Abstract
The European Union’s (EU) financial stability framework is being markedly strengthened. This is taking place on the heels of a severe financial crisis owing to weaknesses in the banking system interrelated with sovereign difficulties in the euro area periphery. Important progress has been made in designing an institutional framework to secure microeconomic and macroprudential supervision at the EU level, but this new set-up faces a number of challenges. Developments regarding the financial stability may assist in the continuing evolution of the European financial stability architecture.
III. The European Systemic Risk Board: Effectiveness of Macroprudential Oversight in Europe11
A. introduction
34. The role of macroprudential policy is to identify risks to systemic stability and to develop and implement a policy agenda so as to mitigate risks and build buffers to cushion the impact when crises occur. To be effective, macroprudential policy rules need to be complemented by an element of discretion that takes into account all information and enables policy to respond flexibly to developments in the financial system. Because of this inherent need for judgment, the institutional framework that underpins decision-making and establishes powers to use macroprudential instruments plays a central role.
35. In the EU context, the complexity of macroprudential oversight of the financial system is exacerbated by strong financial integration, including cross-border banking activities and integrated financial markets. A supranational overlay is thus required to ensure that policy action can take account of the scope for cross-country externalities and interconnectedness. Coordination of policies at the EU level is needed also to reduce the possibility of regulatory arbitrage of national actions that is likely otherwise when capital and financial services can flow freely across national borders.
36. The establishment of the ESRB in January 2011 is a crucial step in providing greater traction for macroprudential oversight at the EU level, as envisaged in the 2009 De Larosière report. However, the ESRB’s legal foundations imply that it has to establish its credibility under a set of institutional constraints—a complex decision making process (with 27 Member States and EU/euro area institutions being involved), no direct access to supervisory data, and no binding powers.12 These constraints put a premium on close collaboration with other EU bodies, such as the ECB, the new European Supervisory Authorities (ESAs) and the European Commission.
37. Perhaps less anticipated by the De Larosière report, the effectiveness of the ESRB also requires a sufficient level of preparedness of national macroprudential authorities and collaboration between the national and the EU level. Effective macroprudential oversight across the EU is likely to require a “bottom-up” element to complete analysis and decision-making on the part of the ESRB, but will also require that “top-down” recommendations of the ESRB are implemented. This national element, conditioned by national information and mandates, will be important for identification of risks and the development of national initiatives to mitigate risk. Strong national frameworks are needed also to ensure effective follow-up to risk warnings and recommendations of the ESRB. However, since macroprudential policy is likely subject to biases for inaction, as we shall discuss, ESRB risk warnings and recommendations are needed to strengthen the national resolve to take action. In this regard it is desirable for national mandates to contain an EU dimension, including a requirement to comply with ESRB recommendations.
38. In addition to strong mandates and decision making powers, effective follow-up on ESRB risk warnings and recommendations will also require that an adequate and common macroprudential policy toolkit is in place in all Member States. The macroprudential toolkit of EU countries should go beyond the small set of instruments on which there is already an international consensus, and should be flexible enough to address a range of sources and channels of transmission of systemic risks.
39. The ESRB should move forcefully in developing the EU macroprudential policy toolkit and recommend legislative action on the part of the EU Commission to ensure that a common set of macroprudential instruments is put in place across EU Member States. Once macroprudential policy is operational across the EU, a key role for the ESRB will be to provide an EU-wide perspective on risks arising from interconnectedness and financial imbalances across the EU. The ESRB will also have a key role to play in sanctioning and coordinating the use of macroprudential tools across the EU, so as to ensure appropriate reciprocity of macroprudential policy and to minimize cross-border arbitrage of national macroprudential action. The ESRB could finally provide a seal of approval when countries decide to use or to modify macroprudential instruments that interrelate with standards established across the EU.
40. The paper is organized as follows. Section B sums up the constraints under which the ESRB operates. Section C explains why national macroprudential frameworks will play an important role in completing risk analysis and ensuring follow-up on ESRB recommendations, and assesses recent changes in institutional models at the national level. Section D explores which tools should be established at national level so as to enable an effective and coordinated response to the build-up of systemic risks across the EU. In Section E, the paper argues for a strong role for the ESRB in establishing the EU macroprudential toolkit and in harmonizing use of macroprudential policies in a manner that ensures reciprocity across the EU. Section F concludes.
B. Institutional Constraints
41. The ESRB was established under a set of constraints that may hamper its effectiveness, as argued by Fund staff in the report of the European Financial Stability Framework Exercise (EFFE).13 The current set-up may be the outcome of a political economy process combined with a desire to maintain a balance of power among EU institutions.
42. A first constraint is that the regulation does not envisage direct or automatic access to supervisory data for the ESRB.14 Difficulties to obtain such data may hamper assessment of systemic risk. This is a strong constraint in particular when there is a need to assess correlated exposures to particular sectors, the web of interconnections between individual financial institutions and their evolution over time.
