Brussels, 1 October 2012
Commissioner Michel Barnier's reaction to the Basel Committee's preliminary "Regulatory Consistency Assessment"
The Basel Committee on Banking Supervision has been carrying out a preliminary assessment of the consistency of the Council's general approach on the Capital Requirements Directive IV (CRDIV) with the so-called Basel III agreement on internationally agreed capital standards. The outcome of its work so far has been published today, along with reports on the regulatory consistency in the US and Japan. The Committee's work will continue with a final report on consistency being issued once CRD IV is adopted.
Commissioner Michel Barnier, responsible for financial services, reacted as follows:
"In 12 out of 14 areas of the preliminary "Regulatory Consistency Assessment" concerning the EU published by the Basel Committee today, the draft European legislation has been fairly assessed to be "compliant" or "largely compliant". I have, however, reservations about the preliminary findings in the remaining two areas, which do not appear to be supported by rigorous evidence and a well-defined methodology. I believe that this has led to an apparently significant lack of consistency in the way judgement and gradings have, in this preliminary phase, been applied in those two areas across jurisdictions. The European Commission and the European members of the Basel Committee have provided extensive information and clarifications to the Basel Committee during the process, but unfortunately this has only been partially reflected in this present preliminary report. Here at the Commission, we stand ready to support the further work by the Basel Committee to improve its assessment of standards implementation and are confident that the final report of the Basel Committee will constitute an improvement both in the assessment of the EU and in the coherence across jurisdictions"
He added that:
"The European Union remains firmly committed to the robust implementation of the internationally agreed capital standards, the so-called Basel III agreement. It will be applied to more than 8,000 banks in Europe, representing more than 50% of world banking assets. EU implementation of Basel III is a key contribution to sustainable growth and financial stability, not just in the Union, but also globally. We presented our proposals in July last year. They are currently being discussed by the European Parliament and the Council. All institutions are determined to successfully conclude the legislative process this autumn.
It is of the utmost importance that the Basel III agreement is applied consistently around the world, in order to ensure global financial stability and a level playing-field. Therefore, I fully support the Basel Committee's intention to assess consistent implementation of these rules for all internationally active banks worldwide and welcome the work of the Basel Committee in this area."
The report assessing the EU is based on the "general approach", a unanimous agreement between all Member States about how to implement Basel III for all EU banks. This "general approach" builds on the Commission's proposals of July 2011 (see IP/11/915) and constitutes an important step in the legislative process, but not the final legislation. The existing Basel II and Basel 2.5 agreements, which form an integral part of Basel III, have already been implemented for all EU banks in accordance with the internationally agreed timeframe1 – this is not the case for all jurisdictions assessed.
However, application to the wide variety of EU banks necessitates some flexibility for non-internationally active banks, which are outside the scope of the Basel Committee's assessment. Nevertheless, the legislation proposed by the Commission last year gives supervisors the necessary tools to ensure that all internationally active banks fully comply with Basel III, as the report acknowledges. However, given the lack of factual evidence of compliance failures, the conclusions are based on speculation that EU members of the Basel Committee will not respect their commitments. Such speculation is unfounded and has been disproved by Basel II and Basel 2.5 implementation. Furthermore, important safeguards provided by disclosure and oversight carried out by the European Banking Authority, for example in the context of the definition of capital, have not been fully taken into account. This is not consistent with the way supervisory discretion and informal rules have been assessed in the other two reports.
