## Elements of Effective Macroprudential Policies: Lessons from International Experience

## Source details

**Canonical URL:** [Elements of Effective Macroprudential Policies: Lessons from International Experience](https://www.bookstore.imf.org/images/IMF_FSB_BIS_2016.pdf)

## Other formats

- [Markdown version](/images/imf_fsb_bis_2016.pdf.md)
- [Structured JSON version](/images/imf_fsb_bis_2016.pdf.json)

---

### Definition, objectives and scope
- Definition and systemic risk:
  - Macroprudential policy is defined as "the use of primarily prudential tools to limit systemic risk" (Crockett 2000, FSB/IMF/BIS 2011, IMF 2013).
  - Systemic risk: "the risk of widespread disruption to the provision of financial services that is caused by an impairment of all or parts of the financial system, and which can cause serious negative consequences for the real economy" (IMF/BIS/FSB 2009).
  - Systemic risk dimensions:
    - time dimension: vulnerabilities related to the build-up of risks over time;
    - cross-sectional (structural) dimension: vulnerabilities from interconnectedness and the associated distribution of risk within the financial system at any given point in time.
  - Fundamental notion: negative externalities from a disruption or failure in a financial institution, market or instrument.
- Intermediate objectives of macroprudential policy:
  - increase the resilience of the financial system to aggregate shocks by building and releasing buffers that help maintain the ability of the financial system to function effectively, even under adverse conditions;
  - contain the build-up of systemic vulnerabilities over time by reducing procyclical feedback between asset prices and credit and containing unsustainable increases in leverage, debt stocks, and volatile funding;
  - control structural vulnerabilities within the financial system that arise through interlinkages, common exposures, and the critical role of individual intermediaries in key markets that can render individual institutions "too-big-to-fail".
- Scope and targets:
  - Primary focus typically on the banking system; need to monitor systemic risks outside banking as capital market activity and market-based financing expand (FSB 2011a, CGFS 2012, IMF 2013).
  - Macroprudential tools must have a systemic orientation in objective, calibration and governance to qualify as macroprudential measures (2011 Progress Report).
- Interactions with other policies and limits:
  - Interacts with microprudential regulation and supervision, crisis management and resolution, monetary policy, and fiscal policy.
  - Macroprudential policy is not primarily aimed at managing aggregate demand or the business cycle; objective is to strengthen financial system defenses to support continued supply of credit and other financial services through adverse conditions.
  - Macroprudential policy and capital flow management measures (CFMs) have different objectives but can overlap; CFMs are designed to limit capital flows by influencing their size or composition. Macroprudential measures that limit capital flows would be considered CFMs as well (IMF 2012, IMF 2015).
  - Neither macroprudential policy nor CFMs should substitute for warranted macroeconomic adjustment or appropriate microprudential regulation and supervision.
- Experience and caveats:
  - Experience growing with many countries establishing dedicated institutional arrangements and tools; empirical evidence remains tentative as many countries' experience does not yet span a full financial cycle.
  - No "one-size-fits-all" approach; accumulated experience highlights useful elements for macroprudential policymaking.

### Institutional arrangements and powers
- General principles:
  - Adequate institutional foundations are essential; arrangements must suit country-specific circumstances.
  - Effective macroprudential policy is often supported by a clear mandate with well-defined objectives, adequate powers, and strong accountability (BIS 2011, CGFS 2012, IMF 2013).
  - Institutional design can mitigate biases toward inaction when benefits are uncertain and costs more immediate.
- Mandate, governance and accountability:
  - Clear mandate forms basis for assigning responsibility; many jurisdictions assign main mandate to an influential central body with convening power and a broad system-wide view.
  - Roles for central banks:
    - Approaches include: central bank board (or governor) as decision-maker (examples: Ireland, New Zealand); governor chairs policymaking committee (examples: Malaysia, South Africa, UK); central bank provides systemic risk analysis and proposals (examples: France, Germany); leading role in regulation/supervision of SIFIs (example: US).
  - Involvement of regulatory and supervisory authorities:
    - Regulatory/supervisory authorities usually participate to provide expertise and because implementation often rests with them.
    - Potential conflicts when macro- and microprudential perspectives diverge; coordination mechanisms include coordinating/advisory bodies or strong central bank roles (examples: Australia, Sweden, Finland).
  - Role of the Ministry of Finance:
    - Participation varies: non-voting member (UK), voting member (Poland), or chair (France, Germany, US).
    - Arrangements to mitigate political economy risks include strong central bank voice (Mexico, Netherlands) or veto powers (Germany).
  - Use of independent external experts:
    - External experts can be voting members, advisory committee members (ESRB), or invited ad hoc to reduce groupthink.
- Objectives, transparency and accountability mechanisms:
  - Well-defined objectives foster willingness to act and provide basis for accountability (BIS 2011, CGFS 2012, IMF 2013, IMF 2014).
  - Secondary objectives can ensure consideration of costs and trade-offs (examples: ESRB, UK).
  - Transparency tools: financial stability reports, policy statements, meeting records; sometimes legally required (France, Germany, UK).
- Decision-making practice and supporting units:
  - Meeting frequency often quarterly or semi-annually; voting typically simple or qualified majorities to avoid paralysis (examples: Germany, Ireland, UK).
  - Dedicated financial stability units within central banks increasingly common; these analyze systemic risks, develop and monitor systemic risk indicators, and prepare analysis and proposals (examples: Germany, India, Netherlands, UK, US).

