A. The European System of Financial Supervisors (ESFS)

5. The establishment of the ESFS represents a large step forward for the single financial market project, but at the same time forms a complicated network of regulatory and supervisory institutions (see Figure 1). It was set up to facilitate cooperation among EU and national supervisors, secure the exchange of relevant information, and ensure effective decision-making with respect to cross-border financial institutions. The ESFS is designed to be an integrated network of national and EU-wide supervisory authorities leaving the day-to-day supervision to the national level. An advantage of this network is that it engages the existing national supervisory authorities. However, it remains to be seen how this network will operate in practice, and whether all relevant information will be shared among participating authorities.

B. The European Supervisory Authorities

C. European Systemic Risk Board

17. The ESRB, established at the start of 2011, has a broad mandate and is responsible for the macro-prudential oversight of the financial system in the EU. It is to issue risk warnings when risks are deemed significant and, when necessary, provide policy recommendations to mitigate the risks identified.

18. The ESRB is not a separate legal entity. Formally, it is an independent EU body responsible for the macroprudential oversight of the financial system within the EU¹³ The ECB provides analytical, statistical, logistical and administrative support to the ESRB.¹⁴

19. The institutional set-up of the ESRB is relatively complex, and could prove unwieldy. Decisions within the General Board of the ESRB (comprising 37 voting members and 27 non-voting members) will be taken by simple majority, but a majority of two thirds will be needed to adopt recommendations or to make a warning public. The functioning of the ESRB is based on participation involving all key players—national authorities and other EU institutions. This is a pragmatic approach for a new organization. However, there is a risk that such an approach could lead to overly cautious warnings. Moreover, the size of the decision-making group in the ESRB is large, so the timing and the procedures of the decision-making process risk becoming cumbersome. This may be especially costly in time of emerging risks. Moving forward, the General Board should adopt an approach under which puts in place rules that provide broad direction for ESRB activities and ensure accountability but leave ESRB management operational flexibility.

20. The decision-making processes will need to be flexible enough to facilitate the timely identification of emerging risks. Currently, the Steering Committee (a sub-group of the General Board) effectively sets the specific work program. In order to promote the “bottom up” identification of risks, it is suggested that the ESRB staff propose the work program to the Steering Committee and thus the General Board, after consultation with national supervisors, the ESAs, and the Advisory Scientific Committee. The ESRB could also consider how to use the Steering Committee beyond setting the work program in order to be the most effective for decision-making.

21. The ESRB’s enforcement of its recommendations will depend on the effective operation of the “comply or explain” principle. But making warnings and recommendations public may not be a credible threat because doing so could trigger an adverse market reaction. The European Parliament has an explicit role in monitoring actions taken in response to a warning, but such a role can be performed only when recommendations are made public. This limits the potential usefulness of oversight by the Parliament. A possible solution to this concern could be to mandate publication, after a lapse of time, of risk warnings, and recommendations that were not made public, as well as the responses by the addressees, much as central banks do with their market-sensitive decisions. If “comply or explain” does not work well, serious consideration should be given for more direct involvement of the ESRB in ensuring that its recommendations have effect.

22. The credibility of the ESRB will depend crucially on its ability to speak “truth to power,” and to weigh in on policy decisions. To this end, it needs early on to publish clear and well-argued analyses of the main current macro-prudential risks. The key test for the ESRB is whether it will be able to identify major risks, issue risk warnings on these risks, and recommend policy actions that are followed-up.

23. Successful interaction between the ESRB and the ESAs, in particular to ensure a proper coordination of macroprudential and microprudential instruments, will be important. The ESAs are providing microprudential risk assessments, and are working with the ESRB through the Working Groups on systemic risks and data exchanges. In one direction, decisions on temporary bans of financial products (by an ESA) and various bank capitalization ratios, for example, have macroprudential dimensions which necessitate a close coordination with the ESRB. In the other, the ESRB Board envisages that the ESAs will more broadly contribute to systemic risk monitoring through joint submissions of cross-sectoral risk assessments; the ESAs’ contribution would be built into the governance structure through their participation on the ESRB General Board. Eventually and depending on the implementation process, a clarification of the respective mandates may be needed. The ESAs and the ESRB need to establish processes for sharing information and analyses that can feed from, and inform, risk assessments at both the EU and individual college level, for both on and off site examinations.

