09 Aug 2018

A supervisory architecture fit for CMU

Aiming at a moving target?

Marco Lamandini

Download Publication


European Capital Markets Institute, Commentary No. 55

Marco Lamandini is Professor of Corporate and Capital Markets Law at the University of Bologna and Member of the Academic Board of ECMI. He also sits on the Board of Appeal of the ESAs – appointed by ESMA and the Appeal Panel of the SRB.

There are certainly many ways to look at the intersection between the Capital Markets Union (CMU) and current review of the European Supervisory Authorities (ESAs). In this commentary, I will briefly elaborate on some of the legal challenges that seem to be working against the adoption of the proposed reforms.

We all know that the CMU is a moving target and its institutional architecture is by its very nature evolutionary. In principle, adapting the architecture should bring about increased financial market integration or help remove hidden barriers which still fragment capital markets. However, the momentum for reform also reflects the specific weaknesses exposed by the economic cycle and the prevailing political set-up (the ESMA-proposed CCP Executive Session in the context of Brexit being one clear example). This is precisely why this reform matters.

This is indeed a true green-field to test resilience and adjustment capabilities of the European Union as such and of the euro area in particular in these very difficult times. The daunting challenge is to reconcile the pan-European striving towards a level regulatory and supervisory playing field grounded in a Single Rule Book, while sitting on a two-fold rationale: a single market rationale meant to facilitate capital markets integration and a financial stability rationale meant to protect a highly interconnected financial market. We must add the right degree of flexibility, proportionality and, most importantly, subsidiarity and respect for fiscal responsibility. Legal determinants can also play a role in this very delicate balancing act.

Banking Union

Banking Union was easier to design and implement! I realise that this may seem a false statement, taken on its face, if one only considers the daunting institutional complexities we had to address to deliver. ‘Formal’ banking (at least the one targeted so far by the Banking Union), however, has clear boundaries and a specific legal basis: indeed, Article 127(6) TFEU was key to moving a giant institution like the ECB into the supervisory role. I am not suggesting that implementing the Single Supervisory Mechanism (SSM) and the Banking Union has been a simple exercise. It was not (also because, unlike EMU, which could rely on previous legal convergence based upon Article 131 TFEU, Banking Union was established in a rush, ‘jumping’ from minimum harmonisation to a single rule book). The L-Banken judgment (T-122/15) of May 2017 nicely portrays the gymnastics necessary to reconcile ECB exclusive competences under Article 4 SSMR with the decentralised exercise of those tasks over less significant institutions (LSIs) by national competent authorities (NCAs) under Article 6 SSMR. At the same time, it also indicates (paras 109-111) that, at the margins, there may be hidden threats to the ECB’s competence in national laws and constitutional counter-limits.

If you want to have a quick sample of legal problems currently discussed within the Banking Union – excluding those that are overarching but in the first place of a political nature, revolving around the legal foundations of the European deposit insurance scheme (EDIS) and of the European Monetary Fund as the final backstop for the SRF – let’s consider the following:

i) the unprecedented (and thus untested) practice of enforcement, by an EU institution, of national laws implementing CRD IV (judgment 24.4.2018, Credit Agricole, T-133 to T-.136/16, being an example); this may occasionally lead even to diverging outcomes for the same person depending on the applicable national law e.g. on fit and proper (in particular integrity requirements);

ii) excessive administrative complexity, in particular where compound proceedings, in which both a NCA and ECB procedural “leg” apply (pending case CJEU C-219/17, Fininvest; pending case T-913/16, Fininvest), or where the SSM peculiar governance requires the Governing Council to adopt by no-objection more than 1,500 decisions a year; or

iii) blurred redlines of the scope of special tasks conferred upon the ECB under Article 4 SSMR, with quite visible ‘creeping’ extensions of ECB supervisory competences beyond the micro-prudential tasks enshrined in the CRD IV, CRR and BRRD to also encompass tasks that are only in national law but are considered within the functional remit of the ECB under Article 4 SSMR. [1]

But, to my mind at least (the point was recently made also by the Governor of the Bank of Italy), the most fundamental challenge within the SSM still lies on how to reconcile, at the regulatory and supervisory level, micro-prudential policies with macro-prudential and monetary policy. This is particularly visible during a recession when monetary policy tends to be expansionary and micro-prudential policy requires the strengthening of capital requirements in bad times. The same tension arises when we must reconcile macro prudential needs, such as the issuance of MREL/TLAC securities to ensure the viability of resolution at the PONV, with the conduct risk/investor protection[2] regulation. My point, here, is that where macroeconomic or macro-prudential objectives cannot be reconciled with micro-prudential ones, the former should prevail.