- Powers: purpose, strength and types
  - Macroprudential powers required to obtain information, influence calibration of regulatory constraints, designate systemically important institutions, and change the regulatory perimeter (IMF 2011a, CGFS 2012, FSB 2013a, FSB 2013b).
  - Power types:
    - "hard (direct)" — direct control over macroprudential tools or ability to direct other regulatory authorities;
    - "semi-hard" — formal recommendations to other regulatory authorities, coupled with a ‘comply or explain’ mechanism;
    - "soft" — opinions, warnings, or recommendations not subject to comply or explain (IMF 2013).
  - Soft powers alone are unlikely to be sufficient; combinations can be effective (examples: UK, US).
- Uses and operational implications of hard, semi-hard and soft powers:
  - Advantages of hard powers: avoid delay/frictions, provide credible 'stick', act as 'backup', collect information directly from firms (examples: New Zealand; across EU through CRD IV/CRR; US; ECB top-up powers; Germany, UK, US).
  - Semi-hard (‘comply or explain’) increases chance of action, ensures transparency and accountability, and maintains operational independence (examples: ESRB, Germany, UK, US).
  - Soft powers useful to initiate establishment of new tools or perimeter changes (example: UK) and to influence behavior via communications when cooperation beyond regulators is needed.

### Operationalising the use of tools
- Calibrating policy responses to risks:
  - Choices along dimensions: gradual versus forceful; broad-based versus targeted; single versus multiple tools; rules versus discretion (CGFS 2012, IMF 2013).
  - Gradual vs forceful:
    - Multiple indicators or sharply rising risks may call for more forceful tightening.
    - Moderate readings can justify gradual response, including targeted communications.
    - Partial activation of time-varying tools possible (examples: Hong Kong SAR, Sweden).
    - Early and gradual action can manage uncertainty; some countries calibrate LTV and LTI constraints so they are not immediately binding (examples: Ireland, UK).
  - Broad-based vs targeted:
    - Broad-based vulnerabilities call for broad tools; specific vulnerabilities can be addressed via targeted approaches (example: tight LTV and DSTI caps for household foreign-currency borrowing in Poland).
  - Single vs multiple tools:
    - Marginal benefit of tightening one tool decreases; complementary tools mitigate distortions (CGFS 2012, IMF 2013).
    - Examples: LTV, LTI, and DSTI limits complement capital tools; DSTI limits enhance LTV caps by tying debt burdens to income (IMF 2014a).
  - Rules vs discretion:
    - Rules: act as automatic stabilisers and pre-commitment devices ("set and forget"; CGFS 2010) but prone to arbitrage (Borio 2010).
    - Discretionary measures allow tailoring; most countries use 'guided discretion'.
    - Some use tools calibrated on ex ante rules, notably dynamic provisioning (examples: Peru, Spain, Uruguay; IMF 2014a).
  - Practical calibration constraints:
    - Lack of data and uncertainties mean judgment retains overriding role (CGFS 2016).
    - Gradual tightening common but must be balanced against need for decisive action when risks are elevated.
- Considering costs and benefits (ex ante perspective):
  - Formal cost-benefit analyses important but difficult; costs and benefits are hard to quantify (CGFS 2016).
  - Ex ante calibration considers:
    - (i) adjustment costs to the financial industry;
    - (ii) efficiency costs;
    - (iii) costs to output growth (IMF 2014).
  - Adjustment costs:
    - Phase-in or step-wise tightening can mitigate costs for capital and liquidity tools.
    - BCBS recommends a notice period of up to 12 months to provide banks time to meet a CCyB requirement.
    - Examples: Korea announced a ceiling on the loan-to-deposit ratio in December 2009, expected to be met by end of 2013; New Zealand implemented a core funding ratio in steps.
    - Measures affecting only flow of new lending (e.g. LTV constraints, marginal reserve requirements) do not require phase-in; long pre-announcement can cause frontloading (IMF 2014).
  - Efficiency costs:
    - Prefer initially tools affecting lender behaviour (capital-based) before borrower constraints to reduce distortions (example: Israel).
    - Design features can contain efficiency costs: caps on share of loans at high LTV (New Zealand) or high LTI multiples (UK).
    - Differentiation for first-time buyers used in Finland, Ireland, Israel, Singapore; by property type and region in Malaysia, Korea, New Zealand.
  - Output costs:
    - Balance sheet constraints can have some adverse short-run impact on output but likely weak and overcompensated by long-term benefits (BIS 2010, IMF 2013d).
    - Tools addressing flow of new credit (e.g. LTV, DSTI) generally expected to have greater impact on output (IMF 2013d); gradual tightening recommended when economy is weak (example: the Netherlands post crisis).
- Assessing and addressing leakage:
  - Leakages: migration of financial activity outside scope/enforcement; can be domestic or cross-border.
  - Strategies: expand scope to non-bank and foreign providers, jurisdictional reciprocity agreements (BCBS 2010), greater host control over foreign affiliates (CGFS 2014, IMF 2014).
  - Capital-based tools vulnerable to domestic leakages (Annex 2); cross-border leakages challenging when affiliates are branches and local corporates borrow abroad.
  - Household loan restrictions (LTV, DSTI) easier to contain and can be enforced on regulated institutions including non-banks and foreign branches (BoE 2011); examples: Estonia, Ireland, Romania, UK.
  - Liquidity-related tools: maturity transformation can migrate to non-banks; liquidity tools can be extended to non-bank providers with supervisory cooperation (example: US money market mutual funds).
- Evaluating effectiveness (ex post perspective):
  - Ex post evaluation gauges need for recalibration or additional measures (CGFS 2012, ESRB 2014, IMF 2014).
  - Evaluation should assess:
    - (i) extent to which measures improved resilience; and
    - (ii) effects on behaviour, credit dynamics, and asset prices.
  - Methods: event studies on resilience, repeated stress testing, construction of counterfactual paths (examples: Euro area, Hong Kong SAR, Canada, Korea, New Zealand).
- Considering potential for tools to be relaxed:
  - Ex ante consideration of conditions for relaxing tools is integral (CGFS 2012, IMF 2013, IMF 2014).
  - Some tools can be designed for automatic relaxation (e.g. dynamic provisions); policymakers typically decide on timing and speed of relaxation.
  - Guiding objective: prevent disruptions to credit provision that adversely affect the real economy.
  - Indicators for relaxation differ from activation indicators; market-based and flow variables better capture turning points (Drehmann and others 2010, IMF 2013).
- Improving the information base:
  - Closing information gaps improves effectiveness (FSB/IMF/BIS 2011, IMF 2013).
  - Micro-data on distribution of actual LTV and DSTI ratios helpful (examples: Ireland, Singapore, Sweden, UK).
  - Data on bilateral exposures between financial institutions and on nonbank systemic vulnerability often lacking.
  - Data collection enhancements: improved sharing, expanded supervisory reporting, infrastructure investment (credit registers, surveys).
  - International initiatives: FSB-IMF-BIS Data Gaps Initiative; work on a common data template for global systemically important non-bank financial institutions starting with insurance companies (FSB/IMF 2015).