In 12 out of 14 areas of the report, the draft European legislation has been fairly assessed to be "compliant" or "largely compliant". However, regarding the remaining two areas, which concern the definition of capital and the internal ratings based approach to the calculation of credit risk capital requirements, the EU is deemed "materially non-compliant. In those two areas, the findings do not appear to be supported by rigorous evidence and a well-defined methodology. This has led to an apparently significant lack of consistency in the way judgement and gradings have been applied across the three jurisdictions. As an example, the assessment of the EU's internal ratings-based approach (IRB) implementation is considered "materially non-compliant" because EU banks that use the IRB apply the simpler Standardised Approach permanently to about 5% of their exposures on average, which arguably translates into a negligible impact on overall capital requirements. At the same time, in one of the other assessments, the IRB section is considered "largely compliant", even though none of the banks in the assessed jurisdiction is held to comply with Basel II minimum capital requirements. Instead, those banks still use the outdated, risk-insensitive Basel 1 rules, which, according to their supervisors, produces 20% lower capital requirements overall. Apparently, an alleged deviation from the rule is considered worse than the non-application of the rule itself. In the following, the concerns regarding the preliminary report's findings and gradings are explained in more detail.
Intransparent section grading
It is not transparent how the individual "potentially material" findings translate into a section grading. In the case of the IRB, a single – questionable – finding, regarding permanent partial use, seems to have turned the whole IRB section grading for the EU into "materially non-compliant." This does not seem justified given the unclear Basel rule on this point and the unquantified impact of this legislative choice on capital requirements. This finding is also not comparable to the situation in the securitisation section of one other report, where a "materially non-compliant" grading has been attributed because a methodology for assigning capital requirements is used that is not based on credit quality assessments and therefore completely different from Basel II.
In the section about the definition of capital, a section grading of materially non-compliant has been assigned even though none of the seven "main specific issues" has been clearly demonstrated to be material. Moreover, the assessments of these issues contain material shortcomings, which are explained further below. At the same time, again comparing with one of the other reports, a similar number of individual findings in the definition of capital section that are comparable, or even identical, on substance lead to a section grading of "largely compliant".
Shortcomings in individual findings
The report fails to explain what quantitative importance a finding must have in order to be classified as material. Moreover, the "main specific issues" identified in the report are in all instances of "potential", rather than actual, materiality. The findings are usually ascribed to concerns about possible future behaviour by banks and their supervisors. However, it is never spelled out what future behaviour has been assumed by the assessors for this "potential" to materialise. Worse still, it appears that assumptions have not been made consistently across the three jurisdiction assessments. It appears that in the other two reports, more often than in the EU report, findings have been considered non-material based on an expectation that they will be mitigated by informal rules or certain behaviour of supervisors. A case in point is the recognition of capital instruments, where the proposed rules in one of the other two assessed jurisdictions give supervisory authorities far greater leeway than the EU rules do. However, only in the EU report there seems to be an assumption that competent authorities will not act appropriately, despite the existing safeguards of disclosure and discipline through the European Banking Authority.
It appears that at times, individual bank data has been used, but no account is given of the specific instance of the bank in question and of whether or not the relevant finding can plausibly be extrapolated to other banks. A case in point is the finding on non-joint stock companies. This indeed concerns a single internationally active bank. It is unclear that the issue is really material in the specific case and it is unreasonable to assume that a similar concern, which is specific to the cooperative legal form and applicable laws in one Member State, could arise at other internationally active banks.
Another instance is the permanent partial use of the IRB. The concern seems to relate to the observation that up to "around 20%" of the exposures of one of the banks surveyed are treated under the standardised approach instead of the IRB approach. However, the average for the surveyed banks is only 5.49% and no estimate of the impact on capital requirements is provided in order to confirm the materiality of the finding. Notably, when comparing again the three reports, it turns out that the Basel Committee's assessors are much less concerned if in another jurisdiction, all internationally active banks are still subject to the outdated Basel 1 framework and this leads to around 20% lower capital requirements overall according to local supervisors. In any case, the finding is questionable given that limited permanent and transitional partial use is in principle allowable according to the Basel II accord and it is not clear that the identified partial use in the EU banks goes beyond the allowable extent envisaged by the Basel Committee. It is also not clear that the impact of applying the Standardised rather than the IRB approach to the exposures in question is actually material in terms of resulting capital requirements.
