## _wp11238 ## Source details **Canonical URL:** [_wp11238](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11238.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11238.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11238.pdf.json) --- ### Executive summary: objective, scope, and framework - Aim: contribute to debate on making macroprudential policy operational following the 2008 financial crisis; task requested by the IMF Board to review cross-country experience in choosing and applying macroprudential instruments and draw lessons on which instruments—and under what conditions—appear most effective. - Data sources: 2010 IMF Survey on financial stability and macroprudential policy and an internal survey of desk economists. - Analytical focus: conditions under which macroprudential policy is most effective; broad range of instruments, risks, and countries. - Macroprudential policy three defining elements: - Objective: limit risk of widespread disruptions to financial services and minimize macroeconomic impact. - Analytical scope: focus on financial system as a whole (including interactions with real sector). - Instruments and governance: primarily prudential tools calibrated to target systemic risk; non-prudential tools must be specifically designated and governed to target systemic risk. ### Instruments most frequently applied - Credit-related: - Caps on the loan-to-value (LTV) ratio - Caps on the debt-to-income (DTI) ratio - Caps on foreign currency lending - Ceilings on credit or credit growth - Liquidity-related: - Limits on net open currency positions/currency mismatch (NOP) - Limits on maturity mismatch - Reserve requirements - Capital-related: - Countercyclical/time-varying capital requirements - Time-varying/dynamic provisioning - Restrictions on profit distribution ### Objectives targeted by instruments - Four broad systemic risk categories instruments seek to mitigate: - Risks from strong credit growth and credit-driven asset price inflation. - Risks from excessive leverage and consequent de-leveraging. - Systemic liquidity risk (aggregate shortage of liquidity preventing short-term funding). - Risks related to large and volatile capital flows, including foreign currency lending. ### Use patterns and institutional context - Two-thirds of IMF survey respondents used various instruments for macroprudential objectives since 2008. - Emerging market economies used instruments more extensively than advanced economies, before and after the crisis. - Typical practices: - Instruments used in combination rather than singly. - Instruments used to complement monetary and fiscal policies. - Instruments adjusted countercyclically to act like “automatic stabilizers.” - Choice of instruments depends on: - Degree of economic and financial development. - Exchange rate regime. - Vulnerability to particular shocks. ### Main empirical assessment (cross-country analysis) - Sample and methods: - Cross-country regression analysis uses data from a group of 49 countries (2000–2010); System GMM (Arellano-Bond) used to address endogeneity and dynamic-panel bias. - Panel regressions examine four measures of systemic risk: credit growth, systemic liquidity (proxy: credit/deposit ratio), leverage (assets/equity), and capital flow/common exposure (proxy: foreign liabilities/foreign assets). - Instruments found to help dampen procyclicality in credit: - Caps on the LTV ratio - Caps on the DTI ratio - Ceilings on credit or credit growth - Reserve requirements - Countercyclical capital requirements - Time-varying/dynamic provisioning - Instruments found to reduce common exposures across institutions and markets: - Limits on net open currency positions/currency mismatch - Limits on maturity mismatch - Key regression magnitudes and statistics (reported exactly as in source): - Regression sample: 49 countries over a 10-year period from 2000 to 2010. - Example coefficients: GDP growth = 0.0791; LTV caps = -0.0634 (first column, Table 1). - Caps on the LTV reduce the procyclicality of credit growth by 80 percent (example magnitude cited in source). - Limits on maturity mismatch associated with a 5 percent lower credit/deposit ratio. - Korea: banks’ short-term external borrowing remained some 30 percent below pre-crisis levels as of 2010. - Introduction of LTV caps associated with "0.06 percent" effect in context reported, leaving an overall net effect of "0.02 percent." - For countries with limits on NOP, foreign liabilities are "15 percent" lower per dollar of foreign assets than in countries without the instrument (Table 3). - Additional findings: - No evidence that exchange rate regime or financial sector size affects instrument effectiveness (dummy coefficients statistically insignificant). - Rules-based instruments have larger effects in regressions. - Insufficient granular data to resolve whether single or multiple instruments are more effective on average. ### Before-and-after and case-study evidence (selected outcomes) - Simple before-and-after analysis (country-level averages): - Caps on the LTV: credit growth and asset price inflation decline after implementation in more than half of sample countries. - Caps on the DTI: credit growth declines but asset price inflation does not. - Dynamic provisioning: credit growth and asset price inflation, and to a lesser extent leverage growth, decline. - Reserve requirements: both credit growth and asset price inflation decline. - Macroprudential instruments tend to reduce correlation between credit growth and GDP growth; some instruments sometimes reverse correlation to negative. - Case-study highlights: - China: 2010 steps (fiscal, monetary, administrative measures) lowered credit growth and housing price inflation; bank lending growth slowed to 16.9 percent (yoy) in June 2011 from 31.7 percent in December 2009. - Colombia: 1999 package (LTV 70 percent; DTI cap 30 percent; NOP limits) reduced non-performing loans; private sector credit initially fell then recovered. - Eastern Europe: FX lending caps helped slow credit growth and build buffers but were partly circumvented by migration to nonbanks and cross-border lending. - Spain: Dynamic provisioning (introduced 2000) helped cover rising credit losses during the global financial crisis though coverage was less than full. - Korea: post-crisis measures curtailed banks’ short-term external borrowing; short-term external borrowing remained some 30 percent below pre-crisis levels as of 2010. - New Zealand: liquidity mismatch ratios and a core funding ratio (2010) led banks to lengthen wholesale funding even before formal implementation; short-term debt fell from 64 percent of GDP in December 2008 to 50 percent in December 2010. ### Model-based and simulation evidence (Box 2) - New Keynesian open-economy DSGE model with firms financing via retained earnings or borrowing. - Key simulation welfare losses (sum of inflation and output volatilities in percent of steady state consumption): - Taylor rule (monetary policy only): 2.5 - Taylor rule plus macroprudential policy: 1.3 - Macroprudential measures alone with policy interest rate unchanged: 31.5 - Conclusion: combination of monetary and macroprudential policies is superior to stand-alone policies. ### Policy-design lessons and comparative advantages - Single versus multiple instruments: - Multiple instruments can tackle different aspects of the same risk, reduce circumvention, and increase assurance of effectiveness; avoid excessive complexity and costs. - Broad-based versus targeted: - Targeted measures improve precision but require granular data and raise administrative cost and circumvention risk; broad-based measures useful when data are limited. - Fixed versus time-varying: - Time-varying (countercyclical) adjustments improve smoothing of financial cycles; fixed instruments provide minimum buffers at lower administrative cost. - Rules versus discretion: - Rules-based adjustments (e.g., dynamic provisioning) reduce policy inertia and increase predictability; discretion needed where rules are infeasible, but should be accompanied by formal analysis and clear public communication. - Coordination with other policies: - Instruments more effective when used with monetary or fiscal tools; stand-alone macroprudential policies tend to be inferior. - Mechanisms to resolve conflicts and clear governance/ accountability arrangements are essential. ### Caveats, costs, and preconditions - Costs and calibration: - Using macroprudential instruments entails costs; benefits must be weighed against regulatory costs and potential distortions. - Calibration is difficult; miscalibration could lower growth unnecessarily or create unintended distortions. - Empirical analysis does not address cost-benefit or calibration trade-offs. - Data and arbitrage: - Regulatory and cross-border arbitrage and data gaps limit analysis—particularly for cross-sectional systemic risk and interconnectedness. - Better quality, more granular, and longer time series data are needed. - Institutional preconditions: - Sound regulatory framework, high-quality supervision, good macroeconomic policies, and an institutional framework for macroprudential policy and microprudential coordination. ### Open questions and research gaps - Address regulatory and cross-border arbitrage. - Fill data gaps for cross-sectional systemic risk analysis and interconnectedness of global systemically important institutions. - Understand side-effects and leakages from macroprudential measures (migration to nonbank sector, disintermediation). - Clarify relationship between macroprudential and microprudential regulation and coordinate objectives. - Better quantify costs of instruments and optimal calibration (insurance vs. imbalance-correction; price- vs. quantity-based instruments). ### Appendix I — macroprudential instruments that may also be considered capital flow measures - “Hybrid cases” exist where macroprudential instruments may also function as capital flow measures (CFMs); primary objective clarity is essential. - EU context: - ESRB established as of January 1, 2011; fosters warnings of macroprudential risks and “reciprocity” through “comply or explain” powers. - European Commission focuses on countercyclical capital; other agencies propose wider scope to capture regional/national/sectoral conditions. - Advantages of macroprudential instruments over other policies: - Less blunt than monetary tools; more flexible with smaller lags than most fiscal tools. - Can be targeted to sectors or loan portfolios without generalized reduction in activity. - Some instruments (e.g., caps on foreign currency lending) directly target excessive FX lending effects. - Factors affecting instrument choice: stage of development, exchange rate regime, and type of shocks (examples: emerging markets favor liquidity measures; fixed/managed exchange rate countries use instruments more when interest rate policy is constrained). ### Appendix II — selected case studies (highlights) - European selected countries (Bulgaria, Croatia, Poland, Romania, Serbia): - Credit/GDP increases: Croatia +19 percentage points; Bulgaria +45 percentage points (mid-2000s). - Measures: LTV, DTI, reserve requirements differentiated by currency/maturity, higher risk weights, provisioning, lending ceilings; circumvention via nonbanks and cross-border channels was common. - Outcome: generally slowed credit growth and built buffers; success contingent on broader macro policy mix and enforcement. - New Zealand: - Instruments: liquidity mismatch ratios; core funding ratio (65% April 2010 → 70% July 2011 → 75% planned July 2012). - Outcome: short-term debt fell from 64 percent of GDP in December 2008 to 50 percent in December 2010. - Spain: - Dynamic provisioning introduced in 2000; helped cover rising credit losses but did not fully cover actual losses; calibration relaxed in 2005. - China: - 2010–2011 package: LTV tightening, capital buffers raising minimum CAR to 11.5 percent from 8 percent for large banks, provision coverage ratio raised from 100 percent to 150 percent, reserve requirement raised nine times for total 450 basis points, benchmark lending rate raised five times (total 125 basis points). - Outcome: bank lending growth slowed to 16.9 percent (yoy) in June 2011 from 31.7 percent in December 2009; home-sales and prices moderated. - Korea, Colombia, United States: selected outcomes described in Appendix II with instrument-specific impacts (see case summaries above). ### Appendix III–IV — empirical methods and robustness - Simple approach: event-window country averages around instrument implementation (quarters t-2 to t+4) show average declines in targeted risk metrics for several instruments. - GMM methodology (System GMM / Arellano-Bond): - Fixed-effect dynamic panel with lagged dependent variables, instrument dummies I, controls X (GDP growth, interest rate), and interaction terms IX. - Estimation challenges: endogeneity, dynamic-panel bias; addressed via System GMM and robustness checks (restricted instruments, OLS fixed effects). - Main significant interaction results: negative GDP-growth interactions (reducing procyclicality) for caps on LTV, caps on DTI, ceilings on credit growth, reserve requirements, dynamic provisioning; countercyclical capital requirements also significant in some specifications. - Table-level examples (coefficients and significance reported exactly as in source tables): - Table IV.1 examples: Caps on Loan-to-Value (dummy): 0.0823 (2.65)*** in credit-growth-equation constant dummy; Dynamic Provisioning (dummy): -0.1466 (-22.57)*** in another specification. - Table IV.3 interaction examples: Caps on Loan-to-Value 2 × GDP Growth: -0.0615 (-2.59)**; Caps on Debt-to-Income 2 × GDP Growth: -0.0637 (-2.44)**; Reserve Requirements 2 × GDP Growth: -0.0448 (-2.18)***; Countercyclical Capital Requirements 2 × GDP Growth: -0.1563 (-2.57)***. - Table IV.4 interaction examples: Dynamic Provisioning 2 × GDP Growth: -0.2765 (-3.75)***; Reserve Requirements 2 × GDP Growth: -0.0937 (-4.39)***; Caps on Debt-to-Income 2 × GDP Growth: -0.0526 (-2.69)***. - Diagnostics: Arellano-Bond tests for autocorrelation passed; Sargan test weakened by many instruments; results robust to restricted instrument sets and OLS fixed effects. ### Appendix V — operational details of selected instruments - Loan-to-Value (LTV): - Design features: often combined with other tools (DTI, reserve requirements); partial targeting by property value, mortgage purpose, or borrower type; used fixed or countercyclically; discretionary adjustments common. - Effectiveness: statistical evidence of clear effect on credit growth and property prices; may wear off in dynamic markets. - Dynamic provisioning: - Four main systems: continuous benchmark provisions (Spain, Uruguay); activation mechanisms accumulating in upswing and drawing down in downturn (Colombia, Peru); provisioning by debtor risk classification (Chile, Mexico); discretionary countercyclical provisioning (Bulgaria, Croatia, India, Mongolia, Russia). - Effectiveness: rules-based DP systems effective; discretion-based systems less so. DP smoothes provisioning costs and builds buffers but does not cover all unexpected losses. - Reserve requirements: - Used as macroprudential tools in emerging markets; targeted by currency/maturity; can be raised to very high rates; effective in reducing procyclicality but subject to circumvention. - Reported ranges of RR across selected countries (percent): Argentina: 0-100; Bulgaria: 4-400; China: 6-21.5; Colombia: 0-140; Croatia: 13-55; Indonesia: 1-8; Lebanon: 15-25; Romania: 0-40; Serbia: 5-100; Peru: 0-120; Turkey: 5-16; Uruguay: 25-35. - FX lending measures: - Types: direct caps, DTI by currency, bans, higher risk weights, provisioning, limits relative to capital. - Effectiveness: mixed when tested in isolation; net open position limits effective in reducing external indebtedness; risks include migration to nonbanks and FX-indexed lending. - Selected numeric examples for FX and reserve metrics appear in Appendix V tables (examples preserved above). ### Appendix VI — conceptual basis and country experience highlights - Conceptual rationale summarized for each instrument (LTV, DTI, FX lending caps, ceilings on credit growth, NOP limits, maturity mismatch limits, reserve requirements, countercyclical capital, dynamic provisioning, restrictions on profit distribution). - Country vignettes provide timelines and specific measures for Argentina, Austria, Brazil, Bulgaria, Canada, Chile, China, Colombia, Croatia, France, Greece, Hong Kong, Hungary, India, Indonesia, Ireland, Italy, Korea, Lebanon, Malaysia, Mexico, Mongolia, New Zealand, Nigeria, Norway, Peru, Poland, Portugal, Romania, Russia, Serbia, Singapore, Slovakia, South Africa, Spain, Sweden, Switzerland, Thailand, Turkey, Uruguay, and others — documenting combinations of LTV, DTI, RR, provisioning, NOP, maturity limits, capital surcharges, liquidity metrics, and targeted measures. *Source: IMF staff analysis, Executive Summary and Appendixes of _wp11238.