In turn, the SRM also had its true baptism by fire with Banco Popular. More than 100 pending cases before the GCEU (General Court of the European Union) show that ‘shared’ competences (in the adoption and implementation of the resolution decision) both horizontally between ECB (FOLTF assessment), SRB (adoption of the resolution scheme under Article 18(6) SRMR) and the European Commission and Council (decision over the resolution scheme with endorsement by no objection under Article 18(7) SRMR) and vertically between SRB and the NRAs are a multiplier of legal claims at both national and European level. The good news (if I may say so) is that at least some of these pending cases may offer the opportunity to the GCEU to state the law concerning the conformity of the SRM architecture and the Board’s technical discretion to administer this regulated field with its legal basis, grounded in Article 114 TFEU.

Capital Markets Union

As the legal basis for the CMU supervisory architecture, Article 114 TFEU is indeed one of “the phantom(s) of the opera”. Yet this is not the only reason why CMU was by far more complex than the SSM to achieve.  CMU features an extreme heterogeneity of i) market players (issuers, intermediaries, infrastructures, service providers) and ii) products, and a ‘variable geometry’ of their European regulation and national or European supervision.

Even conceding that EU rules and supervision should come into the picture only for cross-border activities or institutions, it is difficult, perhaps impossible, to adopt a unitary approach. And the De Larosière Group was certainly right when, suggesting the unification of all supervisory activities for cross-border institutions at the pan-EU level, it noted that:

The complexities and costs entailed by such a proposal (which would result in a two-tier supervisory system, one for cross-border institutions and one for domestic institutions), its political implications and the difficulty of resolving cross-border burden sharing are such that the Group has doubts of it being implemented at this juncture. This scenario could become more viable, of course, should the EU decide to move towards greater political integration (De Larosière Report, 2009, paragraph 218).

It was thus no surprise that the Five Presidents report (Juncker, 2015) rightly suggested that a Single CMU Supervisor is only a long-term objective. Therefore, the current reform proposal is just a second step in a long journey and must be assessed in this context. The selection, for example, of only some categories of prospectuses for cross-border offerings or some, newly EU-regulated, funds (EuVECA, EuSEF and ELTIF) among those managed by the same AIFMs is not necessarily an inconsistency but rather a consequence of an evolutionary approach in scaling up at the European level financial supervision. However, even this not revolutionary reform met strong opposition on legal grounds.

Institutional design

At the time of their establishment, the three ESAs were structured on the basis of a single template (and the founding regulations were almost identical). The European System of Financial Supervision (ESFS) operates on a sectoral basis, supervision is carried out by the NCAs or colleges of NCAs and the ESFS is thus ‘largely decentralised’ (Moloney, 2014, p. 943). Should this single template be reconsidered? It should be noted, first, that the (almost) uniform template was already partly derogated by the conferral of limited, enumerated direct supervisory powers to European Securities and Markets Authority (ESMA), first on credit-rating agencies (CRAs) and then on trade repositories. This deeply modified the “functional and structural parallelism” (D’Ambrosio, 2018) among the three ESAs.

The purpose of the ESAs’ reform is two-fold. On the one hand, it is aimed at furthering the parallelism, removing some existing discrepancies in the founding regulations on tasks and powers in Article 8 (extending to ESMA and EIOPA the power to adopt supervisory handbooks and rules on stress testing, already granted in 2010 to EBA). On the other hand, the reform deepens the divide between ESMA on one side, and the European Banking Authority (EBA) and the European Insurance and Occupational Pensions Authority (EIOPA) on the other side, because it envisages the conferral of additional direct supervisory remits to ESMA (alone).

This is also in line with the revised European Market Infrastructure Regulation (EMIR 2) proposal, which envisages the establishment of a dedicated body within the ESMA governance (the CCP Executive Session) and the conferral to ESMA of significant extraterritorial supervisory powers in respect of Tier 2 third-country CCPs. This, in my view, signals that the De Larosière Group was right in predicting that “there may be merit, over time, in evolving towards a system which would rely on only two Authorities, one responsible for prudential issues and financial stability, the second for conduct of business and market issues” (twin peaks approach) (paragraph 216). Politically, it is still too early and the legal basis in the Treaty still in need of further testing. But the direction, to some extent, is clear, and a time will come, in the EU27, when we shall start discussing if and how to possibly merge SSM, EBA and EIOPA (using as legal basis, failing a new Treaty provision, Article 114 instead of Article 127(6), which has limitations on insurance supervision).

Two additional factors, to my mind, are preparing the ground for such (desirable) direction of travel, although in practice, merging EIOPA and EBA with the SSM may prove to be extremely difficult, as the insurance sector is firmly against such a move, and the degree of consolidated capital base is much less advanced than in banking. First, SSM is gaining traction as the leading EU micro-prudential supervisor, as recently witnessed by the relocation of Nordea in Finland and by the strict approach adopted on outsourcing and delegation within the context of Brexit banks’ relocation plans. In turn, the EMIR 2 reform is in fact struggling with, and showing the shortcomings of the current uneven ‘silo’ model: ESMA, as the only one among the ESAs with experience in direct supervision, is granted extraterritorial direct supervisory powers over Tier 2 third-country CCPs, which stretch also into the prudential domain. Its lack of expertise in this area is partly remedied by the fact that it will act in agreement with the ECB and the other central banks of issue.