43. The ESRB will be subject to complex decision-making processes. This second constraint is a consequence of the supranational dimension of the ESRB. Its General Board has a large membership covering all EU countries in addition to the ECB and of other EU institutions (the EU Commission, the ESAs). The sheer size of the board may slow decision-making and will require strong leadership on the part of its Steering Committee in setting the agenda. Another factor that may mitigate the strength of this constraint is that the General Board can make decisions by majority vote.15
44. The ESRB has no binding powers. It has no direct powers, e.g., in the calibration or setting of macroprudential instruments, nor the power to issue binding recommendations to any other EU or national body. Indeed, the fact that the ESRB has no legal personality under the treaty may rule out a material strengthening of its powers beyond those envisaged in the establishing regulation. Its influence will instead need to rely on its communication strategy, through formal risk warnings and non-binding recommendations as well as informal suasion, and on the quality of its analysis.
45. The ESRB has limited resources. This limited resource is complemented by the support provided by the ECB, and that provided by national authorities through their contributions to the work of the ESRB in its various working groups.
C. The Need for Effective National Macroprudential Policy Frameworks
46. Because of the high degree of financial interconnectedness across the EU, the ESRB needs to play a leading role in macroprudential oversight for the region as a whole. However, effective macroprudential policy across the EU also requires effective policy frameworks at the national level. First, local supervisory information and analytical expertise are required to analyze risks and policy options in a manner that takes account of local conditions. Second, the national framework for decision-making will determine the effectiveness of the interplay between the national and the EU level. Macroprudential policy at the national level will need to be able to respond to ESRB risk warnings and recommendations, requiring that adequate mandates and powers are established at the national level. Moreover, effectiveness of macroprudential policy across the EU will require that national authorities have the resolve and the incentive to take action, including in response to risk warnings of the ESRB. Finally, strong national mandates and institutions will enhance cooperation with EU institutions in risk analysis and assessment, enabling the ESRB to tap analytical expertise that already exists at the national level. In sum, while strong national frameworks are not sufficient to ensure effective macroprudential policy across the EU, they are a necessary complement to the existing arrangements at the EU level.
47. Strong mandates are needed at the national level that open up and at the same time constrain discretionary use of powers. The mandate needs to establish safeguarding systemic stability as the primary objective, but should also include secondary objectives to ensure the policymakers consider trade-offs when macroprudential action has costs as well as benefits. For example, a secondary objective could be to ensure that macroprudential action does not unduly impair the capacity of the system to contribute to balanced growth. The strength of powers needs to be commensurate with the public policy interest in mitigating systemic risk.16 These powers should include (i) information collection powers; (ii) rulemaking and calibration powers; and (iii) powers to determine the appropriate perimeter of macroprudential action. Mandates and powers should be flanked by strong accountability mechanism that ensure the policymaker is required to explain its action or inaction and can be challenged by the public and elected bodies. 17
48. A key challenge in establishing effective macroprudential policy frameworks both at the national and at the EU level is to ensure ability and willingness to act. The benefits of taking corrective action—reduction in the probability and impact of a financial crisis—are uncertain and long-term, while the costs of taking action, including their impact on the financial industry and the provision of financial services to the economy, are often felt immediately and are highly visible.18 The gains of macroprudential policies are uncertain and difficult to assess because, in contrast to other policies—such as stabilization policies that aim at affecting a macroeconomic aggregate—macroprudential policies aim at reducing the probability of a tail event—an objective by definition difficult to measure. The costs of macroprudential policies, such as a tightening of prudential requirements, are in contrast more easily quantifiable, because they will affect the financial industry and other sectors immediately and more visibly. This, in general, may create a bias in favor of inaction or insufficiently timely and forceful action.19 20
49. National authorities may have a stronger bias towards inaction than supranational authorities. Because of the high level of financial integration and interconnectedness in the EU, financial crises originating in individual countries will inevitably affect other EU countries directly or indirectly. National authorities will not internalize such externalities unless their mandate contains an EU dimension. Another reason why national authorities could be prone to inaction is because of lobbying and other political economy pressures at the national level. Such political economy pressures can be particularly strong where supervision is prone to be influenced by an explicit or implicit objective to protect the competitiveness of “national champions”. To minimize biases for inaction, national frameworks and mandates should therefore be such that they contain an EU dimension and reinforce the effectiveness of the EU institutions, for example by including a requirement to comply with ESRB recommendations.
50. In sum, national and supranational mandates need to complement each other. Because the ESRB has no binding power, nor the capacity to analyze policy-trade-offs that takes full account of national conditions, effective macroprudential policy requires effective arrangements at the national level that increase ability and willingness to act on the part of the macroprudential policymaker. However, since political economy pressures and cross-country externalities may engender greater resistance to act at the national level, the EU dimension is essential. National mandates need to be complemented by guidance provided by the ESRB and should include a requirement to act on ESRB recommendations.
Institutional models in the EU: changes and status quo
51. Since the crisis, there has been substantial institutional change in a number of countries across the EU (Box 1). In addition, in a number of other member states change is ongoing or planned. A range of institutional models are being established and differ along four dimensions: (i) the degree of institutional integration between agencies, (ii) the existence of a separate committee or council, (iii) the role of the central bank or the treasury as the leading agency (chair) of the committee, and (iv) whether the committee has strong decision-making powers, including the power to direct the actions of constituent agencies.