* ### Executive Summary ...................................................................................................... ### _wp11238 - Executive Summary ### Objective and scope - Aim: contribute to the international debate on making macroprudential policy operational following the 2008 financial crisis. - Task requested by the IMF Board: review cross-country experience in choosing and applying macroprudential instruments and draw lessons on which instruments—and under what conditions—appear most effective. - Data sources used: 2010 IMF Survey on financial stability and macroprudential policy and an internal survey of desk economists. - Analytical focus: identifying conditions under which macroprudential policy is most effective, covering a broad range of instruments, risks, and countries. ### Definition and organizing framework - Macroprudential policy characterized by three defining elements: - Objective: limit the risk of widespread disruptions to the provision of financial services and thereby minimize the impact on the economy as a whole. - Analytical scope: focus on the financial system as a whole (including interactions with the real sector) rather than individual components. - Instruments and governance: primarily prudential tools designed and calibrated to target systemic risk; non-prudential tools must be specifically designated and governed to target systemic risk. ### Instruments used (ten most frequently applied) - Credit-related: - Caps on the loan-to-value (LTV) ratio - Caps on the debt-to-income (DTI) ratio - Caps on foreign currency lending - Ceilings on credit or credit growth - Liquidity-related: - Limits on net open currency positions/currency mismatch (NOP) - Limits on maturity mismatch - Reserve requirements - Capital-related: - Countercyclical/time-varying capital requirements - Time-varying/dynamic provisioning - Restrictions on profit distribution ### Objectives targeted by instruments - Instruments have been used to mitigate four broad categories of systemic risk: - Risks generated by strong credit growth and credit-driven asset price inflation. - Risks arising from excessive leverage and consequent de-leveraging. - Systemic liquidity risk (aggregate shortage of liquidity preventing short-term funding). - Risks related to large and volatile capital flows, including foreign currency lending. ### Use patterns and institutional context - Two-thirds of IMF survey respondents have used various instruments for macroprudential objectives since 2008. - Emerging market economies used instruments more extensively than advanced economies, both before and after the recent financial crisis. - Countries often: - Use instruments in combination rather than singly. - Use instruments to complement other macroeconomic policies. - Adjust instruments countercyclically so they act like “automatic stabilizers.” - The choice of instruments often depends on: - Degree of economic and financial development. - Exchange rate regime. - Vulnerability to certain shocks. ### Empirical assessment of effectiveness - Cross-country regression analysis uses data from a group of 49 countries. - Instruments suggested to help dampen procyclicality in credit: - Caps on the LTV ratio - Caps on the DTI ratio - Ceilings on credit or credit growth - Reserve requirements - Countercyclical capital requirements - Time-varying/dynamic provisioning - Instruments suggested to reduce common exposures across institutions and markets: - Limits on net open currency positions/currency mismatch - Limits on maturity mismatch - Effectiveness observations: - Effectiveness does not appear to depend on exchange rate regime nor the size of the financial sector. - The type of shocks matters: different types of risks call for the use of different instruments. - Analytical approaches used in the paper include case studies, a “simple approach,” and panel regressions (including GMM methodology described in appendixes). ### Policy design lessons and comparative advantages - Single versus multiple: - Multiple instruments can tackle different aspects of the same risk, reduce scope for circumvention, and increase assurance of effectiveness. - Avoid excessive complexity and imposing costs that are too high. - Broad-based versus targeted: - Targeting specific risks by differentiating transaction types improves precision and potential effectiveness. - Use broad-based measures when granular data are not available and risks are generalized; supplement with targeted measures to limit circumvention. - Fixed versus time-varying: - Adjusting instruments over the cycle improves effectiveness in smoothing the financial cycle. - Rules versus discretion: - Rules-based adjustments (e.g., dynamic provisioning) have clear advantages and are effective. - Rules can be difficult to design for some instruments; policymakers need discretion and clear public communication when making discretionary adjustments. - Coordination with other policies: - Instruments are more effective when used in conjunction with monetary or fiscal policy tools; stand-alone policies tend to be inferior. - Establish mechanisms to resolve conflicts and assign clear accountability and governance arrangements. ### Caveats, costs, and preconditions - Costs and calibration: - There are costs in using macroprudential instruments (as with regulation generally); benefits must be weighed against these costs. - Calibrating instruments may be difficult and could lower growth unnecessarily or generate unintended distortions if not done appropriately. - The empirical analysis does not address cost-benefit or calibration trade-offs. - Data and arbitrage issues: - Regulatory or cross-border arbitrage and data gaps limit analysis—especially of cross-sectional systemic risk. - Better quality, more granular, and longer time series data are needed to corroborate initial assessments and confirm causal relationships. - Institutional preconditions for successful implementation: - Strong regulatory framework. - High-quality supervision. - Good macroeconomic policies. - Appropriate institutional framework for macroprudential policy and coordination with microprudential regulation. ### Open questions and next steps - Remaining issues to be answered: - Addressing regulatory or cross-border arbitrage. - Filling data gaps to allow careful cross-sectional systemic risk analysis. - Understanding side-effects of applying macroprudential instruments. - Clarifying the relationship and coordination between macroprudential policy and microprudential regulation. - The paper identifies further work needed and points to appendixes covering methodology, case studies, and conceptual foundations for the instruments. *Source: IMF staff analysis, Executive Summary of _wp11238.* ### Appendix I shows some macroprudential instruments that may also be considered capital flow measures ### _wp11238 - Appendix I shows some macroprudential instruments that may also be considered capital flow measures ### Hybrid cases and primary-objective clarity - Appendix I identifies some macroprudential instruments that may also be considered capital flow measures (CFMs). - In these “hybrid cases,” clarity of the primary objective of the macroprudential instrument is important to ensure the policy is used appropriately to target systemic risk, and not the exchange rate or capital flows. - Macroprudential instruments should not be confused with capital controls. ### Macroprudential Instruments in the European Union - Work on selecting and applying macroprudential instruments is a priority in the European Union (EU), both at a national and at a Union level. - The European Systemic Risk Board (ESRB) was established as of January 1, 2011, to provide warnings of macroprudential risks and to foster the application of macroprudential instruments. - The ESRB has an additional role to foster “reciprocity” through its “comply or explain” powers amongst the national authorities, so that all banks conducting a particular activity in a country will be subject to the same macroprudential instrument irrespective of the bank’s home country. - The European Commission has been focusing on countercyclical capital as the main macroprudential instrument. - Other agencies and some national authorities propose a wider scope to account for regional, national, sub-national, or sectoral conditions. - Example: Real estate lending has been central to past financial crises, suggesting a focus on instruments such as the loan-to-value ratio (LTV). ### Why use macroprudential policy and factors affecting instrument choice - Advantages of macroprudential policy versus other public policies: - Less blunt than monetary tools. - More flexible with smaller implementation lags than most fiscal tools. - Many instruments (e.g., caps on the LTV, DTI, foreign currency lending, and capital risk weights) can be tailored to specific sectors or loan portfolios without causing a generalized reduction of economic activity. - Some instruments (e.g., caps on foreign currency lending) can target excessive lending in foreign currency directly in a way that no other policies can. - Especially useful when tightening monetary policy is not desirable (e.g., when inflation is below target). - Authorities prefer instruments that are simple, effective, and easy to implement with minimal market distortions. - Instrument choice should be consistent with other public policy objectives (fiscal, monetary, and prudential). - Importance of minimizing regulatory arbitrage, particularly in advanced economies with large nonbank financial sectors and complex, highly interconnected financial systems. - Factors influencing instrument choice: - Stage of economic and financial development: - Emerging market economies have used macroprudential instruments more extensively than advanced economies. - Emerging market economies are more concerned about systemic liquidity risk and tend to use liquidity-related measures more often. - Advanced economies tend to favor credit-related measures, although more are beginning to use liquidity-related measures after the recent crisis.10 11 - Exchange rate regime: - Countries with fixed or managed exchange rates tend to use macroprudential instruments more since the exchange rate arrangement limits the room for interest rate policy. - In these countries, credit growth tends to be associated with capital inflows as the implicit guarantee of the fixed exchange rate provides an incentive for financial institutions to expand credit through external funding.12 - Credit-related measures (e.g., caps on the LTV and ceilings on credit growth) are often used when interest rate use is constrained. - Liquidity-related measures (e.g., limits on NOP) are used to manage external funding risks. - Type of shocks: - Capital inflows are considered by many emerging market economies to be a shock with a large impact on the financial sector. - Eastern European countries have used credit-related measures (e.g., caps on foreign currency lending) to address excessive credit growth resulting from capital inflows. - In Latin America, several countries (e.g., Argentina, Brazil, Colombia, Peru, and Uruguay) have used liquidity-related measures (e.g., limits on NOP) to limit the impact of capital inflows. - In the Middle East, some oil exporters with fixed exchange rates have used credit-related measures to deal with the impact of volatile oil revenue on credit growth. - Unlike other policy tools aimed at the volume or composition of flows (e.g., taxes, minimum holding periods), macroprudential instruments are more directly aimed at the negative consequences of inflows: excessive leverage, credit growth and exchange rate induced credit risks that are systemic. ### How instruments are applied - Country experiences show that a combination of several instruments is often used to address the same risk. - Examples: - Caps on the LTV and DTI are frequently applied together to curb rapid credit growth in the real estate sector. - Using a single instrument to address systemic risk is rare. - Rationale for multiple instruments: - To provide greater assurance of effectiveness by tackling a risk from various angles. - Potential trade-off: higher regulatory and administrative burden of enforcing multiple instruments. - Figure 3 and Figure 4 are referenced for empirical patterns and ranges of measures, respectively (figures not reproduced here). *Source: _wp11238 - Appendix I shows some macroprudential instruments that may also be considered capital flow measures* ### Appendix II contains case studies of countries using the instruments. ### _wp11238 - Appendix II contains case studies of countries using the instruments. ### Use and calibration of macroprudential instruments - The survey reports only two countries that use single instruments, and for one of them (Canada), the objective of using the LTV is microprudential. - Many instruments, particularly credit-related, are calibrated to target specific risks. - Caps on the LTV and DTI have been applied according to the loan size, the location and the value of the property (Hong Kong SAR and Korea). - Reserve requirements used for macroprudential purposes have been differentiated by currency, types of liabilities, and applied within a band or on a marginal basis, or if credit growth exceeds the official limit (Argentina, Chile, China, Indonesia, Peru, Russia, Serbia, and Turkey). - Social and developmental aspects have been taken into account when calibrating instruments in some cases (Canada). - Instruments are applied both in targeted and broad-based ways; some countries apply instruments flexibly/time-varying, others keep them fixed. - Countercyclical adjustments are common: - Caps on the LTV, DTI and reserve requirements are adjusted most frequently. - Capital-related measures (countercyclical capital requirements and dynamic provisioning) are designed to work through the cycle by providing a buffer, but some countries adjust them at different phases to increase countercyclical impact. - Design and calibration are usually based on discretion and judgment rather than rules; a few exceptions use rules-based formulas (dynamic provisioning in Spain and several Latin American countries). - Macroprudential instruments are sometimes used in conjunction with other macroeconomic policies (monetary and fiscal). Examples include China, Hong Kong SAR, Singapore (taxes on real estate transactions combined with lowering LTVs), and several Eastern European countries. ### A. Case study evidence (selected countries) - Sample: China, Colombia, Korea, New Zealand, Spain, the United States and some Eastern European countries (case studies in Appendix II). - Country-specific summarized outcomes: - China: Steps in 2010 (including fiscal and monetary measures) managed to lower credit growth and housing price inflation. - Colombia: Measures in 1999 to limit banks’ exposure to default risk — non-performing loans declined and remained low while credit to the private sector recovered after an initial reduction. - Eastern Europe: Measures to curb bank foreign currency lending appeared effective in slowing credit growth and building capital and liquidity buffers, but were partly circumvented by migration of lending to nonbanks and cross-border lending by parent banks. - Spain: Dynamic provisioning introduced in 2000 helped cover rising credit losses during the global financial crisis, though coverage was less than full given the severity of losses. - Korea: Measures post-crisis curtailed banks’ short-term external borrowing, which remained some 30 percent below its pre-crisis levels as of 2010. - New Zealand: Two liquidity mismatch ratios and a core funding ratio