piled data, better understanding by the public, im-proved statistical support for the development ofpolicy, and reductions in respondents’and compil-ers’costs of compiling data. Substantial differencesacross countries will continue for some time though,which requires an approach that works in parallel,both for greater future harmonization of data, butthat also proceeds now on the basis of available, un-harmonized data.

In summary, additional information gatheringand technical research is needed before we come toa decision point on the statistical strategies to fol-low in developing MPIs. Depending on options se-lected for developing MPIs, major resource andprioritization issues as well as organizational orlegal issues could confront international organiza-tions and national entities. Some types of MPIsmay prove difficult and costly to compile, or mayrequire new data collection systems that do not fiteasily into existing statistical arrangements. Con-versely, much of the work of upgrading statistical

systems to encompass MPIs dovetails with the on-going work at the IMF and elsewhere to enhancestatistical, accounting, auditing, and supervisorysystems to keep pace with globalization and rapidchanges in financial markets,83 and thus might beviewed as incremental initiatives to work alreadyunder way.

Following is a list of some of the statistical op-tions available for compiling MPIs. The specificstrategy for following these options will dependgreatly on the willingness, technical strengths, andresources of the various international and nationalentities that might be involved.

(1) Monetary statistics could be augmented withspecific types of data used for macropruden-tial analysis. The additional data soughtwould consist mostly of balance sheet infor-mation, but might also include informationon financial institutions’income, expenses,and profitability. Under this option, the IMFwould augment its existing system for compiling monetary statistics and use it as abasis for compiling MPIs across a range ofcountries.84

(2) A new monthly or quarterly compilation of fi-nancial sector prudential data could be insti-tuted, covering all MPIs (should a decision bemade not to use option 1), or covering onlythose MPIs that are not readily included withina monetary statistics framework. The lead rolein such work could be taken by the IMF orother international organizations.

(3) National entities could be encouraged to com-pile and disseminate unharmonized nationaldata on the condition of individual banks oraggregations of microprudential data.

(4) National entities could enhance their programsto compile financial macrostatistics, especiallysectoral balance sheets and flow of funds ac-counts, to support macroprudential analysis.These accounts are tools to assess the financialstrength or vulnerabilities of the major sectorsof an economy and the potential for transmis-sion of financial stress between sectors.

35

82Important work on the development of standards is being un-dertaken by the Basel Committee on Banking Supervision,IOSCO, the International Accounting Standards Committee, in-ternational statistical organizations, regional organizations, andnational supervisors, among others. An important initiative affect-ing MPIs was the enactment in early 1999 of International Ac-counting Standard No. 39—Financial Instruments: Recognitionand Measurement, which mandates that virtually all financial po-sitions be recorded on balance sheet at market value or equivalentand that impairment and loss of market value be reported on anongoing basis. This standard, where implemented by national au-thorities, would markedly improve the usefulness of accountingdata in the construction of MPIs by providing an accurate andtimely depiction of the value of financial institutions’portfolios.The International Accounting Standards are general standards,however, which may be implemented in somewhat different waysin different countries. An important adjunct of this work is the co-operation of the International Accounting Standards Committeewith IOSCO to extend standards to cover reporting and valuationof securities.

83This fluidity also offers the potential for modification andupgrading of accounting, auditing, supervisory, or statisticalstandards to better extract macroprudential information and tosolidify their methodological bases so that MPIs can be soundlyconstructed and made comparable across countries. Achievingsuch improvements will require close cooperation between sta-tistical, accounting, and supervisory authorities.

84This option has similarities to the ECB’s program of collect-ing MPI information for the EU and its member countries via itsmonetary statistics compilation system (see Appendix I).


(5) Modalities for monitoring or contributing toongoing work to develop international stan-dards could be explored. One possibilitywould be to convene an interdisciplinaryworking group that would follow proposalsfor accounting, auditing, supervisory, and sta-tistical standards as well as changes in disclo-

sure requirements for financial institutions,and that would support harmonization with in-ternational statistical standards.

(6) A handbook or manual on statistical compila-tion of MPIs could be prepared to provideguidance to compilers and to assist users in analyzing MPIs.

36

This section reviews issues related to the publicdissemination of MPIs, in recognition of their

crucial role in the strengthening of financial sectorsurveillance and the oversight of the global financialsystem. It discusses insights from IMF approaches tothe data dissemination standards, taking into accountthe conclusions of the September 1999 consultativemeeting on member countries’and other experiencesin MPI identification and dissemination.

The identification of a core set of MPIs is seen asone of the prerequisites for dissemination of MPIsto the general public. The core MPIs would have tofulf ill the following criteria: (1) usefulness in finan-cial systems surveillance; (2) comparability acrosscountries; (3) feasibility of collecting harmonizeddata; and (4) existence of “best practices” with re-gard to coverage, periodicity, and timeliness of thedata that are being disseminated. Given the substan-tial work ahead in crafting a core set of MPIs, par-ticipants at the meeting believed it was premature atthis time to recommend specific modalities for dis-semination of MPIs, but supported the developmentof incentives to national authorities to compile anddisseminate them.

IMF Initiatives in Data DisseminationStandards

The SDDS and the General Data DisseminationSystem (GDDS) embody a structured approach to en-courage data dissemination.85 The SDDS has oper-ated since 1996 as a system of well-defined guide-lines—that is, a standard—for countries to providethe public with comprehensive, timely, reliable, andaccessible macroeconomic data.86 In order to meet the

standard, the SDDS countries have made significantimprovements in their practices for the compilationand dissemination of the SDDS data categories andunderlying databases.

Whereas the SDDS is intended for countries thatare actively involved in international capital markets,or that aspire to do so, and that have relatively well-developed statistical systems, the GDDS serves as aframework for the long-term improvement of dataand statistical practices across the wider IMF mem-bership. At the time the SDDS was initiated, its re-quirements were recognized as very demanding andnot necessarily applicable or relevant for all coun-tries. Consequently, it was agreed that the GDDS becreated to provide a vehicle to support improvementsin the statistical capacity of the remainder of themembership. The GDDS was established in 1997.

Lessons from the SDDS and GDDS

Three lessons can be drawn from the SDDS andGDDS approach:

(1) The SDDS may be viewed as successful be-cause it continues to provide incentives forcountries to improve their practices on compil-ing and disseminating macroeconomic data.Countries are aware that subscribing to theSDDS may be viewed by market participants ashighly desirable and could enhance their ratingson the international capital markets.