52. There is a strong and welcome trend towards greater institutional integration between central banks and regulatory agencies. The move to more strongly integrated models is likely to enhance effectiveness by increasing ability and willingness to act, fostering the flow of information and enabling the establishment of strong powers on the part of decision-making committees.21 Belgium, Ireland, and the U.K. are key examples to date of countries that have reversed the strong separation between central banks and regulatory agencies that was introduced from the late 1990s in favor of integration between central bank and prudential agency (in the U.K.) or between central bank and the integrated financial regulator (in Belgium and Ireland). The U.K. and Romania are examples of countries were a new macroprudential committee has been or is being set up that is chaired by the central bank and has strong decision-making powers.
53. In a number of other cases, however, the new macroprudential council is instead chaired by the treasury, rather than the central bank. Examples include France, Greece, Italy, and Portugal. A risk of a strong role of the Treasury is that this may reduce independence from the political process and willingness to act. A strong role of the treasury tends to be a feature also and perhaps in particular for countries that are characterized by separation between central bank and prudential agency, as prevailing in Denmark, Hungary, Poland and Austria. In these countries, moreover, the role of the council tends to be coordination, often relying on consensus between agencies, as opposed to a decision-making committee that would have strong independent powers or powers to direct its members.
54. Finally, for a large number of countries a high degree of separation between central bank and regulatory agencies persists. This could become an obstacle to effective macroprudential policy especially if a council charged with macroprudential policy coordination remains absent from the regulatory framework. Examples here include Estonia, Finland, Germany and Luxembourg in the euro area, Latvia, and Sweden in the larger EU, as well as Iceland and Norway outside of the EU, but inside the European Economic Area.
55. Overall, recent changes in institutional frameworks are likely to increase effectiveness of macroprudential frameworks, complementing the introduction of the ESRB at EU level. However, institutional changes appear far from complete in a number of countries and have also tended to reduce independence from the political process in some cases. A strong role of central banks is a key strength of the institutional set-up at the level of the ESRB and should be a guiding principle also for the review of national frameworks going forward.22
Institutional Models
Model 1—Full integration. This model is characterized by a full institutional integration between the central bank, the prudential regulator and the securities markets regulator. Full institutional integration allows for decision-making along the full range of potential macroprudential tools to be made by a committee that is internal to the organization, such as the central bank’s executive board and for such decisions to be fully binding on the respective department. As set out in IMF (2011b), it may be useful for decisions on monetary policy to be made by a separate committee. In Europe, the Czech Republic has long been the only example of the fully integrated model. More recently Ireland moved to a fully integrated model and abandoned the strong separation between central bank and integrated financial regulator that characterized its former model. Belgium is planning to move to full integration between central bank and integrated regulator in 2011, in the process dismantling its macroprudential coordinating council.
Model 2—Partial integration. This model is characterized by strong institutional integration between the central bank and the prudential regulator, with the latter a department or subsidiary of the central bank, and a securities markets authority that is separate from the central bank. This model is widespread in Europe, but there are differences in the way macroprudential decisions are taken. Under one variant of the model (2a), the Governor of the central bank chairs a committee that brings together central bank officials, including the head of prudential regulation and supervision, and the head of the separate securities markets authority. Under a second variant (2b), macroprudential policy is conducted by a committee that again brings together the central bank and the separate securities market regulator, but the committee is chaired by the Treasury. Under a third variant (2c) coordination between the central bank and its prudential department and the separate securities market authority is more informal and there is no dedicated macroprudential committee that spans all three agencies. The U.K. is moving to a model where the macroprudential committee is chaired by the central bank (2a) as is already the case in Romania. Change in the U.K. will also reintroduce close institutional integration between central bank and prudential agency, by the setting up of a new prudential agency as a subsidiary of the central bank. In a number of other countries, including France, Greece Italy, and Portugal the starting point has been integration between central bank and the prudential agency, and this is now being complemented by new macroprudential councils that bind in the securities regulators, but are chaired by the Treasury (2b). A number of other countries, such as the Netherlands have not (yet) introduced a dedicated macroprudential committee or council (2c), and rely instead on strong institutional integration between central bank and prudential regulator as a coordinating device.
Model 3—Separation. In this model there is separation between the central bank and both the prudential regulator and the securities market regulator, with the latter two sometimes integrated to form one integrated regulator. Macroprudential policy coordination that involves the central bank then has to rely on inter-agency coordination that can be facilitated by a council. However, under separation it tends to be difficult to square strong powers to direct with the operational autonomy of each participating institution. Separation therefore favors decision-making by consensus and while a council may issue recommendations, such recommendations will under separation tend to be non-binding on constituent agencies. In practice, in addition, separation may favor a strong role of the treasury, so as to mediate between differences of view between constituent agencies. Consistent with this, In Europe there is no example to date of the separated model in which the macroprudential council is chaired by the central bank (3a). Hungary and Poland have recently set up a council that is chaired by the treasury (3b), as is the case also in Denmark and Austria. Finally, in a number of countries, there is no formal macroprudential council and coordination between central bank and regulatory agencies is more informal (3c). Examples here include Estonia, Finland, Germany, and Luxembourg in the euro area, Latvia, and Sweden in the larger EU, as well as Iceland and Norway outside of the EU, but inside the European Economic Area.