### International consistency of macroprudential policy
- Cross-border effects and coordination:
  - Macroprudential policy in financially integrated economies subject to cross-border effects: positive externalities, leakages, undesirable spillovers, and migration of activities (IMF 2013, Caruana 2016).
  - Positive externalities: effective domestic macroprudential policy that reduces crisis probability can lower negative trade and financial spillovers regionally/internationally (IMF 2013, Caruana 2016).
  - International arrangements can buttress national frameworks, including IMF surveillance and FSAP assessments, FSB peer reviews, and BIS meetings.
  - Cross-border coordination example: reciprocity agreed for the Basel III countercyclical capital buffer (BCBS 2010).
- Leakage and reciprocity:
  - National policies can face leakage from increases in cross-border borrowing; reciprocity applies same constraint to all relevant credit exposures to borrowers in a given country.
  - Basel III CCyB includes reciprocity; EU authorities developed a similar voluntary approach (ESRB 2015a); guidance exists on exposures into third countries that fail to take measures (ESRB 2015b).
  - Empirical patterns: cross-border leakages strong for capital requirements (especially where affiliates are branches), weaker for loan restriction tools such as LTV and DSTI (Reinhardt and Sowerbutts 2015).
  - Macroprudential tools appear less effective in reducing credit growth in more developed/open financial systems (Cerutti and others 2015).
- Spillovers and timing:
  - Tightening in one country can reduce that country’s banks’ cross-border lending, creating undesirable spillovers (IMF 2014).
  - Retrenchment more likely when measures are ill-timed and occur during financial stress.
  - Design and calibration may be modified to minimize adverse effects on other jurisdictions while achieving domestic objectives.
  - Expansionary effects can occur when groups respond to domestic tightening by increasing lending abroad; recipient economies may need counterbalancing action.
- Migration of activities and minimum standards:
  - Strengthening resilience in one country can induce migration to other countries, risking concentration of risky activities in lightly regulated jurisdictions (Viñals and Nier 2014).
  - Responses include minimum standards and supplementary agreements (examples: BCBS framework for domestic systemically important banks (BCBS 2015); minimum internal TLAC requirements for each resolution entity within each G-SIB (FSB 2015)).
  - Ongoing work to ensure resilience and resolvability of systemically important central counterparties (FSB/BCBS/CPMI/IOSCO 2015).
  - International monitoring of implementation complements standards and guidance.
- Use and trends in tool deployment (dataset):
  - The underlying database covers 64 countries, of which 32 are advanced economies according to IMF (World Economic Outlook) classification.
  - “Usage” counts number of countries using instruments and assumes continued use once introduced.
  - Tightening and loosening episodes coded as +1 and -1 respectively; cumulative sums shown over time for groups of instruments.
- Institutional models for policymaking:
  - No single model fits all; rising prevalence of assigning main mandate to a well-identified authority, committee, or interagency body, generally with an important central bank role.
  - Illustrative typology:
    - Model 1: Main mandate to the central bank, Board or Governor makes decisions (examples: Czech Republic, Ireland, New Zealand, Singapore).
    - Model 2: Main mandate to a dedicated committee within the central bank (examples: Malaysia, UK).
    - Model 3: Main mandate to an interagency committee outside the central bank, with the central bank participating (examples: France, Germany, Mexico, US).
  - Variants exist (e.g., Sweden, Canada, Australia, Japan).
- Effectiveness of macroprudential tools (evidence summary, Annex 2):
  - Evidence expanding but not yet conclusive on which instruments work best (Akinci and Ohmstead-Rumsey 2015; Claessens 2015; Cerutti and others 2015).
  - Capital-based tools:
    - Support resilience and credit supply during downturns; limited effects in upswing.
    - SIFI surcharges enhance resilience (BCBS 2010, BIS 2015).
    - Some studies suggest short-term dampening on credit growth; longer-run impact limited (Dagher and others 2016).
    - Dynamic provisioning can help smooth post-crisis credit supply (Jiménez and others 2012).
  - Sectoral capital requirements:
    - Increase resilience via additional buffers; evidence on effects on credit growth varies.
    - Some studies find sectoral requirements or risk weights limit loan growth of targeted sectors (BoE 2014; IMF 2013).
    - Leakage may prevent sectoral tools from containing credit booms (Crowe and others 2013).
  - Borrower-based tools:
    - Limits on LTV and DSTI ratios enhance borrower resilience (Hallissey and others 2014) and moderate lending growth (Igan and Kang 2011; Akinci and Ohmstead-Rumsey 2015).
    - Effects on house price growth appear limited (Kuttner and Shim 2013).
  - Liquidity-related tools:
    - Changes in reserve requirements may help moderate credit growth (IMF 2013a; Lim and others 2011), though some studies find no or weak impacts (Kuttner and Shim 2013; Bruno and others 2015).
    - Experience with other liquidity-related tools remains limited.
  - Evidence of leakage:
    - Studies document leakage to non-bank providers of credit (Cizel and others 2016).
    - Cross-border leakages particularly affect capital requirements where affiliates operate as branches; loan restriction tools show weaker cross-border leakage (Reinhardt and Sowerbutts 2015).
    - Evidence on cross-border spillover effects summarized in Buch and Goldberg 2016.