24. The ESRB is dependent on the ESAs for the provision of microprudential data, which will be crucial for its oversight of the financial system. Under the current regulation the ESRB may only submit a “reasoned” request to the ESAs to receive non-aggregated information on an ad hoc basis. The need for ad hoc and motivated requests may give rise to problems in practice, for example, if it is necessary to compare bank-specific data over time on a systematic basis. The ESAs and the ESRB are working on protocols for information sharing and should be encouraged to develop guidelines and processes for handling such requests.

25. The ESRB is developing the institutional framework, concepts and tools to ensure macro-prudential stability at the EU level. In close cooperation with the ESAs, it will elaborate a color-code system (dashboard) corresponding to situations of different risk levels.

D. Supervisory Colleges

26. One step to achieve effective supervision in the EU of financial groups operating across border has been the establishment of colleges of supervisors. In this setting supervisors competent in the jurisdictions where a financial institution belonging to a certain financial group is established (and now also the respective ESA) meet periodically in order to exchange information about the condition of the group. These colleges will enable supervisors to strengthen their working relationships with their counterparts, and to have a much better understanding of the overall state of the institutions that they are supervising. However, since many of the largest institutions operate in a substantial number of countries, such colleges might be unwieldy: hence “core colleges” have been established, in which participation is limited to those jurisdictions where the institution has its most significant presence. Both full colleges and core colleges are operational for a number of cross border groups.

27. Although a major step forward, these colleges are not a complete answer to the cross border supervision issue:

• There are challenges involved in ensuring that the representation of host supervisors in the colleges is sufficiently broad: in particular, it may be difficult to include in the core college the authority of a host country where a financial institution’s activity may not be very significant from the group’s viewpoint even though such activity may be significant from the host country’s perspective. To mitigate this concern, it should be ensured that all decisions and information from the core college are circulated among other college members;

• College meetings are time consuming, and the burden may fall on a limited number of officials. There is already anecdotal evidence of supervisory agencies declining to send representatives to meetings of colleges;

• The colleges provide a forum for discussion, but may not lead to prompt action, although, in the case of the EBA, banking colleges are required to reach joint decisions in relation to the solvency of a banking group;

• It is not always clear that the incentives are such as to maximize information sharing. While such sharing is in principle mandatory, there may be scope for information not to be brought to the attention of college colleagues, or not brought on a timely basis. For instance, as long as supervisors maintain a national focus, the authorities may be reluctant to share information about a troubled financial institution if there is concern that this could lead the authority receiving the information to ring-fence the assets of the institution. If, perhaps, a bank is preparing to downsize, it will likely focus its efforts in those places where it has problems, and supervisors may wish to protect their own jurisdictions by not volunteering information about any problems there; and

• Moreover, there are in some cases legal prohibitions on the sharing of information, so that supervisors from one member state cannot provide more than generalities about the activities of the bank/insurer they are supervising.

28. To some extent these problems can be seen as teething troubles that will likely fade over time as the process becomes established, but a number of specific measures to enhance cross-border supervision can be suggested:

• Collaboration on risk assessment, and adopting a common template, methodology, and scoring scales has already been developed for EBA colleges and similar approaches should be encouraged elsewhere. The risk assessment would best be carried out at the group level, within the colleges, and be interactive, structured, and detailed. One of the key challenges for the EBA will be to further harmonize the convergence on Pillar 2 (common methodologies for risk assessment);

• Collaboration in conducting onsite supervision is essential for supervisory effectiveness. It is already the case that many onsite inspections are joint between home and hosts. There could be a presumption that home supervisors would invite subsidiary and branch hosts to participate, and subsidiary hosts would invite parent home supervisors. Moreover, where the cross-border presence is significant, particularly where the supervisory agency is a member of the core college, the supervisory plan would be discussed at the college level. Participation of all relevant supervisors would also be expected. Such an approach would also serve to mitigate the concerns that have been voiced about the limitations on the powers of national supervisors in the face of branching under the European “single passport” policy. Clearly, this is a significant resource issue, but supervising financial institutions with substantial cross border activity has to be seen as resource intensive;

• Legal restrictions on the sharing of information should be identified and, to the maximum possible extent, removed;

• Agreement on ex ante “burden sharing” principles should be designed in such a manner that supervisors have an interest in the outcome of the institution as a whole, so that the incentives for sharing information are maximized; and

• Where there is uncertainty as to the powers of national authorities, for instance to decline giving “no objections” approval on safety and soundness grounds to a cross-border takeover, such uncertainty could be resolved through the actions of the ESAs.