Lack of powers at national level
56. Lack of powers to use macroprudential instruments at national level could become a further important obstacle to effective macroprudential policy in the EU. The need to constrain a dynamically evolving financial system requires powers to act, including the power to choose the appropriate tool in any given conjuncture and to calibrate those tools to the prevailing level of systemic risk. National authorities need also to have the ability to respond swiftly to ESRB recommendations, by employing appropriate macroprudential instruments to reduce the build-up of vulnerabilities.
57. There is considerable heterogeneity across the EU as regards the independent ability of regulatory agencies to introduce new regulations or to recalibrate existing regulations without a change in law. In a number of countries, such as Sweden and Romania the regulatory agency has strong independent powers to set financial regulations, without such regulatory acts requiring the consent of the treasury or parliament. In a number of other countries the rulemaking power of the regulatory agency is more constrained, often requiring consultation with or approval of the treasury, as in Finland, Germany, and the Netherlands, or in some cases even a parliamentary act.
58. This underscores the need for regulatory authorities across the EU to be empowered to use and calibrate macroprudential tools in a manner that is independent of the political process. Such powers on the part of the regulatory authorities need to be guided by a mandate that opens up as well as constrains the discretionary use of macroprudential tools and should be flanked by appropriate accountability mechanisms.23
D. The Macroprudential Policy Toolkit
59. In this section of the paper, we examine which macroprudential tools should be established at the national level. In the context of the EU this issue has two important dimensions. First, tools that are established at national level should be effective and comprehensive, enabling the authorities to respond effectively to the build-up of systemic risk. Second, it is desirable that there are common sets of tools, shared by all member states, so that mitigation of systemic risk can be coordinated through the ESRB and regulatory arbitrage avoided.
60. We first briefly outline principles that should guide the choice of the EU macroprudential policy toolkit. Second, we summarize the findings of the IMF survey of macroprudential policies performed by the Monetary and Capital Market department of the IMF. We argue that the macroprudential toolkits considered by national authorities according to this survey are unlikely to be sufficient to address a range of possible systemic risks. We next suggest additional instruments that should be included in the macroprudential toolkit.
Choice and use of macroprudential instruments: normative considerations
61. Most instruments that have been used so far are prudential instruments recalibrated for macroprudential purposes. Most experiences with macroprudential instruments in the EU were in Central and Eastern European countries, where instruments included enhanced capital requirements and increases in reserve requirements. Examples of use of macroprudential instruments in more mature economies are the introduction of dynamic provisions in Spain in the early 2000, and limits on LTV more recently introduced in Sweden.
62. Macroprudential instruments are needed to address two distinct dimensions of systemic risk. They need to mitigate systemic risks arising from the tendency of the system to become overexposed to aggregate or correlated risks over time (such as aggregate or sectoral credit imbalances, asset price bubbles and balance sheet mismatches) and those arising from the systemic impact of the failure of individual institutions (such as amplification and contagion effects arising because of interconnectedness, fire sales or confidence effects).24
63. Experience has shown that financial crisis have various sources and channels of transmission or amplification. The current crisis in the euro area is the consequence of real estate bubbles and sovereign risks; the 2008 financial crisis in the United States was exacerbated by weak underwriting standards at origination in the U.S. real estate market and of failures in the securitization process; the Asian crisis of the 1990s was mainly caused by balance sheet maturity mismatches in the corporate sectors, with high reliance on short-term debt supplied by merchant banks to finance long-term investments; emerging markets (Korea) also experienced episodes of financial stress resulting from the rapid expansion of consumer credit through banks and non-bank financial intermediaries.
64. Macroprudential toolkits should include a carefully selected set of instruments sufficient to address most foreseeable sources of systemic risks. The often cited principle according to which instruments should be linked to objectives does not imply that macroprudential policy should rely on only one instrument. Because systemic risk has multiple dimensions and can arise through various institutions, markets, and sectors, the macroprudential policy toolkit must contain sufficient instruments to mitigate risks through each likely channels of systemic risk build-up. By contrast, a too restrictive set of instruments may not allow the policymaker to target new channels of build-up of systemic risks.
65. Another argument in favor of the establishment of a range of instruments is that there is, so far, very limited knowledge about which instruments are effective, and which ones are not. Moreover, experience has shown that to be effective a range of complementary instruments may need to be brought to use when the build-up of systemic risk is fuelled by strong underlying forces that cause each individual instrument to be subject to arbitrage (see for instance Crowe and others, 2011). This also requires that instruments should be broad in scope, rather than applied only to particular institutions, which could result in arbitrage or miss addressing the proper channels of systemic risks.