*Source: IMF, FSB, BIS (2016).*

### 1. Definition, objectives and scope ....................................................................................

### 1. Definition, objectives and scope

### Definition and systemic risk
- Macroprudential policy is defined as "the use of primarily prudential tools to limit systemic risk" (Crockett 2000, FSB/IMF/BIS 2011, IMF 2013).
- Systemic risk: "the risk of widespread disruption to the provision of financial services that is caused by an impairment of all or parts of the financial system, and which can cause serious negative consequences for the real economy" (IMF/BIS/FSB 2009).
- Systemic risk has two dimensions:
  - time dimension: vulnerabilities related to the build-up of risks over time;
  - cross-sectional (structural) dimension: vulnerabilities from interconnectedness and the associated distribution of risk within the financial system at any given point in time.
- Fundamental to the definition is the notion of negative externalities from a disruption or failure in a financial institution, market or instrument.

### Intermediate objectives of macroprudential policy
- The document identifies three interlocking intermediate objectives (FSB 2009, CGFS 2010, IMF 2013):
  - increase the resilience of the financial system to aggregate shocks by building and releasing buffers that help maintain the ability of the financial system to function effectively, even under adverse conditions;
  - contain the build-up of systemic vulnerabilities over time by reducing procyclical feedback between asset prices and credit and containing unsustainable increases in leverage, debt stocks, and volatile funding;
  - control structural vulnerabilities within the financial system that arise through interlinkages, common exposures, and the critical role of individual intermediaries in key markets that can render individual institutions "too-big-to-fail".

### Scope and targets
- Macroprudential policy aims at containing risks across the financial system as a whole and typically applies its policy levers to the banking system, given banks' role as significant providers of credit.
- As capital market activity and market-based financing expand, policymakers need to monitor systemic risks outside the banking system and develop policy responses to contain those risks (FSB 2011a, CGFS 2012, IMF 2013).
- Macroprudential tools must have a systemic orientation in objective, calibration and governance to qualify as macroprudential measures (2011 Progress Report).

### Interactions with other policies and limits
- Macroprudential policy interacts with microprudential regulation and supervision, crisis management and resolution, monetary policy, and fiscal policy.
- Boundaries and interactions can create complementarities and tensions; suitable institutional arrangements may be needed to resolve these and ensure appropriate instrument use and policy mix.
- Macroprudential policy is not primarily aimed at managing aggregate demand or the business cycle; it aims to strengthen financial system defenses to support continued supply of credit and other financial services through adverse conditions and thereby reduce the frequency and severity of financial crises.
- Macroprudential policy and capital flow management measures (CFMs) have different objectives but can overlap:
  - CFMs are designed to limit capital flows by influencing their size or composition.
  - Macroprudential measures designed to limit systemic risks by limiting capital flows would be considered CFMs as well, and common principles would apply (IMF 2012, IMF 2015).
- Neither macroprudential policy nor CFMs should substitute for warranted macroeconomic adjustment or appropriate microprudential regulation and supervision.

### Experience and caveats
- Experience with macroprudential policy is growing, with many countries establishing dedicated institutional arrangements and designing/implementing tools; empirical evidence remains tentative as many countries' experience does not yet span a full financial cycle.
- There is no "one-size-fits-all" approach; accumulated experience highlights a number of useful elements for macroprudential policymaking.

---

### 2. Institutional arrangements

### General principles
- Adequate institutional foundations are essential; arrangements must suit country-specific circumstances and institutional backgrounds.
- Effective macroprudential policy is often supported by a clear mandate with well-defined objectives, adequate powers, and strong accountability (BIS 2011, CGFS 2012, IMF 2013).
- Macroprudential policy may face biases toward inaction or insufficiently timely action when benefits are uncertain and costs are more immediate; institutional design can foster willingness to act and legitimize action.

### Mandate, governance and accountability (Section 2.1)
- A clear mandate forms the basis for assigning responsibility for macroprudential decisions; many jurisdictions assign the main mandate to an influential central body with substantial convening power and a broad system-wide view.
- Roles for central banks:
  - Central banks often play an important role to harness expertise, incentives to act, and independence.
  - Approaches include: central bank board (or governor) as decision-maker (examples: Ireland, New Zealand); governor chairs policymaking committee (examples: Malaysia, South Africa, UK); central bank provides systemic risk analysis and proposals to decision-makers (examples: France, Germany); leading role in regulation/supervision of SIFIs (example: US).
- Involvement of regulatory and supervisory authorities:
  - Regulatory/supervisory authorities usually participate to provide expertise and because implementation often rests with them.
  - Supervisory authorities are incorporating macroprudential perspectives in firm supervision (e.g., intensified supervision of SIFIs).
  - Potential conflicts can arise when macro- and microprudential perspectives diverge; coordination mechanisms include coordinating/advisory bodies or giving central bank a strong role on decision-making boards (examples: Australia, Sweden, Finland).
- Role of the Ministry of Finance:
  - Participation varies: non-voting member (UK), voting member (Poland), or chair (France, Germany, US).
  - Involvement can create political legitimacy and enable discussion of policy choices in other fields.
  - Arrangements to mitigate political economy risks and protect independence include strong central bank voice (Mexico, Netherlands) or veto powers (Germany).
- Use of independent external experts:
  - Independent external experts can be voting members (France, UK), advisory committee members (ESRB), or invited ad hoc (Germany, Netherlands) to reduce groupthink and inject independent perspectives.

