Financial Stability Reports: What Are They Good For?

Financial Stability Reports: What Are They Good For? M. Čihák, S. Muñoz, S. TehSharifuddin, and K. Tintchev June 20, 2012 FPD Academy Comments from staff at IMF, the World Bank, and several central banks gratefully acknowledged. Views expressed here are those of the authors and do not necessarily represent those of the IMF, the World Bank, or any other organization.

Motivation “Banks are well prepared to withstand increased delinquency and loan losses, which have been extremely low to date…” “Nowhere has it been suggested that the banking systems of individual countries should be subjected to size limitations…” Iceland’s Financial Stability Report (2008) “Low probability that in 2008 that existing risks might materialize to the extent that it will have an impact on bank performance…” Latvia’s Financial Stability Report (2007)

Introduction, objectives, and preview of results Need for effective policy framework for financial stability Wide range of instrumentstools and devices are being considered Background Current financial stabilityframework includes financial stability reports (FSRs), which aim to communicate key risks and vulnerabilities Early reviews of FSRs experience have been mixed Objective Ascertain to what extent and in what form FSRs should be part of the new framework Simply publishing a report has no significant impact. However, higher quality reports tend to be associated with a more stable financial environment. Preview of results Areas for improvement: forward-looking assessment, interconnectedness analysis, data issues

Financial stability reports: what are they good for? Trends in financial stability reporting Case studies Empirical evidence Conclusions, further research

FSRs: a growing “industry” (1996–2011) Rapid growing “industry” of FSR publications Less rapid growth in number of FSRs published *Estimate: Newly published FSRs in 2011 and assumed all the central banks that have published FSRs previously will continue to publish in 2011

Global trends reflect the evolving experience with FSRs and dynamics of financial stability communications • Ireland • Ireland ceased to publish its FSRs from 2008 onwards • United Kingdom • Bank of England’s (BoE) FSR went through a marked revamp in 2006 • Existing elements shortened and new features added • France • Banque de France stopped publishing its FSR in 2007 • FSRs currently for internal use only • Central bank website contains a link to FSRs, but the FSRs are a collection of articles or conference materials on a featured topic • Israel • Bank of Israel stopped issuing FSRs as a stand-alone publication in 2005 • Financial stability issues are covered as a chapter in its annual report • USA • US recently published its first FSR • Report is published by the Financial Stability Oversight Council (FSOC) • Central bank is represented in the publishing body, but not the sole publisher

General trends of FSR publication have not changed dramatically in the last 15 years • Frequency, length and structure have not changed dramatically • Several countries have switched frequency of FSR publication • Average length of FSR has declined somewhat, but marginally

Other common trends observed in worldwide FSR publication Coverage • Coverage of issues in FSRs tends to increase over time Market-based indicators • Increasing use of market-based indicators such as credit-default swaps (CDS) Data publication • Increasing number of central banks publishing underlying data

Eight case studies of FSRs – Background • Eight countries selected with a view to having a reasonably balanced coverage; geographically, advanced vs. emerging markets • Analyzed based on “good practices” proposed by Čihák (2006) – CCC framework Iceland Brazil Canada Advanced Emerging market Advanced New Zealand Latvia Korea Spain South Africa Minimal crisis in 2008 Major crisis in 2008 Minimal crisis in 2008 Emerging market Advanced Advanced Emerging market Emerging market Major crisis in 2008 Minimal crisis in 2008 Crisis in 2008 Minimal crisis in 2008 Minimal crisis in 2008

Framework introduced in Čihák (2006) “Three Cs” Key elementsof the report E.g.: . Is it clear what data were used to arrive at the results presented in the report? (5 % weight) Based on methodology used by Fracasso, Genberg, and Wyplosz (2003) to assess central banks’ inflation reports. They showed that for inflation reports, higher ‘quality’ measured this way was associated with lower dispersion in inflation expectations and lower inflation, on average.

Framework introduced in Čihák (2006) “Three Cs” Key elementsof the report E.g.: . Is it clear what assumptions were used to arrive at the results presented in the report? (5 % weight) Based on methodology used by Fracasso, Genberg, and Wyplosz (2003) to assess central banks’ inflation reports. They showed that for inflation reports, higher ‘quality’ measured this way was associated with lower dispersion in inflation expectations and lower inflation, on average.

Framework introduced in Čihák (2006) “Three Cs” Key elementsof the report E.g.: Is the overall assessment of financial stability presented consistently with previous reports? (5 % weight) Based on methodology used by Fracasso, Genberg, and Wyplosz (2003) to assess central banks’ inflation reports. They showed that for inflation reports, higher ‘quality’ measured this way was associated with lower dispersion in inflation expectations and lower inflation, on average.

