Systemic indicators Developing inputs on system-wide risks for financial stability analysis and macroprudential policy Paul Van den Bergh Head of Information, Statistics and Administration Monetary and Economic Department IMF-FSB Users Conference, Washington DC, 8-9 July 2009 Views expressed are those of the author and not necessarily those of the BIS or its associated organisations 1 What are systemic indicators? Early warning indicators Financial stability/vulnerability indicators Financial soundness indicators Macro-prudential indicators Mixture of “individual” indicators and “composite” indicators 2 Commonly used variable in financial stability reports Real economy: GDP, government fiscal position, inflation Corporate sector: debt to equity, earnings (to interest and principal expenses),fx exposure, defaults Household sector: assets and liability positions, income, consumption, debt service levels External sector: (real) exchange rates, fx reserves, CA+capital flows, maturity/currency mismatches Financial sector: money, (real) interest rates, bank credit, bank leverage, NPLs, CDS premia, capital adequacy, liquidity ratio, credit ratings, sectoral/regional concentration of exposures Financial markets: stock index, corporate bond spread, market liquidity, volatility, house prices 3 Macroprudential policy/framework Much work in Basel, still confusion over its definition Two features 1. Focus on financial system as a whole 2. Treat aggregate risk as dependent on collective behaviour of financial institutions (endogenous) Contrast with microprudential which focuses on individual institutions and treats aggregate risk as exogenous For 1: think of financial system as portfolio of securities with each security representing financial institution For 2: think of link credit extension to economic activity, to asset price inflation, to increase in valuation of collateral to credit extension … 4 Macroprudential policy/framework: dimensions Risk distribution at a given point in time (cross-sectional) • Correlation of exposures across institutions (direct or through linkages) • Contribution of individual institution to system-wide risk • Likelihood of failure if others face distress at same time • Vulnerability to risk concentrations even if individual institutions are diversified Risk evolution over time (time dimension) • How can system-wide risk by amplified by interaction financial system and real economy? • Procyclicality • Impact of macroeconomic sources of risk: asset prices, credit, leverage Important implications for calibration of prudential tools 5 Macroprudential policy/framework: cross-sectional Measure likelihood of systemic event at given point in time Use techniques applied to portfolios of securities Data required • Size of institutions • Institution’s probability of default • Loss-given default in each case • Correlation of defaults Information can be collected from supervisory assessments, prices of bank equity and debt Overall level of systemic risk increases with institutions’ exposures to common risk factors 6 Price of insurance against systemic distress1 By financial segment2 Importance of a common driver3 1 Based on credit default swap (CDS) spreads for 10 commercial and eight investment banks headquartered in North America (NA), 16 universal banks headquartered in Europe and 14 insurance companies headquartered in the United States and Euro pe; in per cent. 2 The “Total” line plots the risk neutral expectation of credit losses that equal or exceed 5% of the four financial segments ’ combined liabilities in 2008 (per unit of exposure to these liabilities). Risk neutral expectations comprise expectations of actual losses and preferences. The shaded areas portray how the total is allocated among the four financial segments. The vertical line marks September 2008, the month in which Lehman Brothers filed for Chapter 11 bankruptcy protection. 3 The average share of institutions’ asset return volatility accounted for by a risk factor that is common to all four financial segme nts. Sources: Bankscope; Bloomberg; Markit; BIS calculations. Graph III.1 7 Macroprudential policy/framework: cyclical Countercyclical capital requirements (CR) • Choose indicator that signals time to build up and release capital buffer • Choose formula to determine CR when indicator changes • Adjust actual CR (rule-based or discretionary) Possible indicators • Credit spreads • Real asset prices • Composite indicator combining credit/GDP ratio and real asset prices • Example for US • Variable presented as deviations from long-term average 8 Alternative indicators and charge-off rate in the United States 1 Loans and leases removed from the books and charged against loss reserves, as a percentage of average total loans. 2 Deviation of long-term BBB-rated corporate bond spreads from their long-term average, in basis points. 3 Exponentially weighted five-year average real credit growth minus its 15-year rolling average, in percentage points. 4 Deviation of each variable from its one-sided long-term trend (that is, a trend determined only from information available at the time assessments are made); credit-to-GDP ratio in percentage points, property prices in per cent. Sources: Moody’s; national data; BIS calculations. Graph VII.B.1 Reprioritisation of Financial Soundness Indicators? Significant statistical project developed with wide consultation Large number of participating countries Distinction core and encouraged sets? Core set • various ratio’s of aggregate balance sheet items of deposit takers • concept of consolidation not clearly understood • difficult to integrate data with other banking datasets 10 Reprioritisation of Financial Soundness Indicators? Encouraged set • Includes indicators on indicators for other sectors, financial markets and real estate price indices • More emphasis on non-bank financial sectors? • Financial positions of other sectors through financial account/balance sheet approach? • More emphasis on housing and housing finance indicators (methodology on residential real estate price indices being developed by IWGPS/Eurostat)? 11 Conclusions Need to be clearer about what we want to measure (individual elements of “financial situation”, composites, leading indicators?) Specific set of information requirements for macroprudential policy/framework (cross-sectional and cyclical analysis of aggregate risk; mixture of micro and macro data) Possible to improve FSI’s both in terms of methodology and content (distinction core vs encouraged, better coverage of non-bank sector, more “agile” methodology) 12