29. Cross-border activity is of course relevant not just within the EU; nearly all major EU financial institutions have substantial activity worldwide. Insofar as institutions have less understanding of non-EU markets than of those in the EU, this increases the vulnerability of the activity. Many of the problems affecting EU institutions in the crisis originated from outside the EU. Many of the suggestions above therefore apply as much to cross border activity outside the EU as they do to activity within the EU. Joint inspections for instance are similarly useful. And convergence of regulatory regimes will help mutual understanding as to what is happening in the respective jurisdictions. MoUs with U.S. authorities, for instance, should include provisions for joint inspections on a regular basis.

30. The establishment of “crisis management colleges” for larger cross-border banking groups is one important, but in itself limited, element of the framework. Already some groups have been established to bring together the various institutions—supervisors, central banks, and finance ministries—that would be involved in dealing with a crisis in such a bank. Such a pre-existing network should facilitate the logistics of intervention and the negotiation of a coherent approach. However, crisis management colleges will tend to be large and therefore unwieldy. More importantly, the existence of a college will not solve the competing interests over burden sharing in cross-border banking resolution, nor does it create strong incentives for individuals to reveal vulnerabilities and failings on a timely basis.

E. The Financial Stability Work of the European Commission

31. The EC plays a central role in proposing prudential and non-prudential regulations for the financial sector, and in overseeing macroeconomic policies that affect the financial sector. The current legislative agenda is unusually full, with forthcoming amendments to the Capital Requirements Directive (CRD), the directive on Deposit Guarantee Schemes (DGS), the banking crisis management framework, a review of the Market in Financial Instruments Directive (MiFID), as well as measures for the implementation of Solvency II. This agenda largely represents an effort to implement lessons from the global crisis and associated internationally-agreed regulatory changes (e.g., G-20 commitments), for example, to raise the level and quality of bank capital.

32. There has been a shift toward greater reliance on EU regulations, which have the force of directly applicable law, and to allow less room for national discretion in the transposition of directives; the aspiration is to achieve a “Single Rule book.” The shift is motivated by the need to equip the EU financial sector with a consistent set of core supervisory rules, so as to further strengthen EU financial stability. Differences that stem from exceptions, derogations, additions made at national level, or ambiguities contained in directives that have a material impact on the market, that are not as stringent as the minimum core standards, or that may induce competition distortions or regulatory arbitrage, will be identified and removed. However, the rules need to be designed to allow flexibility for the sake of proportionality and to deal with differences in economic conditions across the EU: especially where monetary and fiscal policies are constrained, there may be good reasons to differentiate other policies, for example, for macro-prudential purposes, though as noted above, coordination of such policies will remain essential and could be handled by the ESRB (see below).

F. Cross-Institution Coordination

33. Interagency coordination is effected through a Joint Committee. This is composed of the Chairpersons of the ESAs and staff from the ESAs who are responsible for coordinating the work of the three ESAs on cross-sectoral issues and on the supervision of financial conglomerates. As the three ESAs could, in theory, take different positions on the same issues, which could allow room for regulatory arbitrage, the work of the Joint Committee will be important to bridge the gap with the ESRB, and address underlaps and overlaps.

34. In this context, consumer protection concerns will need to be given due attention. All three ESAs have the same consumer protection mandate, which might on the one hand create overlaps and inconsistencies among them and on the other may lead to inaction and weaken the sense of ownership on the part of each ESA. Moreover, the financial crisis has re-emphasized the connection between consumer protection and systemic stability. To ensure that these issues are effectively addressed, the European system needs to accommodate differences across countries in responsibility for consumer protection: some unify prudential and consumer protection responsibilities, others have a “twin peak” model and will require that the ESAs will have to coordinate across agencies.