66. Individual macroprudential instruments belonging to the toolkit should be chosen based on several key characteristics. First and foremost, it is essential that instruments are effective in limiting the build-up of systemic risk and help increase resilience while minimizing dead-weight costs on the industry or to the provision of credit to the economy. Another key dimension in the EU is that macroprudential instruments should be designed and applied to limit opportunities for regulatory arbitrage. To protect the fabric of the single market, reciprocity of macroprudential policies will need to play a crucial role. Finally, clarity and simplicity in the use and calibration of instrument may be needed to ensure that clear signals are sent to market participants.
Developing the EU macroprudential toolkit
67. The survey conducted at the end of 2010 by the Monetary and Capital Market Department of the IMF asked each Member State their views on the set of instruments that should belong to their macroprudential policy toolkits. A list of 30 instruments was suggested in the questionnaire, but the questionnaire also allowed countries to add other instruments.
68. The instruments that were mentioned in the questionnaire can be classified in the following categories: (i) instruments targeting the asset side of banks; (ii) instruments targeting asset-liability mismatches; (iii) capital measures; (iv) measures targeting SIFIs; (v) measures focused on the funding of banks; (vi) measures affecting the functioning of markets; and (vii) other measures (such as tax policy).
69. Respondents favor a small set of measures on which there is already an emerging international consensus. There is a clear consensus on capital measures, such as the countercyclical capital buffer agreed upon in Basel III, dynamic provisioning, and capital conservation measures also included in Basel III (such as restrictions on the distribution of profits). A large majority of countries favor imposing limits on loan-to-value ratios and loan-to-income ratios. A capital surcharge for SIFIs and limits on maturity mismatches also emerges as measures that would be considered by a majority of countries.
Macroprudential Instruments Selected by European Countries (Part I)
Macroprudential Instruments Selected by European Countries (Part I)
| Assets measures | Asset/liability measures | Capital measures | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Country | Caps on loan-to-value ratios | Caps on debt/loan-to-income ratios | Limits on exposures or concentration | Caps on foreign currency lending | Ceiling on credit or credit growth, incl by sector | Limits on net open currency positions/currency mismatch | Limits on maturity mismatch | Reserve requirement | Countercyclical capital requirement | Dynamic provisioning | Restrictions on profit distribution |
| euro area | 8 | 6 | 5 | 1 | 2 | 2 | 7 | 2 | 7 | 7 | 7 |
| Other EU | 5 | 2 | 2 | 2 | 1 | 4 | 2 | 3 | 3 | 3 | |
| EEA | 1 | 1 | 1 | 1 | 1 | 1 | |||||
| All | 14 | 9 | 7 | 3 | 3 | 6 | 8 | 4 | 11 | 11 | 11 |
Macroprudential Instruments Selected by European Countries (Part I)
| Assets measures | Asset/liability measures | Capital measures | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Country | Caps on loan-to-value ratios | Caps on debt/loan-to-income ratios | Limits on exposures or concentration | Caps on foreign currency lending | Ceiling on credit or credit growth, incl by sector | Limits on net open currency positions/currency mismatch | Limits on maturity mismatch | Reserve requirement | Countercyclical capital requirement | Dynamic provisioning | Restrictions on profit distribution |
| euro area | 8 | 6 | 5 | 1 | 2 | 2 | 7 | 2 | 7 | 7 | 7 |
| Other EU | 5 | 2 | 2 | 2 | 1 | 4 | 2 | 3 | 3 | 3 | |
| EEA | 1 | 1 | 1 | 1 | 1 | 1 | |||||
| All | 14 | 9 | 7 | 3 | 3 | 6 | 8 | 4 | 11 | 11 | 11 |
70. There is, surprisingly, more limited support for macroprudential measures addressing other dimensions of macroprudential risks. Few respondents mentioned measures targeting the funding structure of financial intermediaries, in particular the reliance on wholesale funding—such as a levy on wholesale funding, a (time varying) liquidity coverage ratio, or liquidity surcharges for systemically important institutions.25 Measures targeted at the functioning of securities market, such as limits on haircuts or on collateral margins, were also only selected by a few countries. Other types of measures (addressing non-real estate sectors, or the functioning of the interbank market, or fiscal measures) appear to have less or almost no support among the group of respondents.