(2) The prescriptions contained in the SDDS havebeen developed based on evidence of best orpreferred practices in the compilation and dis-semination of macroeconomic data by coun-

VII Macroprudential Indicators and Data Dissemination

37

85They also complement other initiatives undertaken by theIMF to foster macroeconomic stability and financial systemsoundness through enhanced transparency, such as the develop-ment of the Code of Good Practices on Transparency in Mone-tary and Financial Policies and the Code of Good Practices onFiscal Transparency.

86In subscribing to the SDDS, countries commit to bringingtheir national statistical practices into alignment with the SDDS

requirements for data coverage, periodicity, timeliness, access practices (including data release calendars), integrity, and qualityproxies (including summary methodologies). Through the Dis-semination Standards Bulletin Board (DSBB), SDDS subscribersprovide information about their statistical practices (so-calledmetadata) for a total of 20 macroeconomic categories, as well asaccess to actual data. The SDDS countries post at least the twolatest data observations for each SDDS data category on their na-tional summary data page to which the DSBB is electronicallylinked. The DSBB website is http://dsbb.imf.org.

VII MACROPRUDENTIAL INDICATORS AND DATA DISSEMINATION

tries that participate in financial markets orthat aspire to do so.

(3) The establishment of two parallel tracks forwork on data dissemination standards—SDDSand GDDS—entails a recognition of the dif-ferent stages of development of countries’sta-tistical systems and the need for an approachto dissemination standards that takes this real-ity into account.

Conclusions on Dissemination Issues from the Consultative Meeting

The major conclusions on data disseminationreached by the participants of the consultative meet-ing can be summarized as follows:87

• There is no single set of MPIs that is currentlybeing disseminated by many countries or that isseen as superior to others, nor are there clearlyidentifiable best practices with regard to the dis-semination of MPI data.

• In the absence of a consensus concerning a coreset of MPIs, it is premature to decide whetherMPIs should be included in the SDDS or alongother tracks for promoting dissemination.

• Dif ferences in approaches to the degree of dis-seminating data, maintaining confidentiality,and providing assessments on the condition ofthe banking sector often seem to be based ondifferent perceptions of how markets mightreact to this information. Some authorities fearthat markets might react adversely to “bad”news about banking sector soundness. The“bad” news could be either perceived or offi-cially validated by authorities’assessments.Other countries disseminate a wide range ofmacroprudential indicators in the belief thatthese data would enable the markets to make theright decisions.88

• Participants agreed on the need for an effort atgreater harmonization of data in terms of cover-age, periodicity, timeliness, and public access.

• The meeting supported conducting a survey ofnational supervisors, statistical authorities, andusers to evaluate the prospects for compiling anddisseminating MPIs.

Next Steps

MPI Dissemination Standards

Dissemination standards typically evolve based onevidence of best practices that are comparable acrosscountries.89 It is clear that significant work liesahead in developing such good practices for MPIs,but also that it should take place in conjunction withprogress in identifying a core set of MPIs for finan-cial system surveillance, as well as in resolving thestatistical issues of measurement.

Actual practice (still to be further investigated)may demonstrate (1) that certain MPIs are more rel-evant for a particular country or country-group andless for another, and (2) that countries differ in theirstatistical capabilities to collect reliable, frequent,and timely data on MPIs.

The current SDDS would perhaps provide a solu-tion to the first situation by means of its “as relevant”provisions that exempt countries from certain require-ments that may not be relevant for a country’s particu-lar economic structure. In such cases, the countrymust make these differences transparent by providingexplicit information in the metadata explaining howand why the particular data set is not considered to berelevant for the country’s economy.

As for the second situation, the case could bemade for a differentiated approach to eventual MPIdissemination—for example, along the lines of thetwo tiers represented by the SDDS and the GDDS.

Incentives

Incentives for the authorities to disseminate reli-able and timely MPIs could come from differentsources. Among these are international technical as-sistance to improve the collection and compilation ofMPIs, and the “appeal” of a technical standard inwhich MPIs might be included.

With regard to the appeal of a standard, it shouldbe noted that eventually including MPIs in the exist-ing SDDS as a dynamic, evolving standard couldhave the following benefits: (1) further strengthen-ing the role of the SDDS; (2) leveraging the “goodwill” of the international community vis-à-vis theSDDS; (3) improving the complementarity ofmacroeconomic and macroprudential indicators;and (4) broadening the application of the SDDS ad-vantages, including the provision of a comprehen-

38

87See Hilbers, Krueger, and Moretti (1999) for details.88The basic approach taken by IMF staff—subject, of course, to

concerns about confidentiality of data for individual institu-tions—is that information is a public good and that enhancedpublic availability of information is desirable.

89In contrast, in the work on international reserves and, to someextent, on external debt in the SDDS, the IMF took the lead in“pushing the envelope” to promote increased dissemination of re-liable, comprehensive, and timely data. In the case of reserves,standards were developed jointly by the IMF and a workinggroup of the CGFS.

Next Steps

sive set of metadata, eventual access to data for aparticular country, and facilitation of potentialcross-country metadata and data comparisons.

While the SDDS is designed to evolve to meetnew challenges—and has already strengthened itsprescription for dissemination of data on interna-tional reserves and external debt—adding MPIs tothe SDDS may result in concerns on the part of ex-isting subscribers. This requirement was not envis-aged at the time of subscription, and some coun-tries may not be in a position to meet the MPIprescriptions, though they are otherwise in obser-vance of the SDDS. Also, including MPIs in theSDDS could conceivably prove daunting for poten-tial new subscribers. These various considerations

do not need to be resolved now. They can be revis-ited later, when more of the necessary preparatorywork has been undertaken.

Metadata

In light of the diverse practices in compiling MPIsand many potential statistical problems, special em-phasis needs to be placed on developing a strategyregarding the role of metadata—the textual descrip-tion of the data series. Information must be providedto users about the MPIs’coverage, public access, in-tegrity of the data, and quality of the data, includingcompilation methods, and adherence to or departuresfrom relevant international standards.

39

This appendix reviews the frameworks for thecollection of MPIs already in place in interna-

tional and regional organizations to ascertain thetypes of data generally available, to identify gaps orweaknesses in coverage, and to assess the potentialfor exchange of data and cooperation in collectionof MPIs.