introduced in 2010; banks lengthened wholesale funding structure after announcement even before formal implementation. - United States: Minimum leverage ratio for banks adopted in 1991; not adjusted over time and did not apply to investment banks after 2004, contributing to rising leverage at investment banks while commercial bank leverage remained lower. ### B. Simple before-and-after analysis - Approach: Examine target risk variables before and after instrument implementation (see Appendix III for charts). - Findings on specific instruments: - Caps on the LTV: credit growth and asset price inflation decline after implementation in more than half of sample countries. - Caps on the DTI: credit growth declines but asset price inflation does not. - Dynamic provisioning: credit growth and asset price inflation, and to a lesser extent leverage growth, decline. - Reserve requirements: both credit growth and asset price inflation decline. - Macroprudential instruments appear to reduce the correlation between credit growth and GDP growth: - In countries with caps on the LTV, DTI and reserve requirements, the correlation is positive but much smaller than in countries without them. - In countries with ceilings on credit growth or dynamic provisioning, the correlation between credit growth and GDP growth becomes negative. - Caveats: data availability/quality, small number of systematic users, selection bias, and absence of cost/distortion analysis limit causal inference. ### C. Panel regression analysis (49 countries, 2000–2010) - Objectives: Estimate effects of instruments on four measures of systemic risk—credit growth, systemic liquidity, leverage, and capital flows—while controlling for monetary and fiscal policy and country characteristics. - Instruments evaluated (dummy variables): caps on the LTV, caps on the DTI, caps on foreign currency lending, ceilings on credit or credit growth, reserve requirements, countercyclical/time-varying capital requirements, time-varying/dynamic provisioning, restrictions on profit distribution; limits on net open positions and limits on maturity mismatch assessed for liquidity/capital flow proxies. - Controls and methodology: - Monetary policy: interest rate variable included. - Fiscal policy: GDP growth used as proxy. - Country characteristics: dummy variables for exchange rate regime, size of financial sector, degree of economic development; fixed effects capture other unobserved country-specific factors. - Estimation method: System Generalized Method of Moments to address endogeneity and dynamic panel bias. - Regression results (summary): - Credit growth (yoy change in inflation-adjusted claims on the private sector): - Five of the 10 instrument dummies are statistically significant: caps on the LTV, caps on the DTI, ceilings on credit growth, reserve requirements and time-varying/dynamic provisioning. - Example magnitude: caps on the LTV reduce the procyclicality of credit growth by 80 percent. (Coefficients cited: GDP growth = 0.0791; LTV caps = -0.0634; first column, Table 1.) - The dummy for countries that have adjusted LTV caps over time is also significant. - Systemic liquidity (proxy: credit/deposit ratio as wholesale funding measure): - Limits on maturity mismatch are statistically significant; credit/deposit ratio is 5 percent lower in countries with the instrument than in countries without it. - Leverage (assets/equity): - Six of the 10 instrument dummies are statistically significant: caps on the DTI, ceilings on credit growth, reserve requirements, caps on foreign currency lending, countercyclical/time-varying capital requirements, and time-varying/dynamic provisioning. - Dynamic provisioning reduces the procyclicality of both credit growth and leverage. - Limited observations for some capital-related measures reduce ability to detect effects. - Capital flows and common exposure (proxy: foreign liabilities/foreign assets): - Effects estimated for limits on net open positions and limits on maturity mismatch (scope limited by data availability). - Additional findings: - No evidence that degree of economic development, exchange rate regime type, or financial sector size affects instrument effectiveness (dummy coefficients statistically insignificant), though these factors may influence instrument choice. - Instruments remain effective after controlling for macroeconomic policies. - Rules-based instruments have a larger effect in the regressions. - Insufficient granular data to determine whether individual or multiple instruments are more effective. ### Key statistics and sample notes - Regression sample: 49 countries over a 10-year period from 2000 to 2010. - Korea: banks’ short-term external borrowing remained some 30 percent below pre-crisis levels as of 2010. - Panel regression example coefficients: GDP growth = 0.0791; LTV caps = -0.0634 (first column, Table 1). - Limits on maturity mismatch associated with a 5 percent lower credit/deposit ratio. - Note on scoring/usage intensity (survey visualization): 0 represents no use of instruments, and 1 denotes the use of a single instrument. For attributes multiple, targeted, time-varying, discretionary and used in coordination with other policies, the value of 1 is added. Percent of the sample and total number of countries using instruments shown in Figure 5 (survey results). *Italic: Source: IMF Financial Stability and Macroprudential Policy Survey, 2010; Appendix II case studies summarized in _wp11238 - Appendix II contains case studies of countries using the instruments.* ### 0.06 percent when LTV caps are introduced, leaving an overall net effect of 0.02 percent. ### _wp11238 - 0.06 percent when LTV caps are introduced, leaving an overall net effect of 0.02 percent. ### Key empirical findings from panel regressions - Introduction of LTV caps is associated with "0.06 percent" effect in the context reported, leaving an overall net effect of "0.02 percent." - The only dummy variable with a statistically significant coefficient in the reported regression is limits on NOP. - For countries with limits on NOP, the regression suggests: for every dollar of foreign assets held, foreign liabilities are "15 percent" lower than in countries without this instrument (Table 3). - Regression coefficients are averages of country performances; magnitudes depend on: - the number of countries in the sample that have used the instruments, and - the effectiveness of instruments in individual countries. ### Confirmation from other studies and models - A separate study focusing on structural determinants of credit growth corroborates the panel analysis findings (see IMF (2011g)). That study, using a different model and endogeneity assumptions, finds that: - caps on the DTI, - caps on foreign currency lending, - reserve requirements, and - time-varying/dynamic provisioning each have a negative sign on credit to GDP and are statistically significant. - A structural model used in IMF (2011h) confirms that the effectiveness of macroprudential instruments does not depend on the exchange rate regime: the impact is "virtually identical" in economies with either fixed or floating exchange rates. ### Interpretation caveats and limitations - Regression results should be interpreted with caution: - Statistical significance of dummy-variable coefficients does not indicate uniform effectiveness across countries. - Country-specific circumstances matter, including: - quality of supervision, - phase of the credit cycle when instruments are implemented, - extent to which circumvention and arbitrage are possible, - ability of authorities to take coordinated policy actions to limit circumvention, and - responsiveness of authorities to changed conditions. - The use of macroprudential instruments is relatively new, limiting the number of observations and the length of available time series for comprehensive evaluation. - Further research is needed with longer time series and better quality data to corroborate initial assessments and evaluate effectiveness in country-specific contexts. - Future analysis should explicitly account for: - costs involved in using macroprudential instruments, - degree of calibration required, and - potential for regulatory and cross-border arbitrage, which can circumscribe effectiveness. *Source: _wp11238 - 0.06 percent when LTV caps are introduced, leaving an overall net effect of 0.02 percent.* ### Box 2. Monetary and Macroprudential Policy: Are They Mutually Reinforcing? ### Box 2. Monetary and Macroprudential Policy: Are They Mutually Reinforcing? ### Model setup and shock scenario - Open-economy, New Keynesian DSGE model with firms financing investment via retained earnings or borrowing from domestic or foreign sources. - Macroprudential policy modeled as an additional “regulation premium” raising firms’ borrowing costs. - Monetary policy follows a Taylor rule reacting to inflation and output gaps. - Initial shock: a decline in investors’ perception of risk, triggering capital inflows, a decline in financing costs, higher borrowing and investment, and eventual higher leverage that raises risk premia and normalizes financial conditions. - Both monetary and macroprudential policies are analyzed for their ability to mitigate the shock’s effects. ### Key simulation findings - Macroprudential measures that directly counteract increased leverage and eased underwriting standards mute domestic and foreign debt responses to the shock. - Output and inflation responses are smaller when macroprudential measures are used alongside monetary policy. - Welfare loss (computed as the sum of inflation and output volatilities in percent of steady state consumption): - Taylor rule (monetary policy only): 2.5 - Taylor rule plus macroprudential policy: 1.3 - Macroprudential measures alone with policy interest rate unchanged: 31.5 - Conclusion: the combination of monetary and macroprudential policies is superior to stand-alone policies. ### Empirical evidence on instrument effectiveness (sample and estimation) - Estimation period: 2000–2010. - Sample: 48 countries (note: elsewhere the paper draws on a sample of 49 countries for broader analysis). - Dependent variables and data sources: - Credit growth: log change in real level of credit (claims on private sector from bank and non-bank financial institutions; source: IFS). - Leverage growth: log change in level of leverage (assets over capital; source: IMF FSIs). - Cross-sectional risk measures: ratio of financial system liabilities with foreign residents to claims on foreign residents; ratio of banking institutions’ claims to deposits (source: IFS). - Interest rate: nominal long-term interest rate on prime lending (IFS). - Estimation approach: regressions include dummy variables for exchange rate flexibility, country fixed effects, time trend, dummy for use of other MPP instruments; instrumental variables for policy instruments and GMM Arellano-Bond estimator to address selection bias and endogeneity. ### Empirical results — which instruments reduce procyclicality of credit and leverage - Credit procyclicality (interaction of GDP growth with instrument dummy): - Instruments associated with a statistically significant negative interaction (reducing procyclicality of credit) include: - Caps on Loan-to-Value (LTV) × GDP Growth (coefficient reported as negative and significant) - Caps on Debt-to-Income (DTI) × GDP Growth - Limits on Credit Growth × GDP Growth - Reserve Requirements × GDP Growth - Dynamic Provisioning × GDP Growth - Some instruments (e.g., Limits on Forex Lending × GDP Growth, Countercyclical Capital Requirements × GDP Growth) show weak or non-significant effects in some specifications. - Leverage procyclicality (interaction of GDP growth with instrument dummy): - Statistically significant negative interactions (reducing procyclicality of leverage) include: - Caps on Debt-to-Income × GDP Growth - Limits on Credit Growth × GDP Growth (weakly significant) - Reserve Requirements × GDP Growth - Dynamic Provisioning × GDP Growth - Countercyclical Capital Requirements × GDP Growth - Restrictions on Profit Distribution × GDP Growth sometimes show positive coefficients (statistically significant) in the leverage regressions. - Cross-sectional risk reduction: - Limits on Net Open Positions in Foreign Currency are associated with a negative and statistically significant effect on foreign liabilities / foreign assets (reducing dependence on foreign funding). - Limits on Maturity Mismatch are associated with a negative and statistically significant effect on cross-sectional liquidity risk. ### Lessons and policy messages - A number of instruments can be effective in addressing systemic risks; effectiveness does not appear to depend strongly on stage of development or exchange rate regime. - Emerging market economies with fixed or managed exchange rates tend to use macroprudential instruments more extensively, but instruments appear equally effective when used by countries with flexible exchange rate regimes and by advanced economies. - Costs and calibration challenges: - There are costs to using macroprudential instruments; benefits should be weighed against regulatory costs and potential distortions. - Calibrating instruments is difficult; miscalibration could lower growth unnecessarily or generate unintended distortions. - Institutional prerequisites: - Effectiveness rests on a sound regulatory framework and high-quality supervision. - Institutional arrangements should ensure policymakers’ ability and willingness to act, including clear mandates, control over instruments commensurate with mandates, operational independence safeguards, accountability, transparency, and clear communication. - Instrument choice by risk type: - Credit growth or asset price inflation: credit-related instruments (LTV, DTI caps), possibly supplemented by reserve requirements or capital-related instruments (dynamic provisioning); targeting by currency if foreign currency lending is the risk source. - Systemic liquidity risk: liquidity-related instruments (limits on liquidity mismatch), limits on net foreign currency positions; core funding ratios or levies on non-core liabilities (not examined in this paper) could be candidates. - Excessive leverage: capital-related instruments (adjustable capital requirements, risk weights, provisioning); may be supplemented by credit-related instruments. - Risks from capital flows: use of liquidity-, credit-, and capital-related instruments in combination (e.g., limits on net open positions in foreign currency to limit dependence on foreign funding). - Design and calibration considerations (five dimensions): - Single versus multiple instruments: multiple instruments can tackle a risk from various angles and reduce circumvention but impose higher costs and complexity. - Broad-based versus targeted: targeting can be more precise but requires granular data, higher administrative cost, and may increase circumvention risk. - Fixed versus time-varying: time-varying (countercyclical) instruments can smooth cycles but require transparent principles and careful timing; fixed instruments provide minimum buffers at low administrative cost. - Rules versus discretion: rules-based instruments (e.g., dynamic provisioning, capital conservation buffer) reduce policy inertia and increase predictability; many instruments will require constrained discretion and formal analysis when rules are infeasible. - Coordination with other policies: coordination with fiscal and monetary policy enhances effectiveness; macroprudential policy should complement monetary policy and be primary when risks are sector- or funding-specific. Mechanisms to resolve conflict and clear accountability are essential. ### Implementation guidance and operational issues - Use multiple instruments judiciously to minimize circumvention while avoiding excessive costs on regulated institutions. - When targeting, supplement with broader-based measures where appropriate to limit circumvention and avoid excessive complexity. - Prefer rule-based instruments where feasible to reduce inertia and improve predictability; where discretion is necessary, base actions on formal analysis and communicate rationale publicly. - Establish a macroprudential framework to identify and monitor systemic risk, define procedures for instrument use, select specific objectives, and evaluate effectiveness. - Ensure coordination mechanisms among macroprudential, monetary, and fiscal authorities to prevent conflicts and ambiguity in accountability. ### Next steps and research gaps identified - More work needed on cross-sectional dimension of systemic risk; data limitations constrain analysis of interconnectedness of global systemically important institutions. - Need for more granular data to identify and monitor systemic risk and make macroprudential policy operational. - Deeper understanding required of design and calibration issues, including whether instruments should be insurance versus imbalance-correction tools, and whether price-based or quantity-based instruments are more effective. - Further research needed on costs of implementing macroprudential instruments (regulatory burden, distortions, unintended consequences, migration of systemic risk or “leakages”). - Clarify the relationship between macroprudential and microprudential regulation to coordinate objectives and avoid conflicts when prudential tools begin serving macroprudential purposes. *Source: IMF staff analysis.