Macroprudential Instruments Selected by European Countries (Part II)
Macroprudential Instruments Selected by European Countries (Part II)
| SIFIs | Funding | Markets | Fiscal | |||||
|---|---|---|---|---|---|---|---|---|
| Country | Size dependent leverage limit | Capital surcharge | core funding ratio | levy on wholesale funding | time varying liquidity coverage ratio | haircut or margin on collateral | CCP for derivatives | Sector specific taxes |
| euro area | 1 | 7 | 4 | 2 | 4 | 1 | 2 | |
| Other EU | 1 | 1 | 2 | 1 | 1 | 2 | 2 | |
| EEA | 1 | 1 | 1 | |||||
| All | 2 | 9 | 6 | 1 | 4 | 6 | 1 | 5 |
Macroprudential Instruments Selected by European Countries (Part II)
| SIFIs | Funding | Markets | Fiscal | |||||
|---|---|---|---|---|---|---|---|---|
| Country | Size dependent leverage limit | Capital surcharge | core funding ratio | levy on wholesale funding | time varying liquidity coverage ratio | haircut or margin on collateral | CCP for derivatives | Sector specific taxes |
| euro area | 1 | 7 | 4 | 2 | 4 | 1 | 2 | |
| Other EU | 1 | 1 | 2 | 1 | 1 | 2 | 2 | |
| EEA | 1 | 1 | 1 | |||||
| All | 2 | 9 | 6 | 1 | 4 | 6 | 1 | 5 |
71. While the choices of respondents fully reflect the growing international consensus, completing this small set with additional instruments would be justified, for two main reasons. First, as discussed earlier, there remains a lot of uncertainty over each instrument’s effectiveness. Hence, enlarging the set of instruments admitted to the macroprudential toolkit would limit the likelihood that macroprudential instruments might turn out to be ineffective when needed. Second, some instrument may be suited only for specific sources or for specific channels of transmission of systemic risk, and are unlikely to be useful in mitigating other sources of systemic risks, as highlighted in the previous section.
72. The most popular instruments so far are not without their limitations.
While broad in scope, the countercyclical capital buffer may suffer from some a number of limitations. First, in the Basel III framework, there will be long lags (up to one year) between the announcement of capital add-ons by national supervisors and its implementation. Second, the deviation of the credit-to-GDP ratio from its trend—used as an indicator to calibrate the buffer—is an imperfect indicator of the macrofinancial cycle. Third, during downturns, the decision to release the buffer by the macroprudential supervisor may be inconsistent with the microprudential principle under which banks should not deplete capital when non-performing assets are building up. Finally, the capital buffer is a blunt tool: when the build-up of imbalances is concentrated in particular sectors this could lead to a crisis well before the buffer is triggered by aggregate developments.26
The second most popular measure, a contingent upper bound on loan-to-value ratios, potentially completed by an upper limit on the debt-to-income ratio to ensure ability to repay, seems to be emerging as an effective instrument; but the experience so far suggests that implementation could be challenging, as there are risks of regulatory arbitrage, through both non-bank and cross-border lending. Moreover, to be effective, coverage must be comprehensive including by covering second-lien mortgages (Crowe and others, 2011). Its benefits in terms of buffers and mitigation of strategic defaults may also differ across countries, according to characteristics of the mortgage market (including, for example, whether mortgages are full recourse or not). Finally, relaxing the LTV limit in a downturn may also create conflicts between the macroprudential and the microprudential perspectives. From a macro perspective, it may be optimal to raise limits on loan-to-value ratios to support the real estate market during a crisis. But such a policy may increase the credit risk of individual financial institutions, and may not be optimal from a micro-prudential perspective.
73. The set of instruments emerging from the international consensus appears too focused on taming broad-based credit booms and real estate bubbles, and not sufficiently aware of other specific channels and sources of risks. The macroprudential toolkit of EU countries should therefore go beyond this set and consider including additional instruments targeting:
a. Time varying exposures to specific sectors. Time and sectoral contingent risk weights would usefully complement the countercyclical capital buffer for two reasons.27 First, they would allow to target more specifically the sectors where systemic risk is developing, thus allowing a cross-sectional differentiation of risks. Second, in contrast to broad-based measures, such as the countercyclical capital buffer, sectoral measures would more easily identify the build-up of sectoral vulnerabilities that may not be well captured by the private credit to GDP ratio. Examples of such measures would include contingent risk weights on interbank lending, lending to sovereigns, corporates, or households.28
b. Funding of financial intermediaries. Balance sheet expansions during the boom were financed by relying on wholesale funding, often of a short-term nature. A possible instrument, in addition to the two quantitative liquidity constraints included in Basel III, would be a contingent levy on non-core funding (as currently introduced in Korea) which would help address both the time dimension and the cross-sectional dimension of systemic risk.
c. Collateralized lending markets. A possible instrument would be contingent margins or valuation haircuts on existing securities used as collateral in the securitized lending markets (such as for repos). This instrument would be used to regulate the supply of secured funding which would help reduce the risks of fire sales. It would also affect the contribution of the shadow banking system to pro-cyclicality by affecting their funding conditions in wholesale markets. Ideally, strong margining should be extended to OTC markets to incentivize the move to central clearing of derivatives.
E. The Role of the ESRB in a Financially Integrated EU
The role of the ESRB in the design and use of macroprudential tools across the EU
74. Because of the integration of financial systems among EU countries, effective national mandates and tools are necessary but not sufficient to ensure effective macroprudential oversight across the EU. They are necessary because without these national elements there can be no effective macroprudential action for the EU as a whole. But a pure national approach is not sufficient in a highly financially integrated region where capital should continue to flow freely across borders. First, decisions to act need to have a EU dimension to overcome cross-country externalities and the risks of regulatory capture at the national level. Second, strong coordination of policy actions at the EU level is needed to avoid regulatory arbitrage by financial institutions that are located outside of the country setting macroprudential policies. The ESRB therefore has a strong role to play in ensuring effective macroprudential oversight for the region as a whole.