International Monetary Fund

The IMF collects and disseminates a wide varietyof macrostatistics, but does not systematically col-lect financial microdata. The IMF’s monetary statis-tics comprise a very extensive database on bankinginstitutions. These data are potentially very impor-tant for macroprudential analysis because manyMPIs and the monetary statistics compiled by theIMF are derived from the same data sources—cen-tral bank and commercial bank balance sheets. Thedata have a monthly periodicity and are provided tothe IMF as soon as possible after the reference date.An analytical presentation of monetary data is pub-lished in International Financial Statistics. The pub-lished data only highlight a limited number of mone-tary and credit aggregates. These aggregates arepresented in a modified balance sheet format thatdoes not present total assets, total liabilities, and de-tailed capital account information, and thus does notprovide the structural information needed for macro-prudential analysis. Also, the monetary statisticshave not been constructed under standard accountingrules, such as for valuation or provisioning, whichdetracts from their usefulness for macroprudentialpurposes. Similarly, data on bank income, expenses,and profitability, which are used in many MPIs, arenot collected.

A number of changes to the methodology for com-piling these monetary statistics, which will bringabout greater standardization and harmonization be-tween countries, will be introduced when the IMF’snew Manual on Monetary and Financial Statisticsispublished. The manual recommends that all countriesapply the statistical standards presented in the Sys-tem of National Accounts, 1993, which will result in

standard statistical accounting treatments, definitionsof the financial sector, and classifications and treat-ments of financial instruments. The manual also pro-vides for compilation of aggregate balance sheets forthe domestic banking sector, which in most countrieswould be the only aggregate statistics on the financialpositions and condition of financial institutions. Im-plementation of the standards in the manual wouldaid countries in producing MPIs in a number ofways, including by providing a framework for theclassification and the measurement of financial deriv-atives, and recording assets at their fair market value.The framework of the manual was not designed withMPIs in mind, but it could be extended—after someconceptual work is done—to accommodate furtherinformation on MPIs, such as on asset quality, creditconcentration, capital adequacy, and relations withforeign affiliates. Perhaps half of the proposed MPIscould be integrated into the monetary statisticsframework, with varying degrees of difficulty.

Bank for International Settlements

The BIS publishes international banking statisticsin the form of a semiannual consolidated report ofstatistics on the amount, maturity, and sectoral andnationality distribution of international bank lend-ing. These data are available to the public throughthe BIS website (http://www.bis.org). The data arealso included in a joint BIS/IMF/OECD/World Bankquarterly statistical release on external debt, whichwas recently introduced in order to facilitate timelyaccess to a single set of debt indicators. These dataare also analyzed in depth in the BIS InternationalBanking and Financial Market Developments, whichalso presents discussions of conceptual and statisti-cal issues related to the data, as needed.

The BIS staff, partly in support of the Committeeon the Global Financial System and its predecessor,the Euro-Currency Standing Committee, have car-ried out work following the financial crises in theearly 1990s with a view to identifying indicators offinancial risk, and data have been collected asneeded to support these analyses.

Appendix I Existing Data CollectionFrameworks

40

Appendix I

European Central Bank

The ECB has initiated an MPI project that hasidentified the EMU monetary statistics as a source ofmacroprudential information. The EMU countriescompile a harmonized set of monetary statistics toprovide the statistical basis for the operation of theEMU single monetary policy.90 The statistics arecompiled on a timely basis according to statisticalstandards based on the European System of Ac-counts, 1995 (ESA95).91 Because universal bankingprevails in Europe, the EMU monetary data covermost of the EMU financial sector.

The EMU monetary statistics are presented in astraightforward balance sheet format, with a rea-sonable amount of detail on financial instrumentsand counterparty sector. The data compiled foreach country cover each institution’s activity withinthe country, and separate information is providedon positions with other countries within the EMUand with nonresidents of the EMU. This statisticalconstruction permits the European System of Cen-tral Banks (ESCB) to produce a comprehensivepicture of the financial positions of domestic finan-cial institutions vis-à-vis residents of the countryand residents of the EMU. Although there are somelimitations to the data because the reporting systemwas designed primarily to serve monetary statisticspurposes, the EMU monetary statistics frameworkis comprehensive and methodologically strong. Theframework is also being enhanced to better incor-porate the needs of macroprudential analysis.

The Banking Supervision Committee of theESCB has initiated a project to identify MPIs forthe EU banking sector, and has established a Work-ing Group on Macroprudential Analysis to that end(see Section IV). The ECB has recently completeda “gaps exercise” to inquire about the availabilityof data at the EU national central banks needed tocalculate MPIs from existing data sources. A selec-tion has been made of indicators to follow and datasources to use. The ECB and national central banksare now putting into place mechanisms for compil-ing the data. Most of the balance sheet data soughtwill be taken directly from monetary statistics:monthly balance sheets for the banking sector, sup-plemented by quarterly information that providesgreater detail on borrowing from and lending tononbank financial institutions, corporations, and

households (including a split between consumercredit and mortgage lending). There is also infor-mation on particular types of lending, deposit rates,and interest rate spreads. Data collection at thisstage is limited to the banking sector. In addition tothe collection of data through the monetary statis-tics system, other data are being gathered from na-tional supervisory sources within the Banking Su-pervision Committee. This exercise draws also ondata collection carried out in other supervisory fo-rums, notably the “Groupe de Contact” (composedof representatives of the supervisory authorities ofthe countries in the European Economic Area).Subsequent actions will depend on the results ofthe exercise.

EMU member countries also prepare financialaccounts that detail financial assets and liabilitiesof all major sectors of an economy, and the ECBand Eurostat jointly prepare the Monetary UnionFinancial Accounts (MUFA). The statistical stan-dards for financial accounts are based on ESA95and thus are harmonized with the standards formonetary statistics, so that it is possible to embedthe analysis of the banking sector within the statis-tical framework for financial activity for the entireeconomy and its key components. The specific im-portance of financial accounts for MPIs is that rela-tionships between the financial sector and its credi-tor and debtor sectors are made explicit in a waythat allows tracking of the influence of macroeco-nomic trends on the condition of the financial sec-tor. The sectoral accounts also permit analysis ofthe financial strength or vulnerabilities of the vari-ous sectors, thus supporting the analysis of trans-mission of financial strains between the rest of theeconomy and the banking sector.