* ### APPENDIX I. MACROPRUDENTIAL OR CAPITAL FLOW MEASURES? ### APPENDIX I. MACROPRUDENTIAL OR CAPITAL FLOW MEASURES? ### Overview - Many countries have recently undertaken measures that can be considered both macroprudential and capital flow management measures (CFMs). - Common theme: concerns of preserving financial stability and macroeconomic stability (exchange rate appreciation, overheating, etc.) are often intertwined. - These cases are further described in IMF (2011f), Recent Experiences in Managing Capital Inflows—Cross-Cutting Themes and Possible Policy Framework. ### Brazil - Policy context: Managing large capital inflows has been a main policy issue since the global financial crisis. - Measure introduced in January 2011: - A 60 percent unremunerated reserve requirement on banks’ short foreign exchange (FX) positions in the spot market exceeding $3 billion or Tier 1 capital (whichever is lower). - Motivations and effects: - Motivated by concerns that banks or the local currency market could face disruptions following a large shock to the exchange rate, given banks’ large short FX spot positions. - Complemented Brazil’s IOF (Imposto sobre Operações Financeiras) tax on bond and equity inflows by reducing attractiveness of non-residents’ long local currency positions. - Targeted forward positions in the onshore and offshore markets (a form of carry trade) that were facilitated by local banks taking the other side of non-resident investors’ positions and hedging by borrowing FX. - By raising the cost of short FX positions, expected to affect an important channel for carry trades and reduce potential vulnerabilities in the banking sector. ### Korea - Policy context: After the global financial crisis, Korea experienced a pronounced sudden stop of short-term external bank debt; such debt had grown rapidly prior to the crisis partly due to demand for currency forward contracts by the corporate sector on expectations of won appreciation. - Measures: - June 2010: Introduced ceilings on banks’ foreign derivatives positions to reduce short-term external debt resulting from banks’ provision of forward contracts to corporates. - Ceilings expressed as a ratio to bank capital: 50 percent for resident banks and 250 percent for foreign bank branches. - Late 2010: Authorities announced a macroprudential stability levy on banks’ non-deposit foreign currency liabilities, with increasingly penal rates on shorter maturities. - This measure became effective on August 1, 2011, and is characterized as a CFM since it is designed to affect capital inflows. ### Turkey - Policy context: Facing rapidly rising capital inflows. - Measures implemented from the fourth quarter of 2010 as part of a new policy mix to preserve macroeconomic and financial stability: - Unremunerated required reserve ratios on all Turkish lira and FX liabilities of banks were raised in several steps to an average of 14 percent and 11.5 percent respectively (from their 5 percent and 9 percent troughs during the global crisis). - Higher rates were applied to shorter-duration bank liabilities. - The Central Bank of Turkey’s interest rate corridor was widened significantly to facilitate increased volatility of short-term market interest rates. - Intended effects: - Use of reserve requirements served both macroprudential and capital flow management purposes by aiming to moderate inflows and lengthen their duration. ### Distinction and Judgment - Not all macroprudential measures are CFMs (and vice versa). - Whether a macroprudential measure is a CFM depends on whether it is designed to influence capital inflows—a matter of careful judgment based on the totality of circumstances, including timing during an inflow surge. - Examples of macroprudential measures that would not be considered CFMs (if not introduced or intensified during an inflow surge): - Capital adequacy requirements - Loan-to-value ratios - Limits on net open FX positions - Limits on foreign currency mortgages *Source: APPENDIX I. MACROPRUDENTIAL OR CAPITAL FLOW MEASURES?* ### APPENDIX II. SELECTED CASE STUDIES ### APPENDIX II. SELECTED CASE STUDIES ### Selected European Countries (Bulgaria, Croatia, Poland, Romania, and Serbia) — Background - Mid-2000s: buoyant macroeconomic conditions; EU accession by Poland in 2004, Bulgaria and Romania in 2007. - Strong GDP growth between 2003 and 2008; large current account deficits (except Poland) financed by larger net capital inflows. - Credit boom: credit/GDP increased by 19 percentage points in Croatia and 45 percentage points in Bulgaria. - Only Bulgaria accumulated fiscal surpluses during this period; other countries maintained fiscal deficits. - Financial systems dominated by foreign banks. - Key macroeconomic indicators, average 2003–08 (in percent): - Bulgaria: GDP growth 6.3; CAB/GDP -15.7; Fiscal deficit/GDP 2.2; Public debt/GDP 28.7; External debt/GDP 79.0; Net capital flows/GDP 24.3; FX regime CB; FX liab./Total liabilities, 2007 58.6 - Romania: GDP growth 6.6; CAB/GDP -9.7; Fiscal deficit/GDP -2.6; Public debt/GDP 20.5; External debt/GDP 42.3; Net capital flows/GDP 13.5; FX regime Floating; FX liab./Total liabilities, 2007 42.5 - Croatia: GDP growth 4.3; CAB/GDP -6.6; Fiscal deficit/GDP -2.9; Public debt/GDP 35.1; External debt/GDP 73.9; Net capital flows/GDP 13.1; FX regime Stabilized; FX liab./Total liabilities, 2007 73.6 - Serbia: GDP growth 5.7; CAB/GDP -12.5; Fiscal deficit/GDP -1.0; Public debt/GDP 51.7; External debt/GDP 64.3; Net capital flows/GDP 19.0; FX regime Floating; FX liab./Total liabilities, 2007 67.8 - Poland: GDP growth 5.2; CAB/GDP -3.3; Fiscal deficit/GDP -4.1; Public debt/GDP 46.6; External debt/GDP 49.2; Net capital flows/GDP 5.6; FX regime Floating; FX liab./Total liabilities, 2007 21.9 ### Selected European Countries — Action and Instruments - Primary systemic risk: currency-induced credit risk from rapid FX-denominated lending, rising asset prices, and private indebtedness. - Measures used as a package across countries: - Poland: 2006 measures to contain FX lending risks (mortgages); 2008 higher risk weights for FX residential loans; 2010 tighter LTV (e.g., based on loan maturity) and debt service to income ratios for FX mortgage and retail lending. - Croatia, Romania, Serbia: curb FX lending via lending criteria, provisioning, gross exposure limits (Romania), and higher risk weights (Serbia 125 percent; Croatia 150 percent on lending to unhedged borrowers). Serbia introduced an exposure limit for retail lending relative to Tier I capital. - Bulgaria: targeted measures on overall credit growth and asset price growth — credit ceilings, differential risk weights based on LTV, countercyclical provisioning requirements. - All countries: LTV ratios imposed; all but Croatia restricted profit distribution; several implemented debt service to income limits. - Reserve requirements (RR) extensively used except Poland (unified low RR). RRs differentiated by currency, maturity, and source of funding. - Bulgaria and Croatia: marginal RR (MRR) on credit growth exceeding thresholds and additional external borrowing; Bulgaria set MRR very high (200 percent in 2005). - Countercyclical adjustments: RRs tightened pre-crisis and later lowered/removed (Bulgaria and Croatia); FX liquidity requirements relaxed in Croatia and Serbia; some provisioning and capital rules also relaxed later. ### Selected European Countries — Implementation Challenges & Outcomes - Instrument use characterized as discretionary (frequent adjustments) due to circumvention and limited initial effectiveness. - Banks circumvented by shifting maturities, funding through non-bank subsidiaries, FX-indexed loan classifications, and branches under different regulatory perimeters (Poland, Croatia, Bulgaria). - Authorities expanded regulation perimeter and harmonized prudential rules to close arbitrage channels. - Cooperation with other policies: - Monetary policy generally aligned with macroprudential policy. - Fiscal policy insufficiently tight in most countries except Bulgaria (Article IV noted augmented structural balance declined from 1.5 to 0.2 percent of GDP during 2005–2008). - Microprudential consistency improved over time. - Outcomes: - Measures effective in slowing credit growth and building capital and liquidity buffers, helping banking systems withstand the financial crisis fairly well except Romania. - Instruments appear to have altered composition of external debt in some countries as banks’ FX liabilities stopped growing in Croatia and Serbia. - Circumvention persisted via direct cross-border corporate borrowing; indebtedness continued to build. - Simpler RR may have been sufficient (Bulgaria) versus complex high-rate RRs that were hard to administer (Croatia, Serbia, Romania). - Macroprudential approach aligned with macroeconomic policy mix (e.g., Bulgaria) worked better; loose fiscal policy elsewhere shifted adjustment burden to monetary and macroprudential policy. ### New Zealand — Background - Banks reliant on short-term offshore funding due to low national saving; gross external debt exceeded 130 percent of GDP in 2009. - Short-term external debt remained high at almost 60 percent of GDP at end-2009. - Before liquidity rules, non-resident funding share grew to 40 percent of total funding; core funding as share of bank loans lower than many advanced countries. - At onset of crisis 2007, about 60 percent of non-resident funding had residual maturities up to three months. - The four largest banks in New Zealand are subsidiaries of Australian banks. ### New Zealand — Action and Instruments - October 2009: Reserve Bank of New Zealand (RBNZ) introduced quantitative liquidity/funding requirements, effective April 2010 after consultation. - RBNZ solely responsible for monitoring and enforcing the liquidity rules. - Instruments conceptually aligned with Basel III liquidity standards: - Liquidity mismatch ratios: minimum ‘zero’ requirements for one-week and one-month mismatch ratios each business day (mismatch dollar amount to total funding). - Minimum core funding ratio (CFR): 65 percent of total loans and advances from April 2010; increasing to 70 percent from July 2011; and 75 percent from July 2012. - Regulations include differentiated run-off rates, maximum exposures to individual liquidity providers, and definitions of eligible liquid securities. - Instruments treated as rules-based and stand-alone; monetary and fiscal policies were not aligned at introduction but not seen as adversely affecting the instruments. ### New Zealand — Outcome - Consultation and publication prompted banks to rebalance wholesale funding toward long-term funding before formal implementation. - New Zealand’s short-term debt fell from 64 percent of GDP in December 2008 to 50 percent in December 2010 — a 20 percentage point drop in short-term external debt ratio. - Banks competed for retail deposits; estimated impact corresponded to a hike in the policy rate of 100-150 basis points, increasing lending rates. - Post-implementation, all locally incorporated banks met liquidity and funding standards; system-level ratios exceeded required minima by 7-10 percentage points. - Careful calibration aided smooth transition; impact on average funding costs was higher than anticipated. ### Spain — Background - Dynamic provisioning (DP) introduced in 2000 to require banks to build reserves for eventual loan losses. - Prior practice: 1 percent generic provision plus specific provisions matching realized loan losses; DP added a periodic reserve fund based on average loan losses over a full economic cycle and specific provisions. - Post-euro entry: rapid credit growth and real estate lending; home prices rose >10 percent per year initially and reached 20 percent annually by 2004–2005. ### Spain — Action and Instrument Use - DP aimed to cope with increased credit risk and to incentivize prudent loan origination; used as a stand-alone macroprudential measure. - DP applied across loan categories: mortgages (differentiated by high/low LTV), corporates, automobiles, credit cards. - Calibrated expected loss estimates ranged from 0.6 percent to 2.5 percent across categories; average specific provisions varied from 0.1 percent to 1.6 percent. - One-off adjustment: provisioning rates lowered in 2005 after coverage of bad loans rose above 300 percent; liberated provisions retained as “other reserves.” - DP was rules-based: contributions determined by the difference between average provision through the cycle and current specific provision. - Low coordination with monetary policy (ECB conditions were too loose for Spain); only in 2008 did Spain introduce stricter treatment for commercial and residential real estate exposures. ### Spain — Outcome - DP helped cover rising credit losses during the financial crisis; banks used DP buffers as specific provisions rose. - Total provisioning cost (percent of total loans) increased less than specific provisions because dynamic reserves were tapped. - Coverage was less than full because actual loan losses exceeded expected losses used in DP calibration; the 2005 relaxation likely reduced prudence of DP rates. - Distributional issue: by June 2009, significant share of banks had run down buffers while fewer retained larger cushions — DP formula did not sufficiently differentiate bank-level portfolio risk. - DP had limited success in curbing credit growth; after 2000 credit contracted briefly then grew up to 25 percent annually. ### China — Background and Actions - Massive 2008 stimulus and delayed exit fueled credit boom in 2009–2010, driven by lending to local government financing platforms (LGFPs), real estate developers, and residential mortgages. - Signs of overheating: housing prices rising at annual 15-20 percent by early 2010. - Measures since 2010 (packages, fine-tuned, discretionary; include fiscal, interest rate, administrative measures): - LTV caps lowered from 80 percent to 70 percent for primary homes and to 50 percent for second homes (April 2010); mortgages for third homes suspended (September 2010); LTV cap on second home mortgages lowered to 40 percent (January 2011). - Interest rates on mortgages for second homes raised to 1.1 times the officially administered benchmark lending rate (April 2010). - Capital buffers introduced: capital conservation buffer, countercyclical buffer, systemic capital buffer — raising minimum capital adequacy ratio to 11.5 percent from 8 percent for large banks (2010). - Provision coverage ratio raised from 100 percent to 150 percent; provisions required to cover the higher of 150 percent of NPLs or 2.5 percent of total loans (2010). - Taxes increased on resale of properties within five years (January 2010); exemptions on stamp duties and income taxes for home purchases/sales abolished except for family’s only home (September 2010). - In high-price cities: local governments limited number of houses each family could buy; non-local mortgage applicants required proof of local tax payments for at least a year (September 2010). - Official benchmark lending rate raised five times between October 2010 and July 2011 for a total of 125 basis points. - Reserve requirement raised nine times for a total of 450 basis points. ### China — Outcome - Measures effective in lowering credit