### Objectives, transparency and accountability mechanisms
- Well-defined policy objectives foster ability and willingness to act and provide the basis for accountability to the legislature and the public (BIS 2011, CGFS 2012, IMF 2013, IMF 2014).
- Secondary objectives can be included to ensure consideration of costs and trade-offs (examples: ESRB, UK).
- Transparency tools include financial stability reports, policy statements, and meeting records; in some cases these are legally required (France, Germany, UK).
- Communication strategies should convey financial stability assessments clearly, link assessments to policy actions, and manage public expectations about policy capabilities.

### Decision-making practice and supporting units
- Formal meeting frequency for decision-making bodies is often quarterly or semi-annually to foster timely engagement; central bank boards acting as decision-makers may hold dedicated macroprudential meetings (ECB, Norway).
- Voting typically uses simple or qualified majorities rather than unanimity to avoid paralysis; authorities often strive for consensus (examples: Germany, Ireland, UK).
- Dedicated financial stability units are increasingly common:
  - Such units within central banks (or standing subcommittees) analyze systemic risks, develop and monitor systemic risk indicators, and prepare analysis and proposals for decision-makers (examples: Germany, India, Netherlands, UK, US).
  - These units can function as the secretariat for the macroprudential body.

*IMF-FSB-BIS – Elements of Effective Macroprudential Policies: Lessons from International Experience*

### 2.2 Powers

### 2.2 Powers

### Purpose and scope of macroprudential powers
- Macroprudential policy requires powers that ensure the ability to act to: obtain information from other authorities and fill data gaps (information powers); influence the activation and calibration of regulatory constraints (calibration powers); influence the designation of individual institutions as systemically important (designation powers); and initiate changes in the regulatory perimeter to capture financial institutions whose activities may give rise to financial stability risks (IMF 2011a, CGFS 2012, FSB 2013a, FSB 2013b).

### Strength and types of powers
- Powers can be:
  - "hard (direct)" — giving policymakers direct control over macroprudential tools or the ability to direct other regulatory authorities;
  - "semi-hard" — enabling policymakers to make formal recommendations to other regulatory authorities, coupled with a ‘comply or explain’ mechanism;
  - "soft" — enabling policymakers to express an opinion, or warning, or a recommendation that is not subject to comply or explain (IMF 2013).
- Each type can be useful, and effectiveness may benefit from a combination of these powers (as in the UK and the US); soft powers alone are unlikely to be sufficient to ensure effectiveness (CGFS 2010, IMF 2013).

### Uses, advantages and operational implications of hard (direct) powers
- Where macroprudential policymakers have hard (direct) powers, they are usually established with respect to a well-defined set of tools that can:
  - help control the build-up of risks over time (examples: New Zealand; across the EU member states through the Capital Requirements Directive and Regulation CRD IV/CRR and national laws);
  - effect the designation of systemically important institutions (example: the US).
- Advantages of direct powers:
  - avoid delay and frictions that may arise when implementation relies on other policymakers;
  - provide a ‘stick’ that macroprudential policymakers can credibly threaten to use, increasing effectiveness of other policies;
  - serve as a ‘backup’ if other regulatory authorities are unwilling or unable to act (example: the ECB’s top-up powers);
  - can extend to the power to collect information directly from firms (examples: Germany, UK, US).

### Semi-hard powers: recommendations with ‘comply or explain’
- A power to recommend actions, coupled with a ‘comply or explain’ mechanism, can be broad in scope (BIS 2011, IMF 2013) and has become very common (examples: ESRB, Germany, UK, US).
- Benefits of ‘comply or explain’ recommendations:
  - increase chance of action being taken;
  - ensure transparency and accountability of the relevant actors;
  - maintain operational independence of the recipient authority.
- A recommendation by the macroprudential authority may help recipient authorities overcome industry opposition or political pressure.
- Even recommendations not subject to ‘comply or explain’ can gain force where the macroprudential authority is mandated by law to publicly issue periodic recommendations (example: Norway).

### Soft powers, communications, and when they are appropriate
- Soft powers can complement stronger powers and extend influence beyond prudential tools or the existing regulatory perimeter.
- Appropriate uses of soft powers:
  - to initiate establishment of new macroprudential tools or changes in the legal framework to extend the regulatory perimeter (example: the UK);
  - when effective mitigation of systemic risk requires cooperation beyond regulatory authorities, for example where tax distortions fuel the build-up of debt (IMF 2013).
- Communications can be targeted directly at the financial industry to influence behaviour, reduce policy uncertainty, and guide agents’ expectations with respect to the likely policy path (CGFS 2012, IMF 2014, ESRB 2014).