World Bank

As noted in Section IV, the World Bank is in-volved in the analysis of financial sector sound-ness, including through its joint work with the IMFunder the FSAP. The Financial Sector LiaisonCommittee of the World Bank and the IMF is cur-rently discussing options for the joint developmentof a financial sector database for internal use thatwill include qualitative information, macroeco-nomic time series, and aggregated microprudentialinformation. Most of the statistical data will bedrawn from databases maintained by other institu-tions, but will also include information gatheredduring the FSAP missions and other consultationswith member countries. The World Bank alsomakes use of macroprudential information in riskassessment models used in conjunction with itslending operations.

41

90EU countries that are not members of the EMU are requiredto use many of the same statistical standards as the EMU countries.

91The ESA95 standards closely conform with the standards in the IMF’s forthcoming Monetary and Financial StatisticsManual.

APPENDIX I

Organization for EconomicCooperation and Development

The OECD collects a wide range of financial sec-tor data from its member countries for use in its reg-ular analysis of national and international financialconditions, as presented in its Financial MarketTrendsand numerous other analytical and statisticalreports. The OECD does not presently collect spe-cific MPIs, but uses a broad range of macro- and microstatistics and qualitative information in its as-sessments of countries’financial situations. How-ever, two OECD publications are of particular inter-est for macroprudential purposes.

• Bank Profitability presents data on (1) bank in-come, expenditure, and profitability; (2) balancesheets, with substantial detail; (3) capital ade-quacy; (4) supplemental data on the number ofinstitutions and employment; and (5) some lim-ited information on the overall structure of the fi-nancial system. A number of countries providedata disaggregated by major type of bank. Dataare available for all OECD member countries,using a standard set of tables that have a ratherdetailed breakdown. The data are subject to anumber of limitations, however, mostly the re-sult of diversity in national coverage.92 Data

have an annual periodicity, and the latest data inthe 1999 report are from 1997 for all but a fewcountries.

• Financial Accounts of OECD Countriespre-sents standard tables with annual data on flowsof funds and balance sheets of most OECDcountries. Detail is given by type of financialinstrument and counterparty sector, and some-times with links to gross saving and investmentin the national accounts. These data are com-piled in accordance with SNA standards andthus provide links between the financial sectorand the overall national economy. This multi-sector by financial instrument framework is po-tentially useful for macroprudential analysis bypermitting examination of the concentration oflending by sector and the transmission of finan-cial stress across sectors. Although adherenceto SNA standards imparts some comparabilityof data across countries, the foreword warnsthat the “extreme diversity that characterizesthe financial institutions of the member coun-tries and the financial instruments they use maylimit the comparability of the statistics.” Datausers are advised to refer to a methodologicalsupplement for information on standard con-cepts, calculation methods, and individualcountry notes. Other limitations are the re-stricted country coverage, availability of onlyannual data, and the long lags in the productionof data by some countries.

42

92“The institutional coverage of the tables has been largely dic-tated by the availability of data on income and expenditure ac-counts of banks. As a result of the reporting methods used inOECD countries, the tables are not integrated in the system of na-tional accounts and are, therefore, not compatible with the finan-cial accounts of OECD countries. International comparisons inthe field of income and expenditure accounts of banks are partic-ularly difficult because of considerable differences in OECDcountries regarding structural and regulatory features of national

banking systems, accounting rules and practices, and reportingmethods.” (Organization for Economic Cooperation and Develop-ment, 1999, Foreword).

Recent financial crises have given rise to increasedefforts by the international community to foster

macroeconomic stability and financial system sound-ness. Transparency in the functioning of world capitalmarkets and of countries’financial policies is beingpromoted. The IMF has taken numerous steps to en-hance transparency and openness, including the estab-lishment and recent strengthening of disclosure stan-dards to guide countries in a number of areas,including data dissemination. The need for these stan-dards, first highlighted by the Mexican financial crisisin 1994/95, was underscored by the recent financialcrises in Asia and elsewhere. The Special Data Dis-semination Standard, complete with an electronic bul-letin board—and in a growing number of cases, elec-tronic links that enable users to move between themetadata and the actual data—has been in place sinceMarch 1996. The General Data Dissemination Systemwas established in December 1997.

Special Data Dissemination Standard

Countries subscribing to the SDDS undertake tofollow good practices in four dimensions:

• Data—coverage, periodicity, and timeliness;

• Access by the public—dissemination of advancerelease calendars, and simultaneous release ofthe data;

• Integrity—disclosure of information on lawsgoverning the compilation and release of thedata, access to the data by other government offi-cials prior to release, ministerial commentary ac-companying the release of the data, revision pol-icy, and advance notice of major changes inmethodology; and

• Quality—dissemination of documentation onmethodology and sources, and dissemination ofdetailed data that support statistical cross-checks.

Under the SDDS, data dissemination practicesare described for a total of 20 data categories cover-

ing the real, fiscal, financial, and external sectors aswell as for population, and are posted on the Dis-semination Standards Bulletin Board. To date, 47countries—representing a mix of industrial, emerg-ing market, and transition economies—have volun-tarily subscribed to the SDDS. Countries are alsorequired to establish an Internet site containing theactual data disseminated under the SDDS, called anational summary data page that is hyperlinked tothe DSBB.

The SDDS has led to wider availability and en-hanced timeliness of published data and greater useof advance data release calendars. In light of the re-cent financial crises, the IMF has also taken stepsto strengthen the SDDS, particularly in the areas ofinternational reserves and external debt. The newreserves template is more comprehensive than theexisting prescription, with subscribers having untilMarch 31, 2000, to meet the new requirements. Improvements in external debt data are also takingplace.

General Data Dissemination System

The GDDS, like the SDDS, was developed inclose collaboration with a wide range of producersand users of statistics. The primary focus of theGDDS is on improvement in data quality. Thisstands in contrast with the SDDS, where the focus ison dissemination in countries that generally alreadymeet high data quality standards. Against this back-ground, the GDDS is one of the most importantstrategic projects for the IMF in the area of statis-tics, where a long-standing objective has been theimprovement of data and statistical practices amongthe membership. It is hoped that the GDDS will alsobe a valuable resource for bilateral and multilateralproviders of technical assistance, and that theGDDS can provide the basis for enhanced coopera-tion with other providers of technical assistance.The GDDS’s purposes are (1) to encourage membercountries to improve data quality; (2) to provide a

Appendix II Special Data DisseminationStandard and General DataDissemination System

43

APPENDIX II

framework for evaluating needs for data improve-ment and setting priorities in this respect; and (3) toguide member countries in the dissemination to thepublic of comprehensive, timely, accessible, and re-liable economic, financial, and sociodemographicstatistics. The framework takes into account, acrossthe broad range of countries, the diversity of theireconomies and the developmental requirements oftheir statistical systems.