growth and housing price inflation: - Bank lending growth slowed to 16.9 percent (yoy) in June 2011 from 31.7 percent in December 2009. - Home sales rose 6 percent (yoy) in the first four months of 2011 vs. 30 percent (yoy) in same period 2010; sales declined sharply in major cities. - Home prices leveling off; an anticipated house price correction had not materialized by the reporting period. ### Colombia — Background, Action, and Outcome - Late-1990s recession after unsustainable fiscal positions and external shocks; peso floated; high unemployment and current account deficit stressed financial sector; mortgage write-offs and NPLs increased. - Final months of 1999: authorities implemented three prudential measures (package, broad-based, not adjusted, no fiscal/monetary accompaniment): - LTV limit: loan amount limited to 70 percent of collateral value. - Debt-to-income limit: borrower’s monthly debt service payments limited to 30 percent of disposable income. - FX net open position limits: global net open position in foreign currency limited to 20 percent of capital; spot net open position limited to 50 percent of capital. - Outcome: subsequent reduction in non-performing loans; banks’ foreign liabilities declined slowly while foreign assets expanded; private sector credit decreased initially then recovered; NPLs remained subdued for some time. ### Korea — Background, Action, and Outcome - Pre-2008: large buildup in short-term external debt tied to FX forward transactions and maturity/currency mismatches; inability to roll over short-term liabilities during crisis. - Measures introduced post-crisis (packages, fine-tuned, differentiated between domestic banks and foreign branches; discretionary; include fiscal tools): - Required raise of long-term foreign currency borrowing from 80 percent to 90 percent of long-term lending, and hold at least 2 percent of foreign assets in liquid investments rated A or higher (November 2009). - Limit value of FX forward transactions to 125 percent of exporters’ future export revenues (November 2009); lowered to 100 percent (June 2010). - FX derivative position limits: 50 percent of capital for domestic banks and 250 percent for foreign bank branches (June 2010); subsequently lowered to 40 percent and 200 percent, respectively (June 2011). - Withholding tax reinstated on foreign purchases of domestic bonds (January 2011). - Macro-prudential levy planned on banks’ non-deposit foreign currency liabilities (August 2011). - Outcome: - Measures appear to have limited growth in banks’ external liabilities; banks’ short-term external borrowing remained some 30 percent below pre-crisis levels as of 2010. - Measures did not stem portfolio inflows into debt and equity markets; withholding tax impact limited by double-taxation agreements. ### United States — Background, Instrument, and Outcome - 1980s surge in bank failures and rising leverage led to regulatory revision. Under the Federal Deposit Insurance Corporation Improvement Act of 1991, a consolidated leverage ratio introduced as a supplement to risk-based capital ratios. - Leverage ratio defined as minimum ratio of Tier 1 capital to total average adjusted assets (quarterly average total assets less specified deductions). - Minimum leverage ratios: - 3 percent for banks rated “strong”. - 4 percent for all other banks. - Prompt corrective action requires at least 5 percent leverage to be “well capitalized”. - Advantages: simplicity, ease of application, quick adoption, monitoring, and acts as back-up to model-dependent risk-based ratios. - Disadvantages: as a balance-sheet measure, does not capture off-balance-sheet exposures; can be circumvented. - Implementation: introduced broadly by Federal banking agencies; not adjusted over time or accompanied by other policies initially. - 2004: SEC regulation change allowed investment banks to raise leverage from 15:1 (6.7 percent) to 40:1 (2.5 percent) for broker-dealers opting for consolidated oversight. - Outcome: - Leverage of U.S. investment banks rose significantly after 2004 while commercial banks’ leverage remained relatively low. - Leverage ratio helped constrain excessive leverage in commercial banking but effectiveness limited by off-balance-sheet and structural circumvention risks. - Recommendation: coverage should be comprehensive and ratio adjusted countercyclically to reflect rising systemic risk. *Source: _wp11238 - APPENDIX II. SELECTED CASE STUDIES* ### APPENDIX III. THE SIMPLE APPROACH ### APPENDIX III. THE SIMPLE APPROACH ### Figure III.1 — Change in Risk Variables after the Implementation of Instruments - The figure reports the simple average of changes in the risk variables of all countries in the sample across quarterly horizons t-2, t-1, t, t+1, t+2, t+3, t+4 after implementation of each instrument. - Risk-variable panels shown include (examples as labeled in the source): - Loan to Value: Average Credit Growth (YoY Change) — quarterly axis from -1.5% to 1.0% (example tick labels shown). - Loan to Value: Average Real Estate Prices Growth (YoY Change) — quarterly axis from -0.5% to 4.0% (example tick labels shown). - Debt to Income: Average Credit Growth (YoY Change) — quarterly axis from -3.0% to 1.0% (example tick labels shown). - Debt to Income: Average Real Estate Prices Growth (YoY Change) — quarterly axis from -0.5% to 6.0% (example tick labels shown). - Reserve Requirement: Average Credit Growth (YoY Change) — quarterly axis from -2.0% to 3.0% (example tick labels shown). - Dynamic Provisioning: Average Credit Growth (YoY Change), Average Real Estate Prices Growth (YoY Change), Average Leverage Growth (YoY Change) — example axes include ranges such as 0.0% to 6.0% and -12.0% to 8.0%. - Caps on FCL and Limits on Maturity Mismatch: reported on currency mismatch and liquidity growth metrics (example axes include -0.6% to 0.8% and -2.0% to 7.0%). - Source note on the figure: IMF Staff Estimates. ### Key points from Figure III.1 - The presented panels summarize average, country-level quarterly changes in targeted risk metrics around instrument implementation dates, using a symmetric horizon of at least two quarters before and up to four quarters after implementation. - Graphs are plotted in year-on-year (YoY) change terms for credit growth and real estate prices, and in percentage-point scales for mismatch ratios and liquidity metrics. --- ### APPENDIX IV. GMM METHODOLOGY FOR PANEL REGRESSION ### Objective and Research Questions - Two primary questions: - What is the effect of an instrument in countries where it has been introduced? - What would have been its effect in countries that have not actually used it? - Approach: introduce a dummy variable I equal to 1 for country-periods where a particular instrument is used, 0 otherwise; this dummy captures an average “treatment effect” across countries, with non-using countries/periods as counterfactuals. ### Specification - A fixed-effect dynamic panel specification is used. The model (as in equation (2) in the source) includes: - Lagged dependent variable terms to capture dynamics. - Matrix I: time-series of the value of a particular instrument or dummy(s) for instrument use. - Matrix X: macroeconomic control variables (e.g., GDP growth, interest rate). - Interaction matrix IX: interaction between macro controls and the instrument; coefficient ݁ଵ measures change in correlation between the risk variable and the control variable after instrument adoption. - Matrix Y: change in systemic risk after introduction of an instrument. - For four risk variables and 10 instruments, a total of 40 regressions are required to show interactions. ### Estimation challenges and chosen estimator - Ordinary least squares (OLS) may be biased due to endogeneity (countries adopting instruments may systematically differ). - The dynamic panel generates bias in OLS with fixed effects because the lagged dependent variable is correlated with unobserved individual effects. - The Generalized Method of Moments (GMM) system estimator is used to address endogeneity and dynamic-panel bias: lagged variables are used as instruments in levels and differences, ensuring orthogonality between lagged endogenous variables and residuals. - Robustness checks: restrictions on lag lengths for instruments and estimation with ordinary least squares with fixed effects were used; results are stated as consistent across estimators. ### Data - Sample coverage: the sample covers 49 countries for a period of ten years, from 2000 to 2010 (text statement). - Table notes and estimation panels report the sample as composed of 48 countries in the estimation tables. - Sources for variables: - Credit growth: claims on the private sector from banks and non-banking financial institutions (source: IFS). - Leverage: assets over equity (source: IMF FSIs). - Liquidity risk proxy: non-core funding measured as bank credit to deposits. - Capital flow reversal risk proxy: ratio of foreign assets to foreign liabilities for banks and non-bank financial institutions (source: IFS). - Controls: GDP growth and prime lending rate (from IFS). - All variables tested and found covariance stationary. Interaction terms and interactions with dummies (exchange rate regime, degree of financial development, use of other macroprudential policies) were tested without further significant results. - Additional controls (policy rate, fiscal imbalances) tested but not used for reasons stated (policy rate identical across euro-area countries; fiscal imbalances highly correlated with GDP and interest rates). ### Results — Summary of main findings - Diagnostic tests: regressions passed the Arellano-Bond test for autocorrelation; Sargan test weakened by large number of instruments and small N. Robustness checks with restricted lag instruments and OLS fixed effects produced consistent coefficients of similar magnitudes. - Most significant coefficients are on the interaction between GDP growth and five instruments: - caps on LTV, - caps on DTI, - ceilings on credit growth, - reserve requirements, - dynamic provisioning. - Importance of cycle dependence: when instruments are analyzed during economic expansions alone (booms), dummy coefficients tend to become negative and can be significant, supporting a focus on procyclicality rather than only level effects. - Table-level highlights (coefficients and significance reported exactly as in source tables): - Table IV.1 (Effectiveness of Macroprudential Instruments in Reducing Credit and Leverage Growth): - Caps on Loan-to-Value (dummy): 0.0823 (2.65)*** for the credit-growth-equation constant dummy. - Reserve Requirements (dummy): -0.0013 (-0.06) in one specification and -0.0076 (-2.23)** in alternative specifications. - Dynamic Provisioning (dummy): -0.0062 (-0.20) in one specification and -0.1466 (-22.57)*** in another specification. - Restrictions on Profit Distribution (dummy): -0.0260 (-3.96)*** in the credit growth constant-dummy column. - For leverage growth regressions, examples include Reserve Requirements dummy -0.0076 (-1.43)*** and Limits on Forex Lending dummy -0.0340 (-7.03)*** (Table IV.1). - Table IV.2 (Effectiveness during Booms): - Dynamic Provisioning (dummy): -0.1466 (-22.57)*** reported in the credit-growth column (boom-only specification). - Reserve Requirements (dummy): -0.0076 (-2.23)** in a boom-only credit-growth regression. - Several instruments not reported for booms due to not enough observations. - Table IV.3 (Effectiveness in Reducing Credit Growth — level and procyclicality; interaction coefficients reported): - Caps on Loan-to-Value (dummy coefficient): 0.0755 (1.92)** on the credit-growth level. - Caps on Loan-to-Value 2 × GDP Growth (interaction): -0.0615 (-2.59)**. - Caps on Debt-to-Income 2 × GDP Growth (interaction): -0.0637 (-2.44)**. - Reserve Requirements 2 × GDP Growth (interaction): -0.0448 (-2.18)***. - Dynamic Provisioning 2 × GDP Growth (interaction): -0.1463 (-0.90) (interaction reported but t-statistic not significant at conventional levels). - Countercyclical Capital Requirements 2 × GDP Growth (interaction): -0.1563 (-2.57)***. - Limits on Forex Lending (level) 0.0822 (2.81) and (interaction) 0.0295 (0.53) — level positive and significant on credit growth; interaction not significant. - Table IV.4 (Effectiveness in Reducing Leverage Growth — level and procyclicality; interaction coefficients reported): - Caps on Debt-to-Income 2 × GDP Growth (interaction): -0.0526 (-2.69)***. - Reserve Requirements 2 × GDP Growth (interaction): -0.0937 (-4.39)***. - Dynamic Provisioning 2 × GDP Growth (interaction): -0.2765 (-3.75)***. - Limits on Forex Lending (level) -0.0334 (-5.48)*** (negative and significant on leverage). - Restrictions on Profit Distribution (level) -0.0127 (-2.39)** and interaction 0.0823 (2.83)*** (level reduces leverage; interaction with GDP growth positive and significant). - Interpretation emphasized in the source: - Many instruments show limited average level effects when pooled across all cycle phases, but reveal significant negative effects on procyclicality (i.e., reduce the sensitivity of risk variables to GDP growth) when interaction with GDP growth is modeled. - Some instruments (notably dynamic provisioning, reserve requirements, caps on LTV and DTI, ceilings on credit growth) are associated with reductions in procyclicality or in growth of risk variables during expansions. ### Estimation notes and sample particulars (as reported) - Estimation period: 2000–2010. - Dependent variables: log change in the real level of credit (credit growth) or leverage (leverage growth). - Credit measured as claims on private sector from both bank and non-bank financial institutions (source: IFS). - Leverage measured as assets over capital (source: IMF FSIs). - Interest rate: nominal long-term interest rate on prime lending (IFS). - Regression controls: dummy variables for exchange rate regime flexibility, individual (country) fixed effects, a time trend (year effect), and a dummy for the use of other macroprudential policy instruments. - Estimation approach: Instrumental variables for policy instruments and GMM Arellano-Bond estimator to address selection bias and endogeneity. - Significance notation: ***, **, * indicate statistical significance at 1%, 5%, and 10% (two-tail) test levels, respectively. - Footnote reporting limitations: 2/ Not enough observations during booms for some instrument-country combinations; instances where instrument coefficients are "N/A" due to insufficient observations. *Source: IMF Staff Estimates (content unit: APPENDIX III. THE SIMPLE APPROACH; APPENDIX IV. GMM METHODOLOGY FOR PANEL REGRESSION from the supplied PDF content).* ### APPENDIX V. MANAGING RISK WITH SELECT MACROPRUDENTIAL INSTRUMENTS ### APPENDIX V. MANAGING RISK WITH SELECT MACROPRUDENTIAL INSTRUMENTS ### Loan-to-Value Ratios - Purpose and usage - Limits on Loan-to-Value (LTV) ratios are applied to reduce systemic risk from boom-bust episodes, notably in real estate markets. - Popular in Asian and European countries; introduced in two stages: (i) during the mid-2000s boom (especially in Asia) and (ii) in the wake of the global financial crisis (especially in Europe). - By limiting the loan amount relative to property value, LTV limits aim to rein in house price increases by reducing household leverage (the financial accelerator effect). - Tool can also be designed to meet social objectives (e.g., ensure lower-income households access financing). - Early-cycle implementation is important to ensure a preemptive effect and provide a minimum buffer. - Design features and cross-country practice - Single vs. multiple instruments - Most countries combine LTV with other macroprudential tools (DTI/debt service-to-income limits or reserve requirements). - Countries using LTV as a single instrument adjust maximum rates regularly (Canada, Hong Kong, Singapore, Thailand). - Some recent adopters have kept maximum rates unchanged (Norway, Sweden). - Targeting - About half of countries differentiate LTV by mortgage purpose or property value to limit financing for commercial investors or luxury/speculative investments (examples: Canada, Turkey, Singapore; Hong Kong, Malaysia, Singapore). - Hong Kong ties maximum LTV negatively to property value; Korea ties rates to whether property is in a speculative zone. - Carve-outs or mortgage insurance are used to ensure access for social housing or first-time buyers (Canada, Chile, Hong Kong, United States). - Some European countries differentiate LTV by loan currency (Poland, Hungary, Serbia). - Fixed vs. time-varying - Some countries keep LTV rates constant to provide a minimum buffer (Colombia, Lebanon, Malaysia, Sweden). - Other countries adjust LTV countercyclically—tightening in booms and relaxing in downturns (China, Hong Kong, Korea). - In some cases adjustments are reactive rather than countercyclical. - Rules vs. discretion - All countries use discretion when adjusting LTV limits; several tighten limits in measured ways in response to house price developments (Canada, Hong Kong, Korea, Singapore). - Effectiveness and limitations - Statistical evidence shows LTV limits have a clear effect on credit growth and property prices. - Effect may wear off in dynamic markets, requiring successive tightenings (e.g., Hong Kong and Singapore). - Targeted LTVs (e.g., Korea) have inconclusive statistical evidence on effectiveness. - Pros, cons, and side effects - Pros: - Allows targeting of specific risks in housing markets. - Likely to have an immediate effect. - Cons / requirements: - Requires debtor-level data on property values. - May require recalibration if market dynamics render limits non-binding. ### Dynamic Loan-Loss Provisioning - Overview and country practice - Spain introduced dynamic loan-loss provisioning (DP) in 2000; rules-based DP also applied in Latin America. - Four main DP systems: 1. Continuous provisions against a benchmark average provision flow through the credit cycle (Spain, Uruguay). 