A. Crisis Management and Bank Resolution

35. Establishing a comprehensive EU framework for resolution and crisis management is necessary to fill a remaining gap in the current financial stability framework. To underpin the single financial market, resolve the mismatch between the banking passport and national fiscal responsibilities, and align national interests toward financial stability in the EU, setting up a unified framework would be the preferred approach. Such a framework should involve the establishment of a European Resolution Authority with associated financing and fiscal backstops, encompassing ex ante rules to fund resolution of financial institutions, through contributions from the industry and pooled fiscal resources in the event that temporary public financial support is necessary; this would also avoid distortions of competition policy at the national level, as compliance with state aid rules would have been cleared at the ESA level.

36. There does not appear to be sufficient political support to make the quantum leap to an EU wide unified setup. Working within this constraint, the Commission outlined in a consultation document of October 2010 a more gradual approach toward building a comprehensive EU framework for resolution and crisis management that will apply to all credit institutions and some investment firms. The first phase consists of a legislative proposal aimed at giving national authorities a common minimum set of resolution tools, which is expected to be adopted by the Commission in fall 2011, while adoption by the European Parliament and Council is expected by the end of 2012.¹⁵

37. The enhancement of the resolution toolbox envisaged by the Commission broadly goes in the right direction. The aim is to reduce the frequency and severity of banking crises by ensuring that early remedial action is taken, shareholders and other relevant stakeholders have a greater role in burden sharing—thereby reducing the risk of loss for others including governments—and introducing features that facilitate resolution when it is unavoidable. The Commission’s plans would require Member States to ensure that national authorities have an effective range of options for early intervention, attempting to strike a balance between financial stability concerns at the one hand and the need to minimize interference in property rights at the other, and mitigating legal risks deriving from the interference with such rights. Due consideration is also given to the need to coordinate certain elements at the global level, such as on the application of statutory bank debt restructuring or “bail-in.”

38. In a few areas, clarification is still warranted. For instance, the Commission consultation argues that the general rule should be that failing banks should be liquidated under the bankruptcy code through ordinary (or slightly modified) liquidation rules, and that they should be resolved using special resolution tools (such as transfers of assets on a wholesale basis) only as a going concern and only if the public interest of financial stability is at stake; however, there seems a strong case that special resolution tools should be available more broadly, also in the bankruptcy process to allow an orderly liquidation, and for both systemic and non systemic institutions.

39. While the Commission intends to review the EU crisis management framework with a view to possibly establishing an integrated regime in 2014, the authorities must be able to take appropriate measures in the interim if problems arise in cross border institutions. The Commission consultation envisages that group-level resolution plans will be developed by resolution colleges under the leadership of the home country authority and with a strong coordination role for the EBA; such plans would include measures for a coordinated resolution of cross-border group entities, with losses being absorbed by private stakeholders. For EU cross-border banks having a significant presence outside the EU (or for EU sub-groups that are part of global financial conglomerates) such plans should be developed with relevant third countries authorities, to ensure coordination in the resolution process and, ultimately, mutual recognition of resolution measures. It would be useful for the above mentioned resolution plans to be developed soon, well before the 2014 review.

40. A framework for intra-group financial support also deserves consideration, provided that the complex underlying legal and economic challenges are addressed. The Commission envisages in its recent consultation that group financial support agreements be concluded among the various group entities in case of troubles affecting the financial stability of the group as a whole. Such a framework may provide a useful tool to facilitate private funding of troubled groups and may give an early intervention mechanism that stabilizes the financial soundness of the groups. Again, the EBA could play a role in facilitating joint decision.

41. Agreement ex ante on principles for specific burden sharing is necessary for a fully satisfactory resolution regime, given that, due to market failures, temporary funding of a resolution by Member States might be necessary.¹⁶ In particular, the resolution plans discussed above should provide for ex ante burden sharing principles, with appropriate participation by national fiscal authorities. Any agreement on burden sharing should cover both the financing of restructuring and the allocation of any net costs. Reaching agreement may be difficult, but if the resolution framework is to reduce the expected cost to government from bank failures, the effort will be well worthwhile.¹⁷

42. Two elements of the new regime have an important bearing on the effectiveness of the financial stability framework: bank creditor bail-ins and sovereign risk. While the framework for bank creditor bail-ins is being discussed in a global context (FSB), uncertainty about its parameters and timing is complicating current crisis management. In addition, the explicit recognition of the need for market discipline on sovereign borrowing has important ramifications for the approach to risk weighting of government debt on the books of financial institutions. Clarifying how both will be handled is urgent and seems essential to help resolve current conjunctural strains.