Dissemination Standards Bulletin Board

The DSBB website (http://dsbb.imf.org), which ismaintained by the IMF, serves as a tool for marketanalysts and others who track economic growth, in-

flation, and other economic and financial develop-ments in countries around the world. The aim of theDSBB is to strengthen the availability of timely andcomprehensive information on economic and finan-cial statistics and to contribute to the pursuit ofsound macroeconomic policies and improved func-tioning of financial markets.

The DSBB describes the statistical practices of theSDDS-participating countries in the collection, com-pilation, and dissemination of key macroeconomicindicators. DSBB users also have access to actualcountry data on the national summary data pages. Aproject is under way to enhance the DSBB websitewith regard to (1) presentation and functionality; (2) tools for metadata management; (3) provision ofa database for data that are accessible on or via theDSBB; and (4) marketing of the DSBB.

44

Akerlof, George A., 1970, “The Market for ‘Lemons’:Quality Uncertainty and the Market Mechanism,”Quarterly Journal of Economics, Vol. 84 (August),pp. 488–500.

Altman, Edward I., 1968, “Financial Ratios, DiscriminantAnalysis and the Prediction of Corporate Bank-ruptcy,” Journal of Finance, Vol. 23 (September), pp.589–609.

Baig, Taimur, and Ilan Goldfajn, 1999, “Financial MarketContagion in the Asian Crisis,” IMF Staff Papers, Vol.46 (June), pp. 167–195.

Bank for International Settlements (BIS), Basel Commit-tee on Banking Supervision (BCBS), 1988, Interna-tional Convergence of Capital Measurement andCapital Standards(Basel).

———, 1996, Amendment to the Capital Accord to Incor-porate Market Risks(Basel).

———, 1997, Principles for the Management of InterestRate Risk(Basel).

———, 1999a, A New Capital Adequacy Framework(Basel).

———, 1999b, “Capital Requirements and Bank Behav-iour: The Impact of the Basel Accord,” BCBS Work-ing Paper No. 1 (Basel).

———, 1999c, “Supervisory Lessons to Be Drawn fromthe Asian Crisis,” BCBS Working Paper No. 2(Basel).

———, and International Organization of Securities Com-missions, Technical Committee (IOSCO), 1998, Su-pervisory Information Framework for Derivativesand Trading Activities(Basel).

Bank for International Settlements, Committee on theGlobal Financial System, 1999, A Review of Finan-cial Market Events in Autumn 1998(Basel).

Bank of England, 1999,Financial Stability Review, Issue 6(London).

Berg, Andrew, and Catherine Pattillo, 1999, “Are CurrencyCrises Predictable?: A Test,” IMF Staff Papers, Vol.46 (June), pp.107–38.

Bussière, Matthieu, and Christian Mulder, 1999, “ExternalVulnerability in Emerging Market Economies: HowHigh Liquidity Can Offset Weak Fundamentals andthe Effects of Contagion,” IMF Working Paper 99/88(Washington: International Monetary Fund).

Caprio, Gerald, and Daniela Klingebiel, 1996, “Bank Insol-vencies: Cross-Country Experience,” Policy ResearchWorking Paper No. 1620 (Washington: World Bank).

Cole, Rebel A., Barbara G. Cornyn, and Jeffrey W. Gunther,1995, “FIMS: A New Monitoring System for Banking

Institutions,” Federal Reserve Bulletin, Vol. 81 (Janu-ary), pp. 1–15.

Corsetti, Giancarlo, Paolo Pesenti, and Nouriel Roubini,1998, “What Caused the Asian Currency and Finan-cial Crisis? Part II: The Policy Debate,” NBER Work-ing Paper No. 6834 (Cambridge, Massachusetts: Na-tional Bureau of Economic Research).

Davis, E. Philip, 1996, “Institutional Investors, UnstableFinancial Markets and Monetary Policy,” in RiskManagement in Volatile Financial Markets, ed. byFranco Bruni, Donald E. Fair, and Richard O’Brien(Boston: Kluwer Academic Publishers).

———, 1999, Financial Data Needs for Macropruden-tial Surveillance—What Are the Key Indicators ofRisks to Domestic Financial Stability?, Handbooksin Central Banking, Lecture Series No. 2, Center for Central Banking Studies (London: Bank of England).

———, Robert Hamilton, Robert Heath, Fiona Mackie,and Aditya Narain, 1999, Financial Market Data forInternational Financial Stability, Center for CentralBanking Studies (London: Bank of England).

Demirgüç-Kunt, Asli, and Enrica Detragiache, 1998, “TheDeterminants of Banking Crises in Developing andDeveloped Countries,” IMF Staff Papers, Vol. 45(March), pp. 81–109.

———, 1999, “Monitoring Banking Sector Fragility: AMultivariate Logit Approach with an Application tothe 1996–97 Banking Crisis,” Policy Research Work-ing Paper No. 2085 (Washington: World Bank).

Diamond, Douglas W., and Philip H. Dybvig, 1983, “BankRuns, Deposit Insurance, and Liquidity,” Journal ofPolitical Economy, Vol. 91 (June), pp. 401–19.

Dimson, Elroy, and Paul Marsh, 1997, “Stress Tests ofCapital Requirements,” Journal of Banking and Fi-nance, Vol. 21 (December), pp. 1515–46.

Dornbusch, Rudiger, Ilan Goldfajn, and Rodrigo O.Valdés, 1995, “Currency Crises and Collapses,”Brookings Papers on Economic Activity: Macroeco-nomics 2, Brookings Institution, pp. 219–94.

Downes, Patrick T., David Marston, and Ínci Ötker, 1999,“Mapping Financial Sector Vulnerability in a Non-Crisis Country,” IMF Policy Discussion Paper 99/4(Washington: International Monetary Fund).

Ediz, Tolga, Ian Michael, and William Perraudin, 1998,“The Impact of Capital Requirements on U.K. BankBehavior,” Economic Policy Review, Federal ReserveBank of New York, Vol. 4 (October), pp. 15–22 (London: Bank of England).

References

45

REFERENCES

Eichengreen, Barry, and Andrew K. Rose, 1998, “StayingAfloat When the Wind Shifts: External Factors andEmerging-Market Banking Crises,” NBER WorkingPaper No. 6370 (Cambridge, Massachusetts: NationalBureau of Economic Research).