Framework introduced in Čihák (2006) “The three Cs” The key elements of the report E.g.: Does the report use a wide range of tools, both quantitative (stress tests etc) and qualitative (supervisory, market participant surveys, etc.)? (5 % weight) Based on methodology used by Fracasso, Genberg, and Wyplosz (2003) to assess central banks’ inflation reports. They showed that for inflation reports, higher ‘quality’ measured this way was associated with lower dispersion in inflation expectations and lower inflation, on average.

Report’s objective: clearly indicatedDefinition of financial stability: mixed experience • Financial stability definition is typically featured at the beginning of the report • Aims or objectives of FSR are clearly indicated in most of the FSRs reviewed • Objectives are usually presented at the beginning of the report • Significant variation in FSRs objective is observed across the reports: • Identifying and analyzing risks to the financial system • Provide information for major participants in the financial industry to evaluate and manage risks • Stimulatedialogue and discussions on financial stability issues • Highlighting efforts and policies of regulatory authoritiesin mitigating risks Definition of financial stability included? Yes No • Canada • Iceland • Korea • South Africa • Brazil • Latvia • New Zealand • Spain

A fair degree of consistency in covering key systemic risk factors • Issues and developments in the FSRs examined are consistently followed-up • Some countries provided detailed analysis on specific risks

Less attention to analysis on contagion risks and interconnectedness in the FSRs reviewed • Not reported regularly in the sample FSRs • Some countries do perform ad-hoc analysis depending on current macroeconomic and financial conditions: • No analysis or explanation of linkages or exposures among domestic banks • Two exceptions: • Korea: November 2010 FSR, box article • South Africa: March 2011 FSR, working paper appendix • All eight countries’ FSRs feature a regular assessment of corporate sector and households • Description of the health of household and corporate sector and impact on banks’ • No analysis on sovereign exposures of the banking system reported in our sample FSRs

Forward looking views, assessments or projections are neither consistently nor comprehensively reported • Bulk of analyses and discussions concentrate on current level of ratios and indicators • FSRs reviewed seldom include statements, assessments or even survey results, e.g: • “Going forward, substantial upward pressure on real estate prices seem to still exist..” • “…a modest increase in the aggregate rate of impairment are likely to continue into 2011.” • Nonetheless, most of the sample FSRs do provide qualitative outlook on credit risk

Four countries’ FSRs in our sample reported stress test results on a regular basis • FSRs for Canada, Korea, Latvia and Brazil usually report at least one type of stress test results in each publication • In some cases, stress tests done by other regulators: Australia Prudential Regulatory Authority (APRA) in New Zealand’s FSR, Financial Supervisory Authority (FME) in Iceland’s FSR and ECB in Spain’s FSR

Reporting of stress test results across countries (2008-2011) • Legend • For each of these questions, a response of ‘yes’ gets a value of 1, while a 0 is assigned to a ‘no’ • Assumptions: Are the assumptions to the stress tests included whenever the results of a stress test are reported? • Graphically: Does the reporting of the stress test results include any graphical presentations, either as a diagram or in a table? • Granular: Do the FSRs report stress test results beyond aggregated level i.e. there is some degree of granularity e.g. no. of institutions, percentage of banks? • Methodology: Is the methodology of the stress test s explained in the FSR? • Text: Are the results to the stress tests reported as part of regular text of the report?

In practice, FSR discussion on policies extends beyond macroprudential policies • Ideally, policy discussions (regardless of the type) should be tied back to its impact on financial stability • Policies and measures to mitigate the impact of the global financial crisis, especially post-Lehman • E.g. new responsibilities granted to central bank and announcement of new mandates • E.g. developments and progress on financial stability made at the G-20, analysis and reforms regarding the Basel II framework • E.g. measures on improving access to financing for SMEs, merger between exchanges

Scope to improve standardization in FSR …… and to be more open about data constraints • FSRs with standardized period of coverage also have standardized timing of release • Most FSRs rarely caution or raise concerns on data gaps issues

Is there an empirical link between FSRs and financial stability? Financial Stability Banking system fragilities Macroeconomic and institutional environment ? FSR quality ? FSR publication

Motivation: diverse country experiences during crisis United States Iceland Latvia Canada Australia No FSR (until 2011) FSR since 2000 (2005) FSR since 2003 FSR since 2002 FSR since 1999 Major crisis Major crisis Major crisis Little crisis impact Little crisis impact