B. Financial Stability, State Aid, and Competition

44. The EU has had to adapt its rules on state aid to take into account the exceptional conditions generated by the global financial crisis. The Treaty on the Functioning of the European Union contains strict limitations on state aid being used to distort competition and the internal market, but also allows some derogation when a country’s economy is subject to a serious disturbance. With this justification in mind, the European Commission has issued a series of Communications¹⁹ defining how it will assess state aid to the financial sector in the context of the current financial crisis. These communications emphasize the temporary nature of the crisis; the objective of ensuring that institutions receiving aid are viable and are quickly restored to a condition where they can safely operate without aid; the need to take into account systemic stability and the maintenance of credit flows; and the desirability of containing moral hazard by ensuring that shareholders assume a large share of the costs associated with the restructuring of the troubled institution and do not unduly benefit from the aid received. Against this background, the Commission has issued rulings on numerous cases of state aid to individual financial institutions and sectoral support schemes.

45. The European authorities are working to formulate a more permanent special framework for state aid in the financial sector, in recognition of certain special features of the financial services industry: the financial sector plays a unique systemic role in the economy, but it is also subject to systemic vulnerabilities and very sudden crises.²⁰ The financial sector is heavily regulated for prudential purposes, and is protected by certain safety nets, such as a DGS, the availability of Emergency Liquidity Assistance, and, in some countries, a Financial Stability Fund or a resolution fund.

46. More specifically, a dedicated regime for state aid to the financial sector should acknowledge systemic stability concerns, which recognize that:

• Risk is reduced by diversification of sectoral and country exposures. A business line or subsidiary that is “non-core” can contribute to stability if its returns are not strongly correlated with those of other business lines;

• The effect on a bank’s margin of support received by an institution is ambiguous. A stigma may attach to receiving state aid, such that the recipient’s marginal funding and capital costs are high; on the other hand, a bank may benefit from the public’s perception of state support;

• Support for one institution could sometimes have positive spill-over effects on others, for example, by underpinning investor confidence; and

• Moral hazard factors are already pervasive in the financial sector, even in the absence of an institution requiring state aid; prudential regulations and mechanisms already go some way to limit this effect. For example, a systemically important financial institution (SIFI) may be subject to a capital surcharge or a higher levy in part because it may well be rescued if it gets into difficulty. Such a measure serves to achieve a similar objective to that of the state aid regime, and indeed a degree of burden sharing with shareholders is achieved in advance.

47. The speed with which financial crises can develop suggests that the operational procedures for assessing state aid need to be implementable under very tight deadlines. In some cases it may be necessary to provide assistance in a matter of days, and a delay to review acceptability under state aid rules could be very costly.²¹ So far, the European Commission has always managed to decide on assistance within the market constraints on timelines since existing procedures are already made sufficiently flexible in this respect. However, also during normal times, and not just during a crisis, it will be useful to validate in advance the pricing of assistance that is available from various standing facilities, such as a bank resolution fund. In the absence of such standing facilities, emergency state aid to a financial institution could be presumed to be acceptable, but subject to ex post review and possible correction in line with the predetermined principles.

48. The new bank resolution regime, by facilitating earlier and more market-based restructuring of problem banks, should reduce but not eliminate the need for a regime for state aid to the financial sector. The two regimes should be consistent with each other, so that, for example, principles applied to divest activities under early intervention should not contradict the principles of bank restructuring under the state aid regime. State aid has been interpreted widely, to encompass also temporary government financing or guarantees to facilitate restructuring. Moreover, such assistance may turn into losses for government if the restructuring does not succeed and/or an appropriate remuneration for the State funds has not been paid. Lastly, the circumstances under which the financing from a compulsory DGS or a bank resolution fund would count as state aid should be clarified. The Commission will need to ensure that the effectiveness of its crisis management framework is not undermined by other EU rules and directives that were largely developed before the recent financial crisis and may not fully recognize the implications of the crisis for effective resolution. This could include the Financial Collateral Directive, in respect of termination rights.

49. A dialogue with prudential regulators and supervisors is advisable. Competition and prudential authorities may have parallel responsibilities in areas such as the vetting of major acquisitions and changes in ownership. They may also have complementary information, for example, on competition between banks and nonbanks and the scope for cross-border entry. Competition authorities need to take prudential regulations into account in assessing what is a “normal” balance sheet structure and what action is needed for a weak bank to become securely viable, because meeting prudential norms is one element of viability.