Esquivel, Gerardo, and Felipe B. Larraín, 1998, “ExplainingCurrency Crises,” Harvard Institute for InternationalDevelopment, Development Discussion Paper No. 666(Cambridge, Massachusetts: Harvard University).

Federal Deposit Insurance Corporation, 1997, History ofthe Eighties: Lessons for the Future (Washington).

Federal Reserve Bank of Kansas City, 1997, MaintainingFinancial Stability in a Global Economy: A Sympo-sium(Kansas City).

Fisher, Irving, 1933, “The Debt-Deflation Theory ofGreat Depressions,” Econometrica, Vol. 1 (October),pp. 337–57.

Fitch IBCA, 1998, International Bank Rating Methodol-ogy, October 1, 1998 (New York).

Folkerts-Landau, David, and Carl-Johan Lindgren, 1998,Toward a Framework for Financial Stability, WorldEconomic and Financial Surveys (Washington: Inter-national Monetary Fund).

Frankel, Jeffrey A., and Andrew K. Rose, 1996, “CurrencyCrashes in Emerging Markets: An Empirical Treat-ment,” Journal of International Economics, Vol. 41(November), pp. 351–66.

Fratzscher, Marcel, 1998, “Why Are Currency CrisesContagious? A Comparison of the Latin AmericanCrisis of 1994–1995 and the Asian Crisis of1997–1998,” Weltwirtschaftliches Archiv, Vol. 134,No. 4, pp. 664–91.

Furman, Jason, and Joseph E. Stiglitz, 1998, “EconomicCrises: Evidence and Insights from East Asia,”Brookings Papers on Economic Activity: Macroeco-nomics 2, Brookings Institution, pp. 1–135.

Gavin, Michael, and Ricardo Hausmann, 1996, “TheRoots of Banking Crises: The Macroeconomic Con-text,” Inter-American Development Bank WorkingPaper No. 318 (Washington: Inter-American Devel-opment Bank).

Glick, Reuven, and Andrew Rose, 1998, “Contagion andTrade: Why Are Currency Crises Regional?” NBERWorking Paper No. 6806 (Cambridge, Massachusetts:National Bureau of Economic Research).

Goldstein, Morris, and Philip Turner, 1996, BankingCrises in Emerging Economies: Origins and PolicyOptions, BIS Economic Paper No. 46 (Basel: Bankfor International Settlements).

González-Hermosillo, Brenda, 1999, “Determinants ofEx-Ante Banking System Distress: A Macro-MicroEmpirical Exploration of Some Recent Episodes,”IMF Working Paper 99/33 (Washington: Interna-tional Monetary Fund).

———, Ceyla Pazarbas,iogvlu, and Robert Billings, 1997,“Determinants of Banking System Fragility: A CaseStudy of Mexico,” Staff Papers, International Mone-tary Fund, Vol. 44 (September), pp. 295–314.

Goodhart, Charles, 1995, “Price Stability and FinancialFragility,” in Financial Stability in a Changing Envi-ronment, ed. by Kuniho Sawamoto, Zenta Nakajimaand Hiroo Taguchi (New York: St Martin’s Press).

Guttentag, Jack, and Richard Herring, 1984, “Credit Ra-tioning and Financial Disorder,” Journal of Finance,Vol. 39 (December), pp. 1359–82.

Haque, Nadeem Ul, Nelson C. Mark, and Donald J.Mathieson, 1998, “The Relative Importance of Polit-ical and Economic Variables in CreditworthinessRatings,” IMF Working Paper 98/46 (Washington:International Monetary Fund).

Hardy, Daniel C., and Ceyla Pazarbas,iogvlu, 1998, “Lead-ing Indicators of Banking Crises: Was Asia Differ-ent?,” IMF Working Paper 98/91 (Washington: Inter-national Monetary Fund).

Hendricks, Darryll, 1996, “Evaluation of Value-at-RiskModels Using Historical Data,” Economic Policy Re-view, Federal Reserve Bank of New York, Vol. 2(April), pp. 39–69.

Hilbers, Paul, Russell Krueger, and Marina Moretti,1999, “Macroprudential Indicators—Seminar Dis-cusses Ways to Assess Soundness of Financial Sys-tem to Improve Surveillance,” IMF Survey, Vol. 28,No. 18, pp. 296–97.

Honohan, Patrick, 1997, “Banking System Failures in De-veloping and Transition Countries: Diagnosis andPrediction,” BIS Working Paper No. 39 (Basel: Bankfor International Settlements).

International Monetary Fund, 1998, World Economic Out-look: May 1998, World Economic and Financial Sur-veys (Washington).

———, 1999a, “Communiqué of the Interim Committee ofthe Board of Governors of the International MonetaryFund,” PR/99/46, September 26, 1999.

———, 1999b, World Economic Outlook: May 1999, WorldEconomic and Financial Surveys (Washington).

———, 1999c, International Capital Markets: Develop-ments, Prospects, and Key Policy Issues, World Economic and Financial Surveys (Washington).

———, 1999d, “Data Template on International Re-serves and Foreign Currency Liquidity: OperationalGuidelines,” IMF Statistics Department. Availablevia the Internet: http://dsbb.imf.org/guide.htm.

Inter-Secretariat Working Group on National Accounts,1993, System of National Accounts 1993. Preparedunder the auspices of the Commission of the Euro-pean Communities—Eurostat, IMF, OECD, UnitedNations, and World Bank (Brussels: Commission ofthe European Communities).

Kaminsky, Graciela, 1999, “Currency and BankingCrises: The Early Warnings of Distress,” IMF Work-ing Paper 99/178 (Washington: International Mone-tary Fund).

———, Saul Lizondo, and Carmen Reinhart, 1998,“Leading Indicators of Currency Crises,” IMF StaffPapers, Vol. 45 (March), pp. 1–48.

Kaminsky, Graciela, and Carmen Reinhart, 1998, “OnCrises, Contagion, and Confusion,” (unpublished;Washington: George Washington University).

———, 1999, “The Twin Crises: The Causes of Bankingand Balance-of-Payments Problems,” American Eco-nomic Review, Vol. 89 (June) pp. 473–500.

Kawai, Masahiro, 1998, “The East Asian Currency Crisis:Causes and Lessons,” Contemporary Economic Pol-icy, Vol. 16 (April), pp. 157–72.

46

References

Keeley, Michael C., 1990, “Deposit Insurance, Risk, andMarket Power in Banking,” American Economic Re-view, Vol. 80 (December), pp. 1183–200.