Literature overview Transparency Financial Stability Nier (2005) Bank-level transparency reduces the likelihood of banking problems and thus enhances financial stability Čihák (2006) Evidence generally supportive but limited data. Income level and time since first publication affect FSR quality Oosterloo et al (2007) Little evidence of a direct relationship between FSR transparency and financial stability Born et al (2011) FSR communication about financial stability moves financial stock prices in the expected direction and reduces price volatility

Baseline model (probit/panel with random effects) Dependent variable: Independent variables: • Probability of a banking crisis • Moody’s Banking Sector Financial Strength Rating • Stock market volatility • ICRG sovereign financial risk rating • 1-year median banking system EDF • Two alternative FSR specifications: • FSR publication dummy • FSR quality index (CCC framework as in Čihák, 2006) • MACRO: Macroeconomic controls • BANK: Banking controls • IQ: Institutional quality controls

Heckman model • To address issues of endogeneity/selection bias • Two equation model: • FSRi,t-1= a dummy =1 if country i published an FSR at time t • Crisisi,t-3 is a dummy =1 if country i had a banking crisis (3rd lag) • GDPi,t-1 and Credit/GDPi,t-1 are GDP per capita and credit over GDP at t-1 • FSRNt-1is the total number of countries publishing FSRs at time t-1 • λt is the inverse of the Mills ratio (defined as the normal probability density of the prediction in the first equation, divided by the cumulative normal density). • Meaning of the other variables same as in the baseline model

Summary results Macroeconomic controls Banking controls Other controls Variables of interest

FSRs effectiveness depend on their quality  Higher-quality reports (those that are clear, consistent, and have good coverage) tend to be associated with more stable financial environments; low-quality FSRs seem associated with sub-par financial stability.  Publication of an FSR by itself does not have a robust empirical link to financial stability ! Despite some improvements in recent years, FSRs still leave much to be desired in terms of clarity/coverage/consistency over time

Major areas for improvement ! ! ! Analysis of interconnectedness Lack of a forward-looking nature of the reports Data issues (clarity on what are the “blind spots”)

Areas for further research … • How (by what channels) does FSR quality promote financial stability? • Improved information in the market, stronger market discipline • Signals the FSR provides for policymakers and regulators • Other factors • Which institutional frameworks for FSRs work better than others?

Institutional models for macro-pru (and FSRs) Source: IMF Staff Discussion Note “Institutional Models for Macroprudential Policy” (SDN/11/18)

Institutional framework: key “desirables”… General Central bank to play an important role in macroprudential policymaking. Complex and fragmented regulatory structures should be avoided. Participation of the treasury in the policy process is useful, but a leading role poses risks. Systemic risk prevention and crisis management should be supported by separate organizational arrangements. Macroprudential policy frameworks should not become a vehicle to compromise the autonomy of other established policies. Arrangements need to take account of country-specific circumstances. Provide for effective identification, analysis, and monitoring of systemic risk Mechanisms for effective sharing of information to assess risks should be in place. At least one institution involved in assessing systemic risk should have access to all relevant information. It should be the one that disposes of the best expertise to assess systemic risk. Mechanisms are needed to challenge dominant views of one institution. Provide for timely and effective use of macroprudential policy tools Institutional mechanisms should support willingness to act against the buildup of systemic risk and reduce the risk of delay in policy actions. Source: IMF Staff Discussion Note “Institutional Models for Macroprudential Policy” (SDN/11/18)

Institutional framework: key “desirables” (continued) A lead authority should be identified and be provided with a clear mandate and powers, in a manner that harnesses incentives of existing institutions to mitigate systemic risk. The mandate needs to be matched by powers, including to initiate the use of prudential tools to address systemic risk. Mechanisms should be established to expand powers when needed. The mandate should give primacy to the mitigation of systemic risk, but include secondary objectives to ensure that the policymaker takes into account costs and trade-offs. To guard against overly restrictive or inadequate policy, accountability and transparency need to be put in place, without unduly compromising the effectiveness of macroprudential policy. Provide for effective coordination across policies to address systemic risk Institutional integration of financial regulatory functions within the central bank can support effective coordination of macroprudential policy with monetary as well as microprudential policy, but also requires safeguards. Where separation of policy decisions and control over policy tools cannot be avoided, legal framework needs to assign powers to recommend or direct action of other policymakers. Where there is distributed decision making among several agencies, establishing a coordinating committee is useful, but may not necessarily be sufficient to overcome collective action and accountability problems. Source: IMF Staff Discussion Note “Institutional Models for Macroprudential Policy” (SDN/11/18)

Institutional frameworks in practice Source: Calculations based on 2011 update of World Bank’s Banking Regulation and Supervision Database. Data for 135 jurisdictions.

Thank you!

Financial Stability Reports: What Are They Good For?