C. Macroprudential Instruments

50. The institutional structure for the implementation of macroprudential instruments remains decentralized at the Member State level. Macroprudential policies should be calibrated mostly based on country-level macro-financial conditions. In some countries work is advanced in developing a macroprudential framework, although in others this is still at an early stage of development. Nonetheless, there is a strong case within the EU for coordination through an EU institution (e.g., the ESRB) to deal with single passport and home-host concerns.

51. One area in which the ESRB’s powers could be usefully expanded is that of the coordination of macroprudential tools and ensuring their consistent application across borders, so as to limit the scope for regulatory arbitrage as well as for leakage, and ensure that similar products are treated similarly for all institutions. The ESRB could take a more proactive role in issuing recommendations to the Commission, as part of the harmonization of national regulatory frameworks, regarding the set of available instruments to be used for macroprudential purposes by the national authorities; similar considerations apply to the imposition of loss absorbency requirements for globally systemic financial institutions. ESRB guidelines would then help operationalize these macroprudential instruments. The tool-box of macroprudential instruments available to the authorities should be easy to implement and to enforce, be effective in mitigating macroprudential risks, have limited distortionary effects, and not result in regulatory arbitrage. Given that systemic crises can originate from various segments of the financial sector (banks, non-banks such as hedge funds, insurance companies), a range of macroprudential instruments should be available for each of these segments. Since possible challenges to financial stability are diverse, one can imagine that instruments might need to be applied to specific institutions, to sectors of the financial system, to regions within a country, regions cross-border, nationally or more widely. A broad range of macroprudential instruments (such as was earlier prepared by the ESRB) should be available to the authorities. If a Member State wishes to apply a macroprudential tool and wants to secure reciprocity from an institution operating in its territory but not subject to its national supervisory oversight, it could submit a proposal to this end for approval to the ESRB.

52. Recommendations on the design of the macroprudential instruments and their calibration should ideally be binding on the Member State. The role of the ESRB could be substantially enhanced, potentially as part of 2014 review that the Commission intends to conduct in 2014 on the new EU financial stability framework, to define the set of instruments and the range over which they may normally be applied, or to review macroprudential measures introduced at a national level with a view to ensuring reciprocity (similar treatment for all institutions performing a particular activity). Nonetheless, careful consideration will need to be given whether it will be possible under the existing EU legal framework to confer upon the ESRB such binding powers toward Member States. A “comply or explain” regime in this regard, analogous to that in place concerning its warnings, would be a weaker alternative.

53. For banks, the set of instruments to be included in a macroprudential legal framework should encompass a broad range of tool. addressing both the demand and supply of credit. Work is ongoing to identify and put in place a comprehensive set of instruments. Amongst the instruments that might be included are:²²

• A countercyclical capital buffer, where the overall buffer for internationally active banks should be a weighted average of capital add-ons set by national jurisdictions, with weights based on the geographic composition of a bank’s portfolio of exposures, as under the Basel III approach. The EU Capital Requirements Directive (CRD4) implementing Basel III should set an ambitiously high common bar that enhances the safety of the European financial system. At the same time, it should allow sufficient scope for the calibration of macroprudential instruments such as countercyclical capital buffers to cope with asynchronous credit cycles, and recognize that there may be still a need for the ESRB to issue recommendations and endorse the introduction of other macroprudential tools. The countercyclical capital buffer has some limitations: (a) there will be long lags in implementation (banks are likely to have one year to adjust their countercyclical buffers); (b) it is a fairly complex instrument; (c) its effectiveness remains untested.

• Limits on loan-to-value ratios could form part of the policy toolkits of national macro-prudential authorities to allow a swift reaction to emerging risks in real estate markets. Existing experience with limits on LTVs suggests that this measure can be effective in dampening mortgage credit growth and in slowing house price appreciation. As with the limit-to-debt ratio below, to prevent regulatory arbitrage, such limits should be applied not only to banks but to non-banks as well. To avoid circumvention through cross-border lending, such limits could be reviewed by the ESRB, which could have the power to ensure reciprocity.