*IMF_FSB_BIS_2016 - 2.2 Powers*

### 3.3 Operationalising the use of tools

### 3.3 Operationalising the use of tools

### Calibrating policy responses to risks
- Policy approach should be commensurate with the profile of risks, involving choices along dimensions set out in CGFS 2012 and IMF 2013.
- Gradual versus more forceful approaches:
  - Multiple indicators signalling elevated risks, or sharply rising risks, may call for a more forceful tightening of available tools.
  - Moderate indicator readings can justify a more gradual response, including targeted communications ahead of implementation.
  - Partial activation of time-varying tools (e.g. CCyB) is possible (examples: Hong Kong SAR, Sweden).
  - Early and gradual action can manage uncertainty over effects of instruments; some countries calibrate LTV and LTI constraints so they are not immediately binding but slow imbalance build-up (examples: Ireland, UK).
- Broad-based versus targeted action:
  - Broad-based build-up of vulnerabilities generally calls for activation/tightening of broad-based tools affecting all exposures, potentially including capital and liquidity buffers.
  - Specific or sectoral vulnerabilities can be addressed more efficiently via targeted approaches (example: tight LTV and DSTI caps for household foreign-currency borrowing as in Poland).
- Single versus multiple tools:
  - Marginal benefit of tightening any one tool eventually decreases due to distortions and circumvention incentives.
  - Complementary tools can mitigate such effects by addressing risks from several angles (CGFS 2012, IMF 2013).
  - Examples: LTV, LTI, and DSTI limits address demand and borrower resilience and complement capital tools that act on supply; DSTI limits enhance LTV caps by tying debt burdens to income (IMF 2014a).
- Rules versus discretion:
  - Rules’ main advantage: once in place they do not require recurring justification (“set and forget”; CGFS 2010), can act as automatic stabilisers and pre-commitment devices, and reduce inaction bias.
  - Rules can be prone to arbitrage and may shift risks in unintended ways (Borio 2010); design hampered by lack of reliable indicators to automate instrument use.
  - Discretionary measures avoid difficulties of designing effective rules and make tailoring to prevailing risk profiles easier.
  - Most countries rely on ‘guided discretion’, combining indicators with judgment; some use tools calibrated on ex ante rules, notably dynamic provisioning (examples: Peru, Spain, Uruguay; IMF 2014a).
- Practical calibration constraints:
  - Lack of data and uncertainties over transmission mean quantitative analysis is growing but judgment retains an overriding role (CGFS 2016).
  - Gradual tightening and preference for less intrusive tools initially are common, but must be balanced against need for decisive action when risks are elevated.

### Considering costs and benefits (ex ante perspective)
- Formal cost-benefit analyses are important but difficult because costs and benefits of macroprudential measures are hard to quantify and model-based welfare analyses are not well-developed (CGFS 2016).
- Calibration can consider targeted benefits and potential costs ex ante, including:
  - (i) adjustment costs to the financial industry;
  - (ii) efficiency costs;
  - (iii) costs to output growth (IMF 2014).
- Adjustment costs:
  - Capital and liquidity tools imposing minimum ratios may create adjustment costs for financial firms; mitigated through phase-in or step-wise tightening.
  - BCBS recommends a notice period of up to 12 months to provide banks time to meet a CCyB requirement.
  - Examples: Korea announced a ceiling on the loan-to-deposit ratio in December 2009, expected to be met by end of 2013; New Zealand implemented a core funding ratio in steps.
  - Measures affecting only flow of new lending (e.g. LTV constraints, marginal reserve requirements) do not require phase-in; long pre-announcement can cause frontloading (IMF 2014).
- Efficiency costs:
  - To reduce distortions, preferable initially to introduce tools affecting lender behaviour (capital-based) before quantitative borrower constraints (LTV, LTI, DSTI) as in Israel.
  - Cases exist where both sets should be introduced simultaneously or asset-side tools preferred (IMF 2014a).
  - Design features can contain efficiency costs: caps on share of loans at high LTV (New Zealand) or high LTI multiples (UK) constrain rather than prohibit such credit.
  - Less stringent constraints for first-time buyers used in Finland, Ireland, Israel, Singapore; differentiation by property type and region used in Malaysia, Korea, New Zealand.
- Output costs:
  - Balance sheet constraints (capital and liquidity tools) can have some adverse short-run impact on output, but effects likely weak and overcompensated by long-term benefits from reduced crisis likelihood/cost (BIS 2010, IMF 2013d).
  - Short-run effects are uncertain and can be greater with aggressive tightening or when tightening occurs during financial stress.
  - To avoid strong output effects, tighten balance sheet constraints well ahead of stress.
  - Tools addressing flow of new credit (e.g. LTV, DSTI) generally expected to have greater impact on output (IMF 2013d), calling for more gradual tightening especially when the economy is weak (example: the Netherlands post crisis).

### Assessing and addressing leakage
- Leakages: migration of financial activity outside scope/enforcement of tools, undermining effectiveness; can be domestic (to providers outside initial scope) or cross-border (to foreign providers outside enforcement).
- Strategies to address leakages tend to expand scope to non-bank and foreign providers, e.g. through jurisdictional reciprocity agreements (BCBS 2010) or greater host control over foreign affiliates (CGFS 2014, IMF 2014).
- Capital-based tools:
  - Broad-based capital tools may be subject to domestic leakages (Annex 2).
  - Can lead to increased credit provision by non-bank companies, including bank-affiliated ones (e.g. bank-affiliated leasing companies in Croatia) if consolidated supervision (BCBS 2012) is ineffective, or by finance companies.
  - Cross-border leakages are challenging, especially when foreign affiliates are branches rather than subsidiaries, and when local corporates borrow directly abroad.
  - Additional measures targeting corporate borrowers or fiscal recommendations may be considered (IMF 2014, FSB 2015).
- Household sector loan restrictions:
  - Leakages for LTV and DSTI limits can be more easily contained because such restrictions can, in principle, be enforced on all regulated financial institutions, including non-banks and foreign branches (BoE 2011).
  - Examples show LTV and DSTI constraints can be expanded to non-bank providers within regulatory perimeter, and applied to foreign branches (examples: Estonia, Ireland, Romania, UK).
  - Cross-border provision of mortgage credit in integrated regions has prompted reciprocity of mortgage-related measures (examples: Belgium and the Netherlands).
- Liquidity-related tools:
  - Maturity transformation can migrate to non-banks if liquidity tools target banking system; sizable migration connected to banks can be consolidated onto core system balance sheets (example: China).
  - Liquidity tools can be extended to non-bank providers of maturity transformation, requiring supervisory cooperation (example: US money market mutual funds where liquidity requirements tightened significantly since the crisis).