75. The ESRB should warn that a minimum set of macroprudential instruments common to all Member States needs to be established EU-wide. This approach would ensure that each Member States is able to act in response to rising systemic risks, and that all Member States can adopt common responses to similar risks. The ESRB advice on the harmonization and use of macroprudential policy toolkits should result in recommendations addressed to the Commission for specific EU Directives, to ensure that macroprudential policy is operational in all Member States.
76. The ESRB should also make recommendations on the calibration of individual macroprudential instruments across all EU countries. This will ensure that the calibration is adequate to the risk identified. In particular, the instrument should be calibrated in a way that mitigates risks effectively—without imposing undue costs on the financial sector—and that all national macroprudential authorities adopt similar quantitative responses to systemic risks. The ESRB could also warn that some practices (such as LTVs at or above 100 percent) are excessively risky and should therefore be prohibited. In some cases the ESRB should give advice as to whether a particular instrument can be implemented by way of a policy rule or whether a discretionary overlay is needed. The ESRB can also issue recommendations on what additional elements need to be considered in the discretionary use of tools. If needed, the ESRB could also issue guidance on the combination of instruments that could more effectively help reduce systemic risks.
77. The ESRB should play a key role in facilitating the effectiveness of macroprudential policy across the EU by ensuring reciprocity across EU countries, so as to reduce the scope for regulatory arbitrage.29 To avoid regulatory arbitrage and ensure effectiveness of macroprudential action across the EU, a mechanism is needed whereby home country authorities reciprocate the macroprudential measures put in place by host countries, based on the exposures of the consolidated national financial institutions to the asset class of the host country considered.30 The ESRB should, if it is satisfied that the macroprudential action taken by the host authority is justified, issue a recommendation to other macroprudential authorities to reciprocate the measures taken by the host authority.
78. The ESRB could finally sanction the decisions of Member States to set or modify macroprudential instruments in particular when these decisions interrelate with EU standards.31 For example, under the forthcoming capital requirements directive (CRD IV), member states will require flexibility in the variation of risk-weights for macroprudential purposes. A potential solution may be for such use of risk-weights to be validated by the ESRB, with the exception then formally granted by the Commission. To minimize the burden on the ESRB, such validation process could be designed to be as rule-based as possible.
The need for the ESRB to work with other EU bodies
79. The ESRB will need to work closely with the European Supervisory Authorities. Successful interaction between the ESRB and the ESAs will be important to ensure a proper meshing of macroprudential and microprudential instruments and risk assessments. Strong cooperation is needed also in the exchange of data and information. As noted in the EFFE report, the current regulations stipulate that requests for detailed data from the ESAs will have to be ad hoc and motivated, and it does not provide a dispute settlement mechanism, which could become problematic in practice
80. The ESRB also depends on the ECB for analytical, statistical, logistical and administrative support. Close collaboration between the two institutions will remain essential in the future. However, since the ESRB is an EU institution covering non-euro area countries, there may be a need to also strengthen its analytical resources in the medium-term, independently of the analytical contribution of the ECB.
81. Finally, the ESRB will also need to work closely with the EU Commission. The ESRB needs to warn when legislative action on the part of the Commission unduly constrains macroprudential policy action. And it should recommend that the Commission takes positive legislative action to ensure that common macroprudential toolkits will be available to policymakers across the EU. The ESRB will also have a role to play in the assessment of macroeconomic imbalances that will be performed by the EU Commission under the External Imbalance Procedure. It should use its recommendations to sharpen financial sector advice given in that process.
F. Conclusion
82. Under the current framework, the effectiveness of the ESRB will strongly depend on successful interaction with other EU institutions and with national authorities. Collaborations with other EU institutions have already been initiated; for example, the ESRB and ESAs are already setting up protocols for information sharing and guidelines and processes for handling data requests should be initiated. The ESRB together with the EBA also actively participates in rulemaking to transpose Basel III countercyclical capital buffer into the EU law. It has expressed clear views on the need for some flexibility in the CRDIV to accommodate macroprudential policy across the EU.
83. National macroprudential oversight arrangements will be important elements of the overall EU framework. They will affect the ability and willingness to make use of national information and analytical capacity as inputs in the ESRB’s analysis. They will also shape the follow-up on ESRB risk warnings and recommendations. But the role of the ESRB will remain crucial, because of the high degree of financial integration within the EU and the resulting scope for cross-country externalities and spillovers. Because the ESRB has no binding powers under the current framework, national mandates should explicitly include a requirement to act on ESRB recommendations.
84. To ensure effective follow-up on ESRB recommendations, EU countries will also have to have common macroprudential instrument toolkits in place. Hence, agreement has to be reached at the EU level on a selective harmonized toolset to be established EU-wide. The toolset should be broad enough to provide the capacity to address various potential sources as well as channels of transmission and amplification of systemic risk. The ESRB should use its mandate to take the lead in harmonizing macroprudential toolkits across EU countries. The ESRB should make recommendations to the EU Commission to use its law making power to introduce a common set of macroprudential instruments, and to ensure that an appropriate degree of flexibility for use of macroprudential tools will be embedded in the CRDIV. To safeguard the single rule book, it should also ensure adequacy and consistency of application, including reciprocal use of instruments across the EU.