As mentioned above, the grading in the IRB approaches section is based on a single finding that has been considered material. Regarding the definition of capital section, there are the following concerns with the individual findings:
Common shares: The CRR requires that the instruments fall under the Article 22 of Directive 86/635 and that these instruments meet the 14 criteria set by the Basel III agreement." The term "common share" that is used in the Basel III agreement is not defined in European legislation and different definitions of common shares exist in Member State company law. For this reason, the articles of the proposed regulation do not use the term common share, which the preliminary report correctly points out. The preliminary report however fails to show that the instruments that will be eligible under the proposed regulation and meet the substance of the 14 criteria could indeed fall short of the Basel Committee's expectations as to what "common shares" deliver. Moreover, the recital No. 53 explaining the relevant stipulations clearly expresses the legislator's expectations that the instruments are indeed "common shares" for banks that have such shares listed on a regulated market – the report does not explain why it expects supervisors not to live up to the legislators' expectations. It would have been fair to announce further work in the "level 3" assessment, but the report does not do it. In the case of one other assessment, on a similar issue, the assessors have been satisfied with a simple promise from the local authorities to use their discretions only "in case of emergencies" and to disclose the use when it occurs. Notably, the public disclosure of all instruments recognised is also a requirement in the draft EU law.
Treatment of insurance subsidiaries: the assessment team concludes that the impact is material, disregarding that the issue can arise only in very few banks that formally constitute "financial conglomerates". Further, as the report concedes, the verdict is made without actually knowing what the consolidation method, which is still being defined, is and what its precise impact will be. However, even absent a detailed definition of the consolidation method it is clear that the additional consolidation at the level of the conglomerate effectively prevents double gearing. It ensures that at least the capital requirements of the insurance subsidiary have to be supported by consolidated capital of the group. The approach is prudentially sound, not least because by contrast to a simple deduction it avoids negative incentives for the level of capital held at the insurance subsidiaries. It should be noted that draft rules in another jurisdiction also allow for a consolidation method as an alternative to deduction, but this seems not to have affected the section grading for that jurisdiction in the same way.
Loss absorbency at the point of non-viability: The absence of a requirement for a contractual clause to that effect can have some transitional impact, but that impact cannot be material once a statutory regime will be introduced with the adoption of the bank resolution proposals. This proposal will enter in force at most two years later than the legislation implementing Basel III. However, given that Basel III requires a phasing out of 10% of the relevant instruments only in 2014, there is no impact of this difference in 2013 and a negligible impact in 2014.
Minority interests: The additional discretionary capital requirements that can be included in the minority interests are likely to be of low materiality. Moreover, the proposed treatment is sound, as only minority interest that are used to meet legally binding capital requirements at the level of subsidiaries are included in the consolidated capital. In that sense, Pillar 1 and 2 capital requirements as well as systemic risk and countercyclical buffer requirements are all equivalent. In particular, the EU Council has introduced the systemic risk buffer as a functional equivalent to higher Pillar 1 capital requirements in order to give Member States' public authorities some flexibility; and Pillar 1 capital requirements are actually included in the Basel II eligible minority interests. In addition, some supervisors have indicated that the current Pillar 2 requirements had been introduced to pre-empt the more stringent Pillar 1 minima under Basel III; hence, once the new rules will enter in force, they will be materially reduced as will be the related minority interests.
Non-joint stock companies: The Basel III agreement recognises in a footnote the particularities of non-joint stock companies. This recognition is also reflected in the EU draft law and should not be seen as a compliance issue. Moreover, it is unclear what the report's view on materiality is based on. This issue is of relevance only in the rare case where an internationally active bank is set up as a cooperative and the cooperative shares are by law required to be redeemable. It is not a case where it could be reasonably assumed that other internationally active banks would turn into cooperatives and get into the same position in the future. At the same time, the actual impact of potential redemption on such a bank's capital position has not even been assessed; it is not clear whether potential redemptions could become material given that the redeemable share capital as opposed to reserves etc. constitutes only a fraction of CET1 capital for cooperative banks.
For Basel 2, a phased implementation starting in 2006 and for Basel 2.5, an unambiguous start date of 31 December 2011 have been agreed internationally.