### Evaluating effectiveness (ex post perspective)
- Ex post evaluation helps gauge need for recalibration or additional measures (CGFS 2012, ESRB 2014, IMF 2014).
- Clarity about the policy target is key for effective calibration and communication; implementation should iterate between action, observation/assessment, and corrective action.
- Ex post evaluation should assess:
  - (i) extent to which measures improved resilience; and
  - (ii) effects on behaviour, credit dynamics, and asset prices.
- Assessing effects on resilience:
  - Define one or more measures of resilience and assess whether resilience improved after imposition (event study approach).
  - Feasible to assess whether LTV and LTI constraints changed distribution of actual LTV/LTI ratios for new and existing borrowers and reduced probability of default and loss given default (examples: Euro area, Hong Kong SAR).
  - Repeated stress testing is another way to assess resilience; increasingly conducted systematically by central banks and supervisors.
- Assessing effects on behaviour and credit dynamics:
  - Useful to assess impact on indicators that prompted intervention and whether desired market responses occurred.
  - Difficult to disentangle policy-induced changes from other forces.
  - Constructing counterfactual paths based on historical relationships in absence of intervention can help (examples: Canada, Korea, New Zealand).

### Considering the potential for tools to be relaxed
- Ex ante consideration of conditions for relaxing macroprudential tools is integral to effective implementation (CGFS 2012, IMF 2013, IMF 2014).
- Some tools can be designed for automatic relaxation (e.g. dynamic provisions), but policymakers typically must decide on appropriateness, timing and speed of relaxation.
- Guiding objective: prevent disruptions to credit provision that adversely affect the real economy.
- When systemic risks recede, gradual relaxation of some constraints may be appropriate; buffers may also be relaxed when risks materialise during financial stress to serve macroprudential objectives.
- Decisions to relax must maintain confidence and appropriate resilience against future shocks; experience with relaxation is limited and effectiveness in times of stress remains uncertain (CGFS 2012, IMF 2014).
- Indicators for relaxation differ from activation/tightening indicators:
  - Slow-moving stock variables/ratios (e.g. credit-to-GDP gap) useful for detecting build-up of risks.
  - Market-based indicators and flow variables (e.g. credit growth, changes in default rates) better capture turning points and predict imminent materialisation of systemic risk (Drehmann and others 2010, IMF 2013).
  - Useful indicator sets differ across sources of stress and tools; e.g. interbank market stress indicators useful for relaxing liquidity tools (CGFS 2012, IMF 2014).

### Improving the information base for macroprudential policy
- Effectiveness benefits from sustained effort to close information gaps (FSB/IMF/BIS 2011, IMF 2013).
- Data gaps hinder risk assessment, tool calibration, and ex post evaluation; data needs differ across tightening and release phases.
- Micro-data on distribution of actual LTV and DSTI ratios across existing borrowers helps calibrate constraint impacts (examples: Ireland, Singapore, Sweden, UK) and is usefully collected prior to imposition.
- Data on bilateral exposures between financial institutions and on systemic vulnerability of nonbanks are often lacking.
- Adequate information sharing arrangements and data collection powers are important to address gaps (IMF 2013).
- Data collection enhancements:
  - Leverage existing sources and consider new investment; conduct sound cost-benefit analysis for data collection enhancement.
  - Fill gaps via improved sharing of supervisory/statistical data, require expanded supervisory reporting, and invest in infrastructure (e.g. credit registers, new statistical surveys).
- International initiatives supporting data closure:
  - FSB-IMF-BIS Data Gaps Initiative and FSB work on monitoring shadow banking risks (FSB 2015d and FSB 2016).
  - Data on bilateral exposures and funding between G-SIBs are collected and shared among home supervisors.
  - Work started on a common data template for global systemically important non-bank financial institutions, beginning with insurance companies (FSB/IMF 2015).

*IMF, FSB, BIS — 3.3 Operationalising the use of tools*

### 4.      International consistency of macroprudential policy

### International consistency of macroprudential policy

### Cross-border effects of macroprudential policy
- Macroprudential policy in financially integrated economies is subject to potential cross-border effects, including positive externalities, leakages that undermine domestic action, undesirable spillovers to other countries, and migration of activities due to uneven policy strength (IMF 2013, Caruana 2016).
- Positive externalities:
  - Effective domestic macroprudential policy that reduces the probability of a financial crisis in one country can lower negative trade and financial spillovers at the regional or international level (IMF 2013, Caruana 2016).
  - International arrangements can buttress national macroprudential frameworks and strengthen “enlightened self-interest,” including IMF surveillance and Financial Sector Assessment Program (FSAP) assessments, FSB peer reviews, and BIS meetings of senior central bank officials.
- Cross-border coordination is recognized as useful in international and regional mechanisms, as exemplified by reciprocity agreed for the Basel III countercyclical capital buffer (BCBS 2010).

### Leakage and reciprocity in addressing cross-border borrowing
- National policies to contain risks from rapid domestic credit build-up can be subject to leakage from increases in cross-border borrowing; leakage effects are empirically documented (see Annex 2).
- One approach is agreement on ‘reciprocity’ so that the same constraint applies to all relevant credit exposures to borrowers in a given country, whether provided by domestic or foreign entities.
  - Example: Basel III agreement on the CCyB includes this approach; EU authorities developed a similar voluntary approach aimed at all measures targeting exposures (ESRB 2015a).
  - Guidance exists for EU countries on treating exposures into third countries that fail to take macroprudential measures (ESRB 2015b).
- Empirical patterns:
  - Cross-border leakages appear strong for capital requirements, especially where affiliates are established as branches, but weaker for loan restriction tools such as LTV and DSTI (Reinhardt and Sowerbutts 2015).
  - Macroprudential tools appear less effective in reducing credit growth in economies with more developed or open financial systems (Cerutti and others 2015).