• Limits on debt-to-income ratios could be considered as a macro-prudential instrument. They would help ensure that borrowers have the ability to repay their mortgages. The experience of some countries that have relied upon such limits on DTI for a long time (e.g., France) should prove useful in designing and calibrating this instrument.

• Measures to address the cross-sectional dimension of systemic risk could also be included, such as capital or liquidity surcharges, risk-weighted instruments, or levies on SIFIs.

54. For non-banks such as hedge funds, possible additional macroprudential measures include:

• Margins or valuation haircuts on securities used as collateral in the securitized lending markets (such as repo markets). This instrument would be used to regulate the supply of secured funding. It could also mitigate the pro-cyclicality of the shadow banking system by affecting their funding conditions given that most source their funding in the wholesale markets.

• Leverage limits. A leverage ratio in the hedge funds sector could help dampen lending exuberance. The EU directive on Alternative Investment Fund Management already gives an advisory role in this regard to the ESRB. However, the draft regulation on a ban of short selling and naked CDS currently does not include any role for the ESRB, which instead is given to ESMA.

55. For insurance companies and pension funds, there may be a need to identify more specialized measures, for instance in relation to the low long term interest rates and longevity risk, and risks associated with derivative products.

D. Data Adequacy

56. An important aspect of the new European supervisory architecture regards its ability to eliminate data gaps and foster an efficient flow of information between its various components. Compiling and disseminating appropriate information on risk exposures and interconnections among institutions is a precondition for the establishment of an effective framework targeted to achieve stability in the financial system. The framework should ensure vertical (across the three layers) as well as horizontal (within members in each layer) distribution of data and information in general. While an effective framework that protects the confidentiality of information needs to be put in place, maximum data sharing should be the goal: lack of data or lack of data sharing should be eliminated as a potential cause of the next crisis.

E. Direct Regional Supervision: The Rating Agencies

63. On 1 July 2011, CRAs will be the first cross-border financial institutions in the EU licensed and supervised by a supranational agency—ESMA—and to this end, ESMA has been endowed with a comprehensive mix of supervisory tools. If ESMA manages to instill a supervisory culture and implement supervisory tools tailored to cross border entities, it could pave the way to possible future centralized supervision of financial markets and/or institutions in the future.

64. ESMA will have the right to request a broad set of reports and returns, and also conduct on-site inspections. Hence, it is developing reporting templates and systems on CRAs that will facilitate the processing, monitoring and analysis of the prudential information received. A central repository (CEREP) will gather all the information for prudential purposes. Part of the information will be made public, on an aggregated and bi-annual basis (as from summer 2011); published information will include CRAs’ performance indicators, including rating actions (downgrades/upgrades), defaults, and transition matrices. ESMA can also access staff and management, and can conduct on-site inspections, which may help enhance its credibility as well as its effectiveness.

65. ESMA can impose stringent fines, according to a comprehensive list of breaches attached to the regulation. It can also require corrective measures, including suspending the use of credit ratings for regulatory purposes, issuing a public notice, temporarily prohibiting the credit rating agency from issuing credit ratings and, as a last resort, withdrawing the registration when the credit rating agency has seriously or repeatedly infringed the regulation. With regard to accountability, ESMA will present its annual report on fines and corrective measures to the EU Commission, Parliament and Council.

66. ESMA will possess its own budget for the supervision of CRAs. While the calibration of the fee is still under discussion (secondary legislation on fees is to be adopted by the EU Commission), the Commission should only set a minimum and a maximum percentage, to give some leeway to the Authority and at the same time establish criteria that ensure a degree of predictability.

67. The accumulated delays in hiring staff and preparing tools and methods for supervision at ESMA’ may hinder its efforts to establish credibility for its supervisory function quickly. By the end of 2011 ESMA is expected to have 15 staff working on CRAs. While supervision legally starts in July 2011, it will initially apply to a narrow set of entities; only four rating agencies—including none of the three largest ones—had been registered as of May 2011. The delay was due mainly to the complexity of the license applications, and the continued functioning of a collegial procedure (where ESMA was an observer and not the licensing authority) on a transitional basis. The largest CRAs are now expected to be registered by the summer. Meanwhile, ESMA is undertaking an intensive recruitment process.²⁹ A balance between supervisors and former private sector practitioners should be an objective in order to ensure the build-up of a supervisory culture and independence from the industry.