Kodres, Laura, and Matthew Pritsker, 1998, “A RationalExpectations Model of Financial Contagion,” Financeand Economics Discussion Series No. 1998–48(Washington: Board of Governors of the Federal Re-serve System).

Kwack, Sung, 1998, “The Financial Crisis in Korea:Causes and Cure,” IMF Seminar Series No. 1998–19(Washington: International Monetary Fund).

Lane, William R., Stephen W. Looney, and James W.Wansley, 1986, “An Application of the Cox Propor-tional Hazards Model to Bank Failure,” Journal of Banking and Finance, Vol. 10 (December), pp. 511–31.

Lindgren, Carl-Johan, Gillian Garcia, and Matthew I. Saal,1996, Bank Soundness and Macroeconomic Policy(Washington: International Monetary Fund).

Mishkin, Frederic S., 1996, “Understanding FinancialCrises: A Developing Country Perspective,” NBER Working Paper No. 5600 (Cambridge, Massachusetts: National Bureau of Economic Research).

Moody’s Investors Service, 1999, Bank Financial StrengthRatings(New York).

Norges Bank, 1998, “Financial Sector Outlook: SecondHalf of 1998,” Economic Bulletin, Vol. 69, Issue No.4 (Oslo).

Organization for Economic Cooperation and Develop-ment, 1999, Bank Profitability: Financial Statementsof Banks 1999(Paris).

Radelet, Steven, and Jeffrey D. Sachs, 1998, “The EastAsian Financial Crisis: Diagnosis, Remedies, Pros-

pects,” Brookings Papers on Economic Activity:Macroeconomic 1, Brookings Institution, pp. 1–90.

Rojas-Suárez, Liliana, and Steven R. Weisbrod, 1995, Fi-nancial Fragilities in Latin America: The 1980s and1990s, IMF Occasional Paper No. 132 (Washington:International Monetary Fund).

Sachs, Jeffrey D., Aaron Tornell, and Andrés Velasco,1996, “Financial Crises in Emerging Markets: TheLessons from 1995,” Brookings Papers on EconomicActivity: Macroeconomics 1, Brookings Institution,pp. 147–215.

Scharfstein, David S., and Jeremy C. Stein, 1990, “HerdBehavior and Investment,” American Economic Re-view, Vol. 80 (June), pp. 465–79.

Sinkey, Joseph F., 1978, “Identifying ‘Problem’Banks:How Do the Banking Authorities Measure a Bank’sRisk Exposure?” Journal of Money, Credit, and Bank-ing, Vol. 10 (May), pp. 184–93.

Standard & Poor’s, 1999, Bank Rating Analysis Methodol-ogy Profile, February (New York).

Sveriges Riksbank, 1999, Financial Stability Report, May(Stockholm).

Thomson Financial BankWatch, 1999, BankWatch Rat-ings—Characteristics and Methodology (New York).

Thomson, James B., 1991, “Predicting Bank Failures inthe 1980s,” Economic Review, Federal Reserve Bankof Cleveland, Vol. 27, No. 1, pp. 9–20.

United States, Department of the Treasury, 1999, “State-ment of G-7 Finance Ministers and Central BankGovernors,” LS-120, September 25, 1999.

Whalen, Gary, 1991, “A Proportional Hazards Model ofBank Failure: An Examination of Its Usefulness as anEarly Warning Tool,” Economic Review, Federal Re-serve Bank of Cleveland, Vol. 27, No. 1, pp. 21–31.

47

48

OCCASIONAL PAPERS

Recent Occasional Papers of the International Monetary Fund

192. Macroprudential Indicators of Financial System Soundness, by a staff team led by Owen Evans, AlfredoM. Leone, Mahinder Gill, and Paul Hilbers. 2000.

191. Social Issues in IMF-Supported Programs, by Sanjeev Gupta, Louis Dicks-Mireaux, Ritha Khemani,Calvin McDonald, and Marijn Verhoeven. 2000.

190. Capital Controls: Country Experiences with Their Use and Liberalization, by Akira Ariyoshi, Karl Haber-meier, Bernard Laurens, Inci Ötker-Robe, Jorge Iván Canales Kriljenko, and Andrei Kirilenko. 2000.

189. Current Account and External Sustainability in the Baltics, Russia, and Other Countries of the Former So-viet Union, by Donal McGettigan. 2000.

188. Financial Sector Crisis and Restructuring: Lessons from Asia, by Carl-Johan Lindgren, Tomás J.T. Bal-iño, Charles Enoch, Anne-Marie Gulde, Marc Quintyn, and Leslie Teo. 1999.

187. Philippines: Toward Sustainable and Rapid Growth, Recent Developments and the Agenda Ahead, byMarkus Rodlauer, Prakash Loungani, Vivek Arora, Charalambos Christofides, Enrique G. De la Piedra,Piyabha Kongsamut, Kristina Kostial, VictoriaSummers, and Athanasios Vamvakidis. 2000.

186. Anticipating Balance of Payments Crises:The Role of Early Warning Systems, by Andrew Berg, Ed-uardo Borensztein, Gian Maria Milesi-Ferretti, and Catherine Pattillo. 1999.

185. Oman Beyond the Oil Horizon: Policies Toward Sustainable Growth, edited by Ahsan Mansur and VolkerTreichel. 1999.

184. Growth Experience in Transition Countries, 1990–98, by Oleh Havrylyshyn, Thomas Wolf, Julian Beren-gaut, Marta Castello-Branco, Ron van Rooden, and Valerie Mercer-Blackman. 1999.

183. Economic Reforms in Kazakhstan, Kyrgyz Republic, Tajikistan, Turkmenistan, and Uzbekistan, byEmine Gürgen, Harry Snoek, Jon Craig, Jimmy McHugh, Ivailo Izvorski, and Ron van Rooden. 1999.

182. Tax Reform in the Baltics, Russia, and Other Countries of the Former Soviet Union, by a Staff Team Ledby Liam Ebrill and Oleh Havrylyshyn. 1999.

181. The Netherlands: Transforming a Market Economy, by C. Maxwell Watson, Bas B. Bakker, Jan KeesMartijn, and Ioannis Halikias. 1999.