2. Activation mechanism that accumulates provisions during upswing and allows drawdown during downturn (Colombia, Peru). 3. Provisioning rates set according to debtors’ classification or risk profile in expected loss terms (Chile, Mexico); Chile allows additional countercyclical provisions to cover “unexpected losses”. 4. Countercyclical provisioning with discretionary rate adjustments (Bulgaria, Croatia, India, Mongolia, Russia). - Design and calibration - Aim: distribute loan losses more evenly over the credit cycle by accounting for expected loss rather than incurred loss. - Requires building reserve buffers during upswings to counter low specific loan reserves when credit quality is high. - Best introduced at the beginning of the credit cycle to build adequate reserve cushion; recalibration may be required as circumstances change. - Single vs. multiple and targeting - DP introduced with other macroprudential measures in Colombia, Mexico, Peru, Uruguay; Spain used it stand-alone. - All DP systems apply differentiated fixed rates by loan category (commercial vs. consumer). Chile and Mexico calibrate rates to debtors’ risk profiles. - Fixed vs. time-varying and rules vs. discretion - Except Spain in 2004, provisioning rates in DP systems have not been changed. - Strict-form DP systems are rules-based; some countries use discretionary countercyclical adjustments (e.g., India adjusting rates gradually and by loan type), which are less likely to build adequate buffers. - Effectiveness and limitations - Statistical evidence: rules-based DP systems are effective; discretion-based systems less so. - DP not designed to cover large unexpected loan losses (bank capital covers those) nor to rein in rapid credit growth. - Examples: - Spain: dynamic provisions offset about half of loan losses during 2008–09; eventual loan losses exceeded expected losses. - Uruguay: reserves coverage ballooned because expected delinquencies used for calibration did not materialize. - In countries with short histories of DP, conclusive results are not available. - Pros, cons, and side effects - Pros: - Countercyclical reserves buffer covers rising loan losses and helps sustain credit in downturns. - Smoothes provisioning costs over the credit cycle. - Cons / requirements: - Requires data on provisioning flows or expected loss. - May lead to overprovisioning if calibration is incorrect and incurred loss is much lower than expected loss. ### Reserve Requirements - Purpose and usage - Emerging market countries use reserve requirements (RR) as a macroprudential tool to (i) protect against liquidity risks and (ii) address risks from excess credit growth, sometimes fueled by capital inflows. - As a macroprudential tool RR are usually targeted and can be raised to very high rates; as a monetary tool RR are often reduced to prudential minimums and replaced by indirect instruments. - Design and calibration - Objective should be clearly defined (slow credit growth vs. encourage stable funding). - If slowing credit growth is the goal and banks dominate the financial system, RR should be widely applied with relatively simple design and rates set below prohibitive levels to minimize distortions. - If objective is limited (e.g., encourage stable funding), RR can be targeted—e.g., on short-term foreign borrowing by banks (Peru). - Targeting, timing, and discretion - RR for macroprudential purposes are usually more targeted than traditional single-rate approaches (different rates by maturity, currency, and base). - Applied countercyclically: raised during booms and lowered or lifted during downturns; some marginal RR can be imposed at extraordinarily high levels during booms. - Adjustments are made discretionarily using trial-and-error; both rate and base can be adjusted. - Effectiveness and limitations - Statistical evidence suggests RR reduce procyclicality of credit growth; durability beyond certain horizons is not robustly confirmed. - Country evidence suggests RR need periodic recalibration to preserve effectiveness. - Pros, cons, and side effects - Pros: - Builds useful liquidity buffer. - Immediately effective. - Can work well with other macroprudential tools. - Easy to apply and adjust. - Cons: - Easy to circumvent; effectiveness may wear off. - Possible migration of risk. - Burdensome to enforce if too complex. - May restrict credit to small/medium enterprises. ### Measures Targeted at Foreign Currency (FX) Lending - Rationale and instrument types - Concern when banks engage heavily in FX lending and borrowers do not earn foreign exchange: exchange rate depreciation can create currency-induced credit risk and systemic effects. - Macroprudential measures fall into two groups: 1. Limit exposures: direct caps, debt-to-income caps by currency, targeted restrictions, outright bans. 2. Build buffers: LTV set by loan currency, higher risk weights or capital requirements, higher provisions against FX lending. - Design and calibration - Measures aim to restrain growth in FX lending to unhedged borrowers and build buffers against downturns. - Degree of targeting depends on data availability and understanding of depreciation impact on borrower types. - Ideally adopted before significant exposures build up; many countries adopted measures during the crisis but still have large outstanding FX loan stocks (e.g., Hungary, Ukraine). - Single vs. multiple instruments - Countries often combine FX lending measures with RR to address liquidity and exposure objectives. - Some countries use single measures (e.g., outright FX lending bans: temporary in Austria 2008–2010; permanent in Brazil since the mid 2000s). - Others use multiple measures to build buffers: higher risk weights on FX loans (up to 150 percent in Croatia), limits on FX exposures relative to capital or capital add-ons (Peru, Romania), plus higher provisioning (Peru, Romania, Uruguay). - Targeting, timing, and rules - Targeting varies: measures applying to all FX loans (Lebanon, Peru), those targeting unhedged borrowers (Argentina, Croatia, Serbia, Uruguay), or FX mortgage lending specifically (Poland). - Some countries impose outright bans on certain FX or FX-indexed lending (Argentina, Hungary, Turkey, Ukraine). - Measures typically use fixed parameters (e.g., higher risk weights). Most countries tighten limits during upswings and usually do not relax them during downturns. - Application is often formula-based; some adjustments have been made discretionarily after observing effects on bank behavior. - Effectiveness and limitations - Mixed statistical evidence that FX lending measures restrain credit growth when tested in isolation; effectiveness weakens when other measures are included. - Net open position limits on FX exposures are effective in reducing external indebtedness of the financial system. - Risks: migration of risk to nonbanks or FX-indexed loans; potential disintermediation or distortions; may encourage more local-currency lending. - Pros, cons, and side effects - Pros: - Formula-based measures provide greater predictability. - Can help build buffers. - Can encourage more lending in local currency. - Can be implemented relatively easily. - Cons: - May cause migration of risk to nonbanks or other areas. - May contribute to disintermediation or introduce distortions. ### Selected numeric and tabular highlights from the appendix - Reserve requirements: reported range of rates (in percent) across selected countries - Argentina: 0-100 - Bulgaria: 4-400 - China: 6-21.5 - Colombia: 0-140 - Croatia: 13-55 - Indonesia: 1-8 - Lebanon: 15-25 - Romania: 0-40 - Serbia: 5-100 - Peru: 0-120 - Turkey: 5-16 - Uruguay: 25-35 - Examples of measures to address FX lending (selected entries) - Direct caps on exposures: example entry shows "60 percent of FX deposits (to 2006)". - Additional capital on FX exposures: example entry "2.5 percent of total FX exposure". - Higher risk weights on FX loans (max): - 150 percent for unhedged borrowers (Croatia). - 75 percent for fully secured household mortgages (entry in table). - 125 percent for unhedged borrowers (entries in table). - Limits on FX exposures to capital / capital add-ons: - Example: "Maximum allowed FX exposure = 20 percent of regulatory capital". - Example: "FX loans to unhedged borrowers to own funds, 300 percent. Lifted in 2007." - Other items: - Additional market risk capital charge depends on value at risk from FX volatility. - FX credit to total private sector credit (percent) — entries listed in table V.2 - Argentina: 73 - Austria: 52 - Croatia: 52 - Hungary: 30 - Lebanon: 63 - Peru: 68 - Poland: 52 - Romania: 52 *Source: APPENDIX V. MANAGING RISK WITH SELECT MACROPRUDENTIAL INSTRUMENTS (excerpt).* ### APPENDIX VI. THE CONCEPTUAL BASIS FOR MACROPRUDENTIAL INSTRUMENTS ### APPENDIX VI. THE CONCEPTUAL BASIS FOR MACROPRUDENTIAL INSTRUMENTS ### Conceptual basis of instruments - Caps on the LTV - Imposes a down payment constraint on households’ capacity to borrow. - Limits the procyclicality of collateralized lending since housing prices and collateral-based borrowing interact procyclically. - Set at an appropriate level, the LTV addresses systemic risk whether or not it is frequently adjusted; adjustment makes it a more potent counter-cyclical instrument. - Caps on the DTI - Prudential regulation aimed at ensuring banks’ asset quality when used alone. - Together with LTV, further dampens cyclicality by adding another constraint on households’ borrowing capacity. - Adjustments can be made in a counter-cyclical manner to address the time dimension of systemic risk. - Caps on foreign currency lending - Un-hedged borrowers are exposed to foreign exchange risks, which create lender credit risks; large common exposure can become systemic. - Caps (or higher risk weights, deposit requirements, etc.) may be used to address foreign-exchange-induced systemic risk. - Ceilings on credit or credit growth - Ceilings may be imposed on total bank lending or sectoral credit. - Aggregate ceilings dampen the credit/asset price cycle (time dimension); sectoral ceilings (e.g., real estate) contain asset price inflation or limit common exposure (cross-sectional dimension). - Limits on net open currency positions / currency mismatch - Limit banks’ common exposure to foreign currency risks. - Address externalities from convergent FX purchases/sales by banks that amplify exchange rate fluctuations and increase credit risk for un-hedged borrowers. - Limits on maturity mismatch - Address externality of fire sales of assets: inability to meet short-term obligations can force asset liquidation, imposing fire sale costs on the system and causing systemic liquidity crises via contagion. - Reserve requirements - Affect credit growth and can dampen the credit/asset price cycle (time dimension). - Required reserves provide a liquidity cushion to alleviate systemic liquidity crunches when needed. - Countercyclical capital requirement - Can take the form of a ratio or risk weights raised during upturns to restrain credit expansion and reduced during downturns to provide a cushion. - A permanent capital buffer built in upturns and deleted in downturns serves the same purpose. - Both address cyclicality in risk weights under Basel II based on external ratings that are procyclical. - Time-varying / Dynamic provisioning - Traditional dynamic provisioning is calibrated on historical bank-specific losses, but can be used to dampen financial system cyclicality. - Provisioning can be raised during upturns to build a buffer and limit credit expansion, and lowered during downturns to support lending. - May be adjusted by fixed formula or policymaker discretion to affect bank lending behavior counter-cyclically. - Restrictions on profit distribution - Intended to ensure banks’ capital adequacy; undistributed profits add to bank capital. - Tend to have a counter-cyclical effect if used in a downturn. - The capital conservation buffer of Basel III has a similar role. ### Country experience: selected cases and measures - Argentina (2010) - Background: rapid growth in 2010, improved international liquidity, low advanced economy rates, historically volatile banking profitability correlated with cycles. - Actions: - Restrictions on profit distribution: introduction of a restriction on profit distribution (May 2010). - Note: early 2000s measures included a 30% unremunerated reserve requirement on capital inflows, a limit on net open currency position, and a foreign currency lending capacity requirement. - Austria (2003-2010) - Background: high ratio of foreign currency loans exposing banks and borrowers to exchange rate risk. - Actions (timeline): - 2003 – introducing minimum standards governing foreign currency loans and loans with repayment vehicles. - 2006 – increasing risk awareness of foreign currency borrowers. - 2008 – suspending the granting of foreign currency loans. - 2010 – improving the risk-bearing capacities of individual banks. - Brazil - 2005-2007 - Background: signs of overheating and increasing capital inflows. - Actions: - Time-varying/dynamic provisioning: introduction of forward looking provisioning. - Currency mismatch: decrease in limits on currency mismatch from 60% to 30% of regulatory capital (2007). - 2008-2010 - Background: rapid growth, strong credit expansion, increase in speculative capital inflows. - Actions: - Reserve requirements: raised to reduce credit growth (2010). - Introduction of a 60 percent unremunerated reserve requirement on banks’ short foreign exchange positions in the spot market exceeding US$3 billion or Tier I capital (whichever is lower). - Capital: increased capital requirements for some consumer loan operations with long maturities and high LTV ratios (including car loans) (2010). - Bulgaria - 2004-2007 - Background: household credit grew rapidly during EU accession; 49 percent growth in bank lending in 2004 caused concern. - Actions: - Capital: more stringent rules for classifying claims and determining banks’ capital adequacy by excluding current profit from the capital base (2004 and 2005). - Provisions: higher specific provisions for loans to households introduced (2005) and tightened several times since then. - LTV: introduction of a 70 percent LTV ratio for mortgages risk-weighted at 50 percent. - Reserve requirements: tighter reserves by reducing share of vault cash in eligible assets and broadening the liability base; marginal reserve requirement for banks exceeding average credit growth (2006); rise in reserve requirement ratio (2007). - 2008-2010 - Background: financial crisis led to stagnant growth; authorities promoted credit growth. - Actions: - Reserve requirements: reductions in reserve requirements (2008 and 2009). - Risk weight: reduction in risk weights for loans to households and