References
Borio, Claudio, 2009, “Implementing the Macroprudential Approach to Financial Regulation and Supervision,” Banque de France, Financial Stability Review 13, The Future of Financial Regulation, September (Paris: Banque de France).
Crowe, Christopher W., and others, 2011, “Policies for Macrofinancial Stability: Options to Deal with Real Estate Booms,” Staff Discussion Note No. 11/02 (Washington: International Monetary Fund).
International Monetary Fund, 2011a, European Financial Stability Framework—Preliminary Findings and Recommendations, report prepared by staff of the Monetary and Capital Markets, European and Legal Departments (Washington: International Monetary Fund).
International Monetary Fund, 2011b, Macroprudential Policy: an Organizing Framework, IMF Policy Paper, (Washington: International Monetary Fund).
HM Treasury, 2010, “A New Approach to Financial Regulation: Judgment, Focus and Stability” (London: HM Treasury).
Morris, Stephen and Hyun Song Shin, 2008, “Financial Regulation in a System Context,” Brookings Papers on Economic Activity, Fall (Washington: The Brookings Institution).
Nier, Erlend W., 2009, “Financial Stability Frameworks and the Role of Central Banks—Lessons from the Crisis,” IMF Working Paper No. 09/70 (Washington: International Monetary Fund).
Nier, Erlend W., 2011, “Macroprudential Policy—Taxonomy and Challenges,” National Institute Economic Review, April.
Tucker, Paul, 2011, “Macroprudential Policy: Building Financial Stability Institutions,” Speech given at the 20th Annual Hyman Minsky Conference, New York, April, 14, 2011.
Wong, Eric, and others, 2011, “Loan-to-Value Ratio as a Macroprudential Tool—Hong Kong’s Experience and Cross-Country Evidence,” Hong Kong Monetary Authority Working Paper 01/2011. (Hong Kong: Hong Kong Monetary Authority).
Prepared by Erlend Nier and Thierry Tressel
Regulation (EU) No. 1092/2010 of the European Parliament and of the Council of 24 November 2010 on European Union macro-prudential oversight of the financial system and establishing a European Systemic Risk Board.
European Financial Stability Framework Exercise, Preliminary Findings and Recommendations, May 26, 2011, prepared by staff of the Monetary and Capital Markets, European and Legal Departments of the IMF.
The regulation stipulates that requests from the ESRB for detailed data addressed to the ESAs will have to be ad hoc and motivated, and it does not provide a dispute settlement mechanism.
Decisions within the General Board of the ESRB are generally taken by simple majority, but a majority of two thirds will be needed to adopt recommendations or to make a warning public.
The public policy interest arises because the financial sector provides key services to the real economy.
IMF (2011b) and Nier (2011) offer further discussion of these points. See also Tucker (2011) for an overview of the new arrangements in the U.K.
See IMF (2011b) and Nier (2011).
Another difficulty with macroprudential policy is that the decision by the policy-maker to signal growing systemic risk may trigger market speculation and self-fulfilling expectations of a crisis.
Countries’ crisis experience may also condition their approach to macroprudential policies. For example, countries that experienced a deep financial crisis—such as the U.K.—are more actively developing their macroprudential toolkit than other countries that did not.
See Nier (2009) and Nier (2011)
See IMF (2011b) and Nier (2011) for further discussion of the appropriate roles of the treasury and the central bank in financial regulation.
See IMF (2011b) and Nier (2011) for further discussion.
See for example Borio (2009), Nier (2009), and Nier (2011).
Both levies on wholesale funding and quantitative liquidity ratios act as “automatic stabilizers” mitigating the cross-sectional dimension of systemic risk and potentially reducing the need to readjust these measures in a countercyclical fashion.
The impact on credit supply will depend on the speed at which the buffer is built-up. A fast build-up will presumably be more effective in constraining credit supply, in particular if banks have to resort to costly issuance of equity.
Some of these measures would generalize microprudential exposures limits that are not time varying.
Optimal risk weights may differ from a microprudential perspective than from a macroprudential one. For example, collateralized short-term assets (such as reverse repo transactions) may appear safe from a microprudential perspective, and therefore attract low capital requirements. But they could be systemically important as decisions not to roll-over the transaction may trigger fire sales of assets by the counterparty of the transaction, which may amplify financial crisis (Morris and Shin, 2008).
The experience of Central and Eastern European countries proved that measures taken by host countries can be evaded through cross-border lending, lending through branches or non-bank financial intermediaries that are not within the regulatory perimeter of the host country.
For example, an asset class would be defined as sovereign bonds of country A, or mortgages on properties of country B.
More generally, countries may occasionally have an interest in obtaining a validation of their national macroprudential policies by the ESRB, for example to overcome opposition on the part of the financial industry.