180. Revenue Implications of Trade Liberalization, by Liam Ebrill, Janet Stotsky, and Reint Gropp. 1999.

179. Dinsinflation in Transition: 1993–97, by Carlo Cottarelli and Peter Doyle. 1999.

178. IMF-Supported Programs in Indonesia, Korea, and Thailand: A Preliminary Assessment, by Timothy Lane,Atish Ghosh, Javier Hamann, Steven Phillips, Marianne Schulze-Ghattas, and Tsidi Tsikata. 1999.

177. Perspectives on Regional Unemployment in Europe, by Paolo Mauro, Eswar Prasad, and Antonio Spilim-bergo. 1999.

176. Back to the Future: Postwar Reconstruction and Stabilization in Lebanon, edited by Sena Eken andThomas Helbling. 1999.

175. Macroeconomic Developments in the Baltics, Russia, and Other Countries of the Former Soviet Union,1992–97, by Luis M. Valdivieso. 1998.

174. Impact of EMU on Selected Non–European Union Countries, by R. Feldman, K. Nashashibi, R. Nord,P. Allum, D. Desruelle, K. Enders, R. Kahn, and H. Temprano-Arroyo. 1998.

173. The Baltic Countries: From Economic Stabilization to EU Accession, by Julian Berengaut, AugustoLopez-Claros, Françoise Le Gall, Dennis Jones, Richard Stern, Ann-Margret Westin, Effie Psalida,Pietro Garibaldi. 1998.

172. Capital Account Liberalization: Theoretical and Practical Aspects, by a staff team led by Barry Eichen-green and Michael Mussa, with Giovanni Dell'Ariccia, Enrica Detragiache, Gian Maria Milesi-Ferretti,and Andrew Tweedie. 1998.

171. Monetary Policy in Dollarized Economies, by Tomás Baliño, Adam Bennett, and Eduardo Borensztein.1998.

49

Occasional Papers

170. The West African Economic and Monetary Union: Recent Developments and Policy Issues, by a staffteam led by Ernesto Hernández-Catá and comprising Christian A. François, Paul Masson, Pascal Bou-vier, Patrick Peroz, Dominique Desruelle, and Athanasios Vamvakidis. 1998.

169. Financial Sector Development in Sub-Saharan African Countries, by Hassanali Mehran, Piero Ugolini,Jean Phillipe Briffaux, George Iden, Tonny Lybek, Stephen Swaray, and Peter Hayward. 1998.

168. Exit Strategies: Policy Options for Countries Seeking Greater Exchange Rate Flexibility , by a staff teamled by Barry Eichengreen and Paul Masson with Hugh Bredenkamp, Barry Johnston, Javier Hamann,Esteban Jadresic, and Inci Ötker. 1998.

167. Exchange Rate Assessment: Extensions of the Macroeconomic Balance Approach, edited by Peter Isardand Hamid Faruqee. 1998

166. Hedge Funds and Financial Market Dynamics, by a staff team led by Barry Eichengreen and DonaldMathieson with Bankim Chadha, Anne Jansen, Laura Kodres, and Sunil Sharma. 1998.

165. Algeria: Stabilization and Transition to the Market, by Karim Nashashibi, Patricia Alonso-Gamo, StefaniaBazzoni, Alain Féler, Nicole Laframboise, and Sebastian Paris Horvitz. 1998.

164. MULTIMOD Mark III: The Core Dynamic and Steady-State Model, by Douglas Laxton, Peter Isard,Hamid Faruqee, Eswar Prasad, and Bart Turtelboom. 1998.

163. Egypt: Beyond Stabilization, Toward a Dynamic Market Economy, by a staff team led by Howard Handy.1998.

162. Fiscal Policy Rules, by George Kopits and Steven Symansky. 1998.

161. The Nordic Banking Crises: Pitfalls in Financial Liberalization?by Burkhard Dress and CeylaPazarbasıoglu. 1998.

160. Fiscal Reform in Low-Income Countries: Experience Under IMF-Supported Programs, by a staff team ledby George T. Abed and comprising Liam Ebrill, Sanjeev Gupta, Benedict Clements, Ronald McMor-ran, Anthony Pellechio, Jerald Schiff, and Marijn Verhoeven. 1998.

159. Hungary: Economic Policies for Sustainable Growth, Carlo Cottarelli, Thomas Krueger, RezaMoghadam, Perry Perone, Edgardo Ruggiero, and Rachel van Elkan. 1998.

158. Transparency in Government Operations, by George Kopits and Jon Craig. 1998.

157. Central Bank Reforms in the Baltics, Russia, and the Other Countries of the Former Soviet Union, by astaff team led by Malcolm Knight and comprising Susana Almuiña, John Dalton, Inci Otker, CeylaPazarbasıoglu, Arne B. Petersen, Peter Quirk, Nicholas M. Roberts, Gabriel Sensenbrenner, and JanWillem van der Vossen. 1997.

156. The ESAF at Ten Years: Economic Adjustment and Reform in Low-Income Countries, by the staff of theInternational Monetary Fund. 1997.

155. Fiscal Policy Issues During the Transition in Russia, by Augusto Lopez-Claros and Sergei V. Alexas-henko. 1998.

154. Credibility Without Rules? Monetary Frameworks in the Post–Bretton Woods Era, by Carlo Cottarelli andCurzio Giannini. 1997.

153. Pension Regimes and Saving, by G.A. Mackenzie, Philip Gerson, and Alfredo Cuevas. 1997.

152. Hong Kong, China: Growth, Structural Change, and Economic Stability During the Transition, by JohnDodsworth and Dubravko Mihaljek. 1997.

151. Currency Board Arrangements: Issues and Experiences, by a staff team led by Tomás J.T. Baliño andCharles Enoch. 1997.

150. Kuwait: From Reconstruction to Accumulation for Future Generations, by Nigel Andrew Chalk, Mo-hamed A. El-Erian, Susan J. Fennell, Alexei P. Kireyev, and John F. Wilson. 1997.

149. The Composition of Fiscal Adjustment and Growth: Lessons from Fiscal Reforms in Eight Economies, byG.A. Mackenzie, David W.H. Orsmond, and Philip R. Gerson. 1997.

148. Nigeria: Experience with Structural Adjustment, by Gary Moser, Scott Rogers, and Reinold van Til, withRobin Kibuka and Inutu Lukonga. 1997.

Note: For information on the title and availability of Occasional Papers not listed, please consult the IMF Publications Catalog or contact IMFPublication Services.

Macroprudential Indicators of Financial System Soundness ...

Documents

Transcript of Macroprudential Indicators of Financial System Soundness ...