mortgage loans (2010). - Note: differentiated reserve requirements for funds attracted from abroad vs domestically introduced (2009); lower reserve requirement ratio on funds from abroad. - Canada (2008-2011) - Background: rapid mortgage growth in 2008 and high household debt in 2010 and 2011. - Actions: - LTV: July 2008 – maximum term for mortgages decreased from 40 to 35 years. - February 2010 – selectively tightened LTV ceilings on cash-out refinancing transactions and investment property loans. - April 2011 – maximum amortization period for new government-backed insured mortgages with LTV ratios of more than 80% reduced to 30 years from 35 years. - Chile (2008-2009) - Background: economic decline following global financial crisis; authorities restored credit flow to low-income households and SMEs. - Actions: - LTV: maximum LTV ratio for covered bond-type mortgages raised from 75% to 100% for debtors with higher credit ratings (2009). - Differentiated reserve requirements: introduction for foreign currency (2008). - Note: Chile has a systemic capital surcharge when a merger/acquisition yields a bank market share higher than 15 percent, requiring higher capital adequacy ratio from 10 to 14 percent for a minimum period of not less than a year. - China (2010-2011) - Background: massive 2008 stimulus with delayed exit fueled a domestic credit boom; housing prices rose at an average annual rate of 15-20% from late-2009. - Actions: - LTV: primary homes lowered from 80% to 70% and to 50% on second homes (2010); LTV on second homes lowered further to 40% (2011). - Lending ceiling: caps on credit growth for major banks; verbal guidance to banks to temporarily stop lending. - Reserve requirements: increased 8 times since Jan 2010. - Countercyclical capital requirement: large banks required to have a countercyclical and systemic capital buffer (2010). - Provisions: provision coverage ratio raised from 100% to 150%. - Colombia - Late 1990s - Background: excessive mortgage borrower leverage leading to a mortgage crisis. - Actions: - LTV: introduction of caps on LTV ratios at 70% (1999). - DTI: introduction of caps on debt-to-income ratio by imposing a monthly debt service limit of no more than 30% of disposable income (1999). - NOP: limit on financial institutions’ net open foreign currency positions set at 20 percent of their capital (1999). - 2007-2009 - Background: growth over 7 percent in 2007 with signs of overheating, then slowdown due to global crisis. - Actions: - Maturity mismatch: limit on maturity mismatch introduced (2009). - Reserve requirements: marginal reserve requirements used (2007 and 2008). - Time-varying/dynamic provisioning: introduced (2007). - Restrictions on profit distribution: introduced (2008), used one time. - Croatia (2003-2008) - Background: strong growth, credit and mortgage booms (bank credit to private sector increased by 20-30% a year 2001-2003; mortgage credit grew at 31% annual rate 2003-2007); house prices surged by cumulative 30 percent. - Risks: foreign currency lending to unhedged borrowers, weaker underwriting, rollover/liquidity risks, contagion from foreign borrowing. - Actions: - LTV: introduction of LTV ratio for housing loans at 75% (2006). - DTI: approval of new loans prohibited if debtor’s average monthly income did not cover total repayment obligations (2006). - Lending ceiling: lower credit ceilings (2003 and 2007-2008); banks exceeding growth thresholds required to hold low-yielding central bank bills. - Liquidity: additional liquidity requirement increased ratio of foreign liquid assets to foreign borrowing to 24% (2003) and 35% (February 2005), cut to 32% (March 2005). - Reserve requirements: unremunerated reserve requirement on additional foreign borrowings (2004-2008) and on newly issued securities (2006-2008). - Time-varying/dynamic provisioning: extra provisions for excessive credit growth (2004-2006). - Countercyclical capital requirement: used (2008). - Risk weight: 25% increase in risk weight on loans to debtors with currency mismatch (2005). - France (2010) - Background: global financial crisis caused deterioration in interbank market functioning and bank liquidity problems. - Actions: - Liquidity: imposition of one-month liquidity ratio of 100% (2010). - Exposure limits: banks' exposures to individual clients or a group of connected clients limited to no more than 25% of the bank’s capital (2010). - Greece (1999-2005) - Background: household credit growth accelerated (30 percent in 1998); household debt to disposable income rose from 27% in 2000 to 52% in 2004; mortgage debt to GDP rose from 4 percent in 1995 to 23% in 2004; real house price appreciation reached 67% over the same period. - Actions: - Lending ceiling: introduction of unremunerated reserves equivalent to growth of credit above specified rates (1999-2000). - DTI: imposition of an indicative limit of 40% on the household debt service-to-income ratio (2005). - Provisions: increases in regulatory provisioning ratios for doubtful consumer loans from 84% to 100% (2005). - Hong Kong - 1990s - Background: limited land supply, large public housing sector, volatile real estate markets; property inflation in late 1980s and acceleration in 1993. - Actions: - LTV: reduction from 80-90% to 70% (1991); further reduction to 60% for luxury residences (1997). - Lending ceiling: ceiling on growth of mortgage lending set at 15% per annum (1994); banks’ exposure to property limited to 40% (1994-1998). - 2009-2010 - Background: post-1995–2003 boom-bust, house prices increased 45% since 2007Q2 as of 2010Q3. - Actions (selected): - LTV: 2009 – reduction to 60% for properties valued at or above HK$20 million from 70%. August 2010 – extension of 60% LTV limit to properties valued at or above HK$12 million and non-primary-residence loans. November 2010 – further reductions: (i) residential properties valued at or above HK$12 million from 60% to 50%, (ii) residential properties valued at or above HK$8 million and below HK$12 million from 70% to 60%, and (iii) all non-owner-occupied residential properties, properties held by a company and industrial and commercial properties to 50%, regardless of property value. - DTI: standardization of the limit on DTI at 50% from the previous range of 50-60% (2010). - Lending ceiling: (text truncated in source at this point). *Source: _wp11238 - APPENDIX VI. THE CONCEPTUAL BASIS FOR MACROPRUDENTIAL INSTRUMENTS* ### introduction of loan cap of HK$7.2 million on mortgages subject ### _wp11238 - introduction of loan cap of HK$7.2 million on mortgages subject ### Hong Kong — loan cap on mortgages - Introduction of loan cap of HK$7.2 million on mortgages subject to 70% LTV limit (August 2010). - Reduction of cap to HK$4.8 million (November 2010). ### Hungary — addressing foreign currency mortgage vulnerabilities (2010) - Background and motivation: - The economy had considerable vulnerabilities in the form of high external debt as well as currency mismatch. - A large share of mortgage loans was provided in foreign currency, making unhedged borrowers, especially the household sector, vulnerable to exchange rate volatility. - Main motivation: to address the excessive foreign exchange lending to households. - Actions — macroprudential tools used: - LTV: introduction of LTV limit for FX mortgages (2010). - DTI: introduction of DTI limit for FX mortgages (2010). - FX lending ceiling: ban on foreign exchange mortgage lending (2010). ### India — dampening procyclicality and credit boom (2004–2010) - Background and motivation: - Financial institutions tended to behave procyclically; pre-crisis growth and urbanization led to a real estate boom and rising credit to the private sector. - After the global financial crisis, credit began to decline. Main objective: reduce procyclicality. - Actions — macroprudential tools used: - LTV: introduction of 80% LTVs for residential real estate (2010). - Reserve requirements: increase in cash reserve requirements from 4.5% to 5% (2004), 5.5% (2006), and then to 6% (2007). - Risk weight: increase in risk weight on housing loans from 50% to 75% (2005) and for commercial real estate exposure from 100% to 125% (2005), 150% (2006), and then to 100% (2008). - Provisions: increase in general provisions from 0.25% to 0.4% (2005), 1% (2006), and then to 2% (2007). ### Indonesia — containing inflationary pressure and capital inflow vulnerabilities (2010–2011) - Background and motivation: - Rapid growth with rising inflationary pressure, partly due to massive capital inflows from advanced economies. - Objectives: contain inflationary pressure and reduce vulnerability from capital inflows. - Actions — macroprudential tools used: - Reserve requirements: reserve requirement for local currency deposits raised from 5% to 8% (2010). - Reserve requirements for foreign currency deposits: raised from 1% to 5% (2011) and then to 8% (later in 2011). - Additional reserve requirement introduced for banks with loan to deposit ratios below 78 percent or above 100 percent (March, 2011). ### Ireland — damping credit growth amid rapid mortgage expansion (2006) - Background and motivation: - Rapid mortgage growth between 2000 and 2006 driven by financial deregulation, positive macro outlook, immigration, tax cuts on non-owner-occupied property, and removal of residential property taxes. - Mortgage debt to GDP grew by 159 percent between 1996 and 2005; house prices rose by 217 percent. - Objectives: dampen credit growth and strengthen banks against rapid mortgage growth. - Actions — macroprudential tools used: - Risk weight: increase in risk weight for mortgages from 50% to 100% of the loan value, on the portion of each loan exceeding 80% of the value of the property (2006). ### Italy — reducing lending cyclicality (2007) - Background and motivation: - Bank lending accelerated due to strong corporate demand and recovery in activity; household lending grew fast; higher share of loans linked to real estate. - Motivation: reduce lending cyclicality. - Actions — macroprudential tools used: - LTV: introduction of caps on LTV (2007). Mortgages secured by residential real estate are discouraged when they are beyond 80% loan to value. Tighter capital requirements are requested for loans above 80% loan to value. ### Korea — managing housing cycles, household debt, and external vulnerabilities (2002–2011) - Background and motivation: - Post-Asian crisis expansive policies created credit booms; housing price volatility with 26 percent increase from 2001Q1 to 2003Q3, resumed appreciation with 14 percent increase between 2005Q1 and 2007Q1, then decline after the global financial crisis. - Policy aims: maintain positive but limited house price appreciation; maintain consumer confidence; support construction sector; provide for housing needs; more recently limit household debt. - Actions — macroprudential tools used (housing market and broader financial stability): - LTV: introduction of caps on LTV ratios in 2002; tightened 4 times and loosened once in accordance with property price fluctuations. - DTI: introduction of caps on debt-to-loan ratio in 2005; tightened 4 times and loosened 2 times in accordance with property price fluctuations. - Loan-to-deposit ratio: reduction in banks’ loan-to-deposit ratio to 100% starting in 2014 (November 2009, the deadline was shortened to end-June 2012, in June 2011). - Reserve requirements: increase from 5% to 7% for demand deposits, money market deposit accounts, and other non-savings deposits (2006); reduction in reserve requirement from 1% to 0% for long-term savings deposits (2006); overall reserve requirements increased from 3% to 3.8% (November 2006); reserve requirement on demand deposits in foreign currency increased from 5% to 7% (2006). - Other instruments: tax incentives, subsidized financing, government construction and purchases of unsold houses, direct support for the construction sector, and moral suasion on lenders. - Actions — addressing short-term external debt and foreign currency risks (2009–2011): - Motivations: reduce short-term external debt and capital flow volatility; reduce wholesale financing; strengthen foreign currency liquidity standards to reduce maturity mismatches and improve liquid asset quality; prevent excessive foreign currency bank loans from turning into systemic risks. - Off-balance-sheet limits: introduction of a ceiling on banks’ foreign exchange forward positions (2010) and tightened further in 2011. - Lending ceiling: limits set on foreign currency loans (2010). - Liquidity: use of stronger foreign currency liquidity standards (2009). - Tax: reintroduction of a withholding tax on foreign purchases of treasury and money stabilization bonds and of a macroprudential levy on banks’ non-deposit foreign currency liabilities (2011). - Restriction on investment in foreign currency denominated bonds. *Source: _wp11238 - introduction of loan cap of HK$7.2 million on mortgages subject* ### introduction of restriction on domestic banks and other institutional investors onshore from investing in ### introduction of restriction on domestic banks and other institutional investors onshore from investing in Kimchi bonds (foreign currency denominated bonds issued by Korean banks and corporate) that are intended to be converted into Korean won for domestic use (2011) ### Background and motivations (selected excerpts) - Lebanon (1997-2009): banks carried a substantial maturity mismatch from funding lending largely from short-term deposits and had significant foreign currency exposure from foreign exchange lending to unhedged clients. The central bank introduced measures to reduce open currency positions and resulting foreign exchange risk. - Lebanon (2008-2009): the global financial crisis and slowdown in the Gulf negatively affected capital flow and economic activity; the central bank introduced measures to promote credit growth. - Malaysia (1990s; 2005; 2010): house prices accelerated to 13 percent annual growth in 1995-96 after 3 percent per year in 1993-94; office rents rose 50 percent between 1990 and 1996; later mortgage credit growth and house price increases prompted measures to reduce mortgage growth and property prices. - Mexico (late 1990s, early 2000s; 2010): post-1994-1995 crisis reforms and concerns about liquidity risk in domestic and foreign currency prompted limits on maturity mismatch and interbank exposures; in 2010 authorities increased buffers and reduced procyclicality. - Mongolia (2010-2011): the 2008-2009 crisis highlighted banking vulnerabilities; rising mineral prices and mining investment returned growth in 2010; motivations included addressing procyclicality, exchange rate risk from cross-border exposure, leverage and maturity mismatch, and credit growth and asset price risk. - New Zealand (2010-2011): banks’ dependence on short-term offshore funding and the global financial crisis motivated measures to ensure sufficient liquid assets and stable funding sources. - Nigeria (2008-2010): large bank exposures to the stock market and oil industry, stock crash and oil price collapse raised liquidity and asset quality concerns; authorities acted to mitigate liquidity risk and reduce loan concentration. - Norway (1998; 2010): following rebound from a systemic banking crisis, house prices and credit growth led to measures to curb credit growth and property price inflation; in 2010 household debt and high housing debt motivated limits on LTV and DTI. - Peru (2001-2010): double-digit credit growth and massive capital inflows motivated actions to dampen lending cyclicality, mitigate FX risk in banks' balance sheets, and mitigate short-term capital flows and exchange rate volatility. - Poland (2006-2011): credit boom in 2006-2008 and growth of FX lending to unhedged borrowers raised systemic risk; motivations included mitigating credit and FX risk, and strengthening capital and liquidity buffers. - Portugal (1999): from 1996 to 2000 house prices rose modestly by 17 percent but mortgage debt to GDP doubled from 21% to 41%; motivations were moderating cycles in specific sectors and safeguarding banking soundness. - Romania (2000s): rapid credit growth including over 40% in 2003 and growth of FX loans to unhedged borrowers motivated slowing credit growth, limiting indebtedness, and limiting currency risk. - Russia (2008-2010): contraction by 7.9 percent