### Spillovers from domestic macroprudential actions and timing considerations
- Macroprudential tightening in one country can reduce that country’s banks’ cross-border lending, creating undesirable spillovers for other countries (IMF 2014).
  - Retrenchment is more likely when measures are ill-timed and occur in periods of financial stress, when groups find it harder to raise additional capital or liquidity.
  - Timely use of macroprudential tools is important from an international perspective (IMF 2014).
- Design and calibration of tools may be modified to minimize adverse effects on other jurisdictions while achieving domestic objectives (IMF 2014).
- Expansionary effects can occur when global or regional banking groups respond to domestic tightening (e.g., LTV caps) by increasing lending abroad; such effects can be beneficial or may call for counterbalancing macroprudential action in recipient economies.

### Migration of activities and minimum standards to limit regulatory arbitrage
- Policies strengthening financial institutions’ resilience in one country can induce migration of activities to other countries, risking under-implementation of structural measures (e.g., SIFI capital surcharges) and concentration of risky activities in lightly regulated jurisdictions (Viñals and Nier 2014).
- Responses to migration effects:
  - Minimum standards and supplementary agreements and guidance (examples: BCBS framework for domestic systemically important banks (BCBS 2015); minimum internal TLAC requirements for each resolution entity within each G-SIB (FSB 2015)).
  - Ongoing work to ensure resilience and resolvability of systemically important central counterparties (FSB/BCBS/CPMI/IOSCO 2015).
  - International monitoring of implementation is a key complement to standards and guidance.

### Use and trends in macroprudential tool deployment (dataset coverage)
- The underlying database covers 64 countries, of which 32 are advanced economies according to IMF (World Economic Outlook) classification.
- “Usage” in the dataset counts the number of countries using the various instruments that comprise each group and assumes that once a country introduces an instrument, it continues using it.
- Tightening and loosening episodes in the database are coded as +1 and -1 respectively; cumulative sums are shown over time for groups of instruments.

### Institutional models for macroprudential policymaking (Annex 1)
- Institutional arrangements reflect country-specific circumstances; no single model fits all. A rising prevalence assigns the main macroprudential mandate to a well-identified authority, committee, or interagency body, generally with an important role for the central bank.
- Illustrative typology of models:
  - Model 1: Main mandate assigned to the central bank, with Board or Governor making macroprudential decisions (examples: Czech Republic, Ireland, New Zealand, Singapore).
    - Where supervisory authorities sit outside the central bank, coordination can be achieved via a committee chaired by the central bank (Estonia, Portugal), information sharing agreements, or explicit recommendation powers (Norway, Switzerland).
  - Model 2: Main mandate assigned to a dedicated committee within the central bank (examples: Malaysia, UK).
    - Enables separate monetary and macroprudential decision structures under the central bank roof, and allows external experts and separate supervisory authorities to participate in decision-making.
  - Model 3: Main mandate assigned to an interagency committee outside the central bank, with the central bank participating (examples: France, Germany, Mexico, US).
    - Can accommodate a stronger role for the Ministry of Finance, aiding political legitimacy and broader policy consideration.
- Variants and other setups exist (e.g., more limited central bank role in Sweden; distributed mandates in Canada; prudential-authority-led macroprudential responsibility in Australia, Japan).
- Table of illustrative country examples highlights jurisdictions choosing each model and notes where additional councils or ministerial chairs apply.

### Effectiveness of macroprudential tools (Annex 2)
- Evidence base is expanding but not yet conclusive on which instruments work best in which circumstances (Akinci and Ohmstead-Rumsey 2015; Claessens 2015; Cerutti and others 2015).
- Capital-based tools:
  - Support resilience and credit supply during downturns; limited effects in upswing.
  - SIFI surcharges enhance resilience (BCBS 2010, BIS 2015).
  - Some studies suggest short-term dampening on credit growth; longer-run impact appears limited (Dagher and others 2016).
  - Cross-country analysis suggests these tools help contain credit contraction as better-capitalised banks can continue lending during downturns (IMF 2013a); Buchholz 2015 finds faster post-crisis credit growth in countries with caps on banks’ leverage.
  - Dynamic provisioning can help smooth post-crisis credit supply (Jiménez and others 2012).
- Sectoral capital requirements:
  - Increase resilience via additional buffers; evidence on effects on credit growth varies.
  - Some studies find sectoral requirements or risk weights limit loan growth of targeted sectors (BoE 2014; IMF 2013).
  - Leakage may prevent sectoral tools from containing credit booms (Crowe and others 2013).
  - Sectoral CCyBs have limited effects on loan growth but can shift loan supply toward better-capitalised institutions (Basten and Koch 2015).
- Borrower-based tools:
  - Limits on LTV and DSTI ratios enhance borrower resilience (Hallissey and others 2014) and moderate lending growth (Igan and Kang 2011; Akinci and Ohmstead-Rumsey 2015).
  - Effects on house price growth appear limited (Kuttner and Shim 2013).
- Liquidity-related tools:
  - Changes in reserve requirements may help moderate credit growth (IMF 2013a; Lim and others 2011), though some studies find no or weakly significant impacts (Kuttner and Shim 2013; Bruno and others 2015).
  - Experience with other liquidity-related tools remains limited.
- Evidence of leakage:
  - Studies document leakage to non-bank providers of credit (Cizel and others 2016).
  - Cross-border leakages particularly affect capital requirements where affiliates operate as branches; loan restriction tools show weaker cross-border leakage (Reinhardt and Sowerbutts 2015).
  - Evidence on cross-border spillover effects of macroprudential policies is summarized in Buch and Goldberg 2016.

---


_Source: https://www.bookstore.imf.org/images/IMF_FSB_BIS_2016.pdf_