in 2009 and capital outflows; banking sector suffered bad loans and stagnant credit growth; motivations were to stimulate credit growth, mitigate liquidity constraints, and limit currency risk and manage capital flows. - Serbia (2004-2011): rapid credit growth accompanied by eurorization and systemic risk from FX lending to unhedged borrowers; actions aimed at constraining credit growth and limiting currency risk. - Singapore (2009-2011): real house prices increased 45 percent from 2004Q2 to 2008Q1; private property price index declined almost 25 percent between 2008Q2 and 2009Q2 then rebounded sharply; motivations were to ensure a stable and sustainable property market and reduce speculative demand. ### Actions — Macroprudential tools used (selected country entries and exact measures) - Lebanon: - NOP: introduction of foreign currency exposure limits as a share of bank’s Tier I capital (1997). - NOP: introduction of a foreign currency liquidity ratio (2009). - LTV: discontinuation of LTV ratio in real estate (for housing loans for a first house and loans granted under special programs including housing to military personnel) (2008). - Reserve requirements: introduction of reserve requirement exemptions on local currency deposits in order to promote in local currency (2009). - Malaysia: - LTV: introduction of a maximum LTV ratio of 60% on real estate loans in 1995 (discontinued in 1998). - Lending ceiling: introduction of a limit on property lending equal to 20% of a bank’s portfolio in 1997 (discontinued in 1998). - Reserve requirements: increase in the statutory reserve requirement from 8.5% to 11.5% in 1994, and again to 13.5% in 1996 (reversed to 8% in 1998). - Risk weight: increase in risk weight for non-performing loans from 50% to 100% (2005). - LTV: introduction of 70% of LTV for the third house loan (2010). - Mexico: - Maturity mismatch in foreign currency: significant refinement of limits on maturity mismatch in foreign currency (1997). - Exposure limits: limits on interbank exposure set at 100% of a bank’s Tier I capital (2001). - Provisions: introduction of forward-looking loan loss provisioning (2010). - Mongolia: - NOP: introduction of limits on net open currency positions (the amount of a single foreign exchange open position shall not exceed 15% of the bank’s equity capital) (2010). - Maturity mismatch: introduction of limits on maturity mismatch (the difference between average durations of asset and liability shall not exceed 30% of total assets’ average duration). - Reserve requirements: increase in reserve requirement from 5% to 9% (2011). - Time-varying/dynamic provisioning: change in the rate of provisioning in response to the economic downturn (2010). - New Zealand: - Maturity mismatch: introduction of a liquidity mismatch ratio and a core funding ratio (2010). The minimum ratio of core funds to loans and advances was set at 65% and raised to 70% in July 2011. In 2011Q3, the authorities will review the likely impact of the further increase in the minimum ratio of core funds to 75% from July 2012. - Nigeria: - Lending ceiling: limiting capital market lending to a set proportion of a bank’s balance sheet (2010). - Liquidity: reduction of the liquidity ratio from 40% to 25% between September 2008 and April 2009. - NOP: reduction of foreign exchange open positions from 20% to 1% (2009). - Reserve requirements: reduction of the cash reserve requirement for commercial banks from 4% to 1% (2008-2009). - Norway: - Risk weight: increase in risk weights on loans with LTV above 60% from 50% to 100% (discontinued in 2001). - LTV: 90% cap on LTV for housing loans and 75% cap on LTV for home equity loans (2010). - DTI: introduction of caps on DTI (2010). Note: These limits are guidelines rather than hard caps — e.g., the LTV limit can be exceeded if the lender makes a special soundness evaluation. - Peru: - NOP: introduction of limits on NOP (2010). - Differentiated reserve requirements: use of differentiated reserve requirement for residents/non-residents, and domestic currency/foreign currency. The authorities also apply 60 percent of reserve requirements to external liabilities with maturity of less than two years. - Time-varying/dynamic provisioning: introduction of dynamic provisioning (2008). - Poland: - DTI: 50%-65% caps under stressed scenarios for loans to households starting in 2010. Lower cap of 42% for FX loans to households with banks having until end 2011 to adjust. - FX mortgage lending ceilings and tighter criteria: FX mortgage lending ceiling set at 50% of total mortgage lending introduced in 2010; haircuts in collateral for FX loans and stricter eligibility criteria for FX mortgages introduced in 2006. - Risk weights: Differentiated risk weights for mortgage in Polish zloty (PLN) and FX, with FX weight of 75% and 35% for PLN lending in 2008 (LTV for FX lending was reduced in 2010). The risk weight for all FX loans will be raised at 100% starting in 2012. - Reserve requirements: decrease in reserve requirements from 3.5% to 3% to increase banking sector liquidity (2009). - Capital: restrictions on profit distribution (2009). - Portugal: - Capital: tighter capital requirements for housing loans with an LTV ratio exceeding 75% (1999). - Provisions: tighter provisioning requirements for consumer loans (provisions for general consumer credit risks were raised to 1.5%). - Romania: - LTV: caps on LTV of 75% during 2004-2007. - DTI: caps on DIT of 30% for consumer loans and 35% for mortgages (2004-2007). Introduction of a cap on DIT for total household debt of 40% (2005). - FX lending ceiling: aggregate exposure from FX loans to unhedged borrowers limited to 300% of the credit institution’s own funds during 2004-2007. - Reserve requirements: increases in reserve requirements on deposits in foreign currency from 25% to 30% in 2004, from 30% to 40% in 2006 (the reserve requirement on deposits in domestic currency was reduced slightly from 18% to 16% in 2005). - Provisions: stricter loan provisioning and loan classification rules taking into account the currency risk of the borrowers (2002 and 2005). - Russia: - DTI: easing of DTI (2008-2009). - NOP: limits on net open currency positions (2008-2009). - Reserve requirements: decrease in mandatory reserve requirements (2008). - Provisions: The Bank of Russia eased the requirements for the evaluation of debt servicing on loans in connection with the formation of loan loss. - Serbia: - DTI (for households): introduction of caps on DTI in 2004 and recalibrated several times since then. - FX lending ceiling: introduction of 20% of minimum deposit on FX denominated non-mortgage loans to households in 2007 (it was raised once to 30% but reduced to 0% in 2008). - Differentiated reserve requirements for FX: introduction of differentiated reserve requirement for FX and Serbian Dinar (RSD) funds in 2005 to reduce FX loans. - NOP: tightening of maximum net open currency positions relative to capital from 30% to 10% in 2007 but with the onset of the financial crisis it was raised to 20% (end 2008). - Capital: introduction of 200% ceiling on the ratio of household loan portfolio to Tier I capital in 2006 (reduced to 150% in 2007). - Restrictions on profit distribution: dividend and bonus payments forbidden if the bank was under provisioned in 2008 — remains in place to this day. - Mandatory shortening of cash loans to maximum of 2 years duration: to slow down household lending, preempt large credit risks and prevent excessive leverage of the poorest households. - Singapore: - LTV: reduction of caps on LTV from 90% to 80% for all borrowers (2010). This was lowered to 70% for borrowers who have one or more outstanding housing loans at the point of applying for the new housing loan (later in 2010), and reduced further to 60% for borrowers who have one or more outstanding housing loans and to 50% for non-individuals (2011). - Lending Ceiling: (entry truncated in source content). *Source: _wp11238 - introduction of restriction on domestic banks and other institutional investors onshore from investing in Kimchi bonds (foreign currency denominated bonds issued by Korean banks and corporate) that are intended to be converted into Korean won for domestic use (2011) — extracted content.* ### introduction of caps on banks' ### _wp11238 - introduction of caps on banks' ### Macroprudential motivations and tools (selected highlights) - Cap on banks' loan exposures to the property sector (excluding residential mortgages for owner occupation) at 35 % of total non-bank exposure in 2009. - Disallowance of the Interest Absorption Scheme and interest-only loans for residential mortgages in 2009. - Maturity mismatch: revisions of regulation in 2008 to improve liquidity condition during the financial crisis (such as expansion of the range of eligible liquid assets). - Introduction of a new short-term liquidity indicator (2008). - Restrictions on profit distribution: central bank recommendation that banks would not distribute their entire profit from 2008 but use it to increase own funds (2009). - Capital surcharges: introduction of capital surcharges for systemically important institutions (2008). - Minimum liquid asset ratios: banking sector required to hold a minimum of 5 percent of its liability as liquid assets (since 1996). - Limits on net open currency position: net open position of foreign currency of each bank limited to 10 percent of its net qualifying capital and reserve funds (since the 1990s). - Time-varying/dynamic provisioning: introduction in 2000 (revised in 2004); in some cases not applied universally across the banking sector. - Risk weights: introduction of sector-dependent asset risk weights in 2008, establishing higher risk weight for mortgages that exceed an LTV of 95% for residential property and 80% for others. - LTV caps: example LTV cap of 85% for mortgages (2010). - Leverage ratio: introduction of leverage ratio for systemically important banks (2008). - Caps on foreign currency lending: moderation of FX lending by allowing non-FX earnings companies to obtain FX loans (2009). - FX liquidity: allowing banks to temporarily classify FX loans as FX liquidity to help meet FX liquidity adequacy ratios (2008). - Introduction of caps on DTI (2004). - Introduction of limits on maturity mismatch and a core funding ratio. - Introduction of higher risk weights for FX loans (2005). - Limits on net open currency positions (NOP should not exceed 150% capital). ### Country case summaries (selected entries) - Slovakia (2008-2009) - Background and Motivation: - Small open economy, mainly exports. Impact of global recession started in 2008Q4. Concern about risk of liquidity outflows during the financial crisis. - Main motivations: strengthen short-term liquidity position of banks; increase the capital base; prevent liquidity outflows of liquid assets. - Action – Macroprudential Tools Used: - Maturity mismatch: introduction of a new short-term liquidity indicator (2008). - Restrictions on profit distribution: central bank recommendation that banks would not distribute their entire profit from 2008 but use it to increase own funds (2009). - South Africa (2008) - Background and Motivation: - Global financial crisis led to large capital outflows, lower stock prices, weaker currency. Impact on financial sector was low due to existing macroprudential regulations and prudent risk management. Basel II entered into effect on January 2008, setting capital requirements for exchange rate risk. - Motivation: strengthen the resilience of financial system. - Action – Macroprudential Tools Used: - Capital surcharges: introduction of capital surcharges for systemically important institutions (2008). - Note: - Authorities have used minimum liquid asset ratios (banking sector required to hold a minimum of 5 percent of its liability as liquid assets) since 1996 and limits on net open currency position (net open position limited to 10 percent of net qualifying capital and reserve funds) since the 1990s. - Spain (2000-2008) - Background and Motivation: - Long boom from 1996 to 2007 driven by immigration, foreign property investors, demographic changes, financial liberalization, and convergence with EU. Real interest rates pressed down, mortgage availability increased, tight rental market. House prices rose 121 percent from trough to peak. - Motivations: reduce procyclicality of loan loss provisions and stem credit growth; build up a buffer in good times to be used in bad times. - Action – Macroprudential Tools Used: - Time-varying/dynamic provisioning: introduction in 2000 (revised in 2004); dynamic provisioning not applied universally (Cajas omitted). - Risk weight: introduction of sector-dependent asset risk weights in 2008 (higher risk weight for mortgages exceeding LTV of 95% for residential property and 80% for others). - Sweden (2010) - Background and Motivation: - Signs of excessive indebtedness making borrowers vulnerable to real estate price declines. - Motivations: stem an unsound trend in the credit market; protect consumers. - Action – Macroprudential Tools Used: - LTV: LTV cap of 85% for mortgages (2010). - Switzerland (2008) - Background and Motivation: - Deterioration in general economic and financial conditions in 2008 due to global financial crisis; refinancing difficulties, declining credit growth, losses at financial institutions. - Motivations: reduce procyclicality; strengthen resilience of the financial system. - Action – Macroprudential Tools Used: - Leverage ratio: introduction for systemically important banks (2008). - Capital surcharges: introduction for systemically important institutions (2008). - Thailand (2002-2011) - Background and Motivation: - Rapid credit growth, double-digit rises in housing prices, and massive capital inflows in early 2000s. House prices declined since 2006, spiked 10 percent quarter-on-quarter in 2010Q2. Commercial bank loans grew strongly over the summer of 2010. - Motivations: reduce cyclicality of the real estate sector; reduce currency risk. - Action – Macroprudential Tools Used: - LTV: cap of 70% on the LTV ratio (2003); increase in the LTV ratio for high value mortgages (above 10 million baht) from 70% to 80% (2009). - DTI: introduction of caps on DTI (2004). - NOP: introduction of limits on net open currency positions (2002). - Risk weight: higher risk weight for high value mortgages (above 10 million baht) with LTV above 80% (2009); higher risk weight for residential mortgages (less than 10 million baht) with LTV above 90% (2011). - Turkey (2008-2009; 2009-2010) - Background and Motivation (2008-2009): - Global financial crisis caused an FX liquidity squeeze; banks reduced FX loans and Eurobond holdings. - Motivations: strengthen and preserve financial position of banks; address negative effects of global financial crisis. - Action – Macroprudential Tools Used (2008-2009): - Caps on foreign currency lending: moderation of FX lending by allowing non-FX earnings companies to obtain FX loans (2009). - FX liquidity: change in FX liquidity ratio by allowing banks to temporarily classify FX loans as FX liquidity (2008). - Restriction of profit distribution: introduction of restrictions on profit distribution (2008). - Background and Motivation (2009-2010): - Rapid rebound after crisis, rapid increase in domestic demand, rapid credit growth, increased foreign currency borrowing by banks. - Motivation: slow credit growth and improve credit quality. - Action – Macroprudential Tools Used (2009-2010): - LTV: introduction of caps on the LTV ratio for real estate loans (2010). - Uruguay (most of 1990s; 2001; 2005) - Background and Motivation: - Relatively high economic growth during most of the 1990s accompanied by vulnerabilities: financial dollarization and sharp increase in non-resident deposits. Main motivations: reduce currency risk and liquidity mismatch. - Action – Macroprudential Tools Used: - NOP: introduction of limits on net open currency positions (NOP should not exceed 150% capital). - Maturity mismatch: introduction of limits on maturity mismatch. - Core funding ratio: introduction of a core funding ratio. - Time-varying/dynamic provisioning: introduction in 2001. - Risk weight: introduction of higher risk weights for FX loans (2005). *Source: _wp11238 - introduction of caps on banks'* --- _Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11238.pdf_