But this teaches us a lesson.  Beware of obsessions with symmetry!  Due to the uneven stage of development of the Banking Union and CMU, one institutional model or template cannot fit all. This could also justify conferring the new task of setting supervisory priorities (Article 29a, which raised concerns also by the ECB) only to ESMA, to the extent necessary to ensure a smooth functioning of the coordinated supervision with NCAs, but not to EIOPA and EBA, in this latter case preventing an undesirable overlap of competences that could threaten the ECB’s independence.

Governance and powers

On the ESAs’ governance, I welcome a more independent and efficient Executive Board, with full-time senior members, to some extent mimicking the SRM Board (I would recommend however a bridge between the EMIR 2 and the ESAs reforms on ESMA governance, to the effect that the CCP Executive Session would be designed in the context of the Executive Board, so as to prevent the emergence of a “two-headed” organisational structure within ESMA). Some claim that such an Executive Board would threaten the existing balance of powers and NCAs would become ‘supervised’ by a European supervisor. The rationale, however, is the opposite, and the change is motivated by the need to redress existing imbalances, which make the ESAs less sharp than desirable in contrast to potential NCAs’ strategic behaviour.

And as Karel Lannoo recently reminded us (Lannoo, 2018), “fragmentation of supervision gives way to regulatory competition, reduces investor protection and ultimately increases the cost of capital, as investors stay on the sidelines”. It is also argued that the supervisory board of the SSM is composed of a majority of representatives of the NCAs. However, under Article 26 SSMR, the Chair and Vice-chair are appointed by the ECB with the approval of the European Parliament; the Vice-Chair and four members are representatives of the ECB and all members of the board “shall act in the interest of the Union as a whole”. In turn, it is argued that, within the SRM, the Board’s decisions are endorsed by the European Commission and the Council. This is true, however, for the adoption of the resolution decision, but not for all Board’s decisions (including the public interest test determination under Article 18(5) SRMR).

I see merit, thus, in the establishment of the Executive Board. And to the extent that the Executive Board is granted, in addition to the administrative and preparatory role, the powers and tasks currently assigned to the Board of Supervisors of a non-regulatory nature (those under Articles 17, 19, 22 and 32), I hardly see how the system would be converted into a supervisor of supervisors. It is rather a democratic example of decentralised integration, where the coordinating role of the hub is balanced by the governing attributes reserved to the spokes. With a lot of potential to specialise, each spoke pointing in different directions, can leverage the overall efficiency. At the same time, I still believe, for instance, that the supervisory tool of breach of Union law under Article 17, if exercised (and to this effect, the procedure should be simplified), could convey a strong and right signal to curb national opportunism and to foster effective supervisory convergence. But, as long experience shows, this tool can work only in the hands of an independent and empowered Executive Board.

To dispel concerns, for additional powers a tiered approach could be adopted and more ‘intrusive’ decision-making powers, including those on strategic supervisory planning and independent reviews (if, at the end, they will be granted), could be granted to the Executive Board, but subject to a no-objection decision (with qualified majority) of the Board of Supervisors. This could apply also to guidelines and recommendations under Article 16, with a note of caution for the safeguards currently envisaged in the proposal. If there is a need for a preventative check that the ESAs are not exceeding their competences, the right organ to be vested with this competence, to my mind, is the joint Board of Appeals and not the Stakeholders’ Group. And this could also be extended to ESAs’ decisions of general application on reporting templates, which could nicely replace the current requirement for Implementing Technical Standards (ITS) on this, streamlining the process (without the need of endorsement by the Commission) on a tool that needs regular updates and clearly does not impinge on any policy choice.

All direct supervisory tasks and powers, in turn, once the legislative choice to confer them upon ESMA is made, do not pose a question of balance of powers between the Executive Board and the Board of Supervisors, but instead a question of legitimacy under Article 114 TFEU. Let me spend two words on this. I must say, first, that I read the CJEU ESMA short-selling judgment (C-270/12), which reinterpreted “Meroni” and “Romano” in light of the means of redress now afforded by Article 263 TFEU against the decisions adopted by the agencies, as a comfortable ‘green light’ (and I refer in particular to paragraphs 105-108). I am aware that a more restrictive reading of the judgment is also present in the debate and this makes very welcome further guidance by the Court which is likely to come in one or more of the several pending cases on Banco Popular.

But let me say that, to my mind, the time has come to accept that, at least where we are confronted with regulated industries as it is the case with financial markets, delegation to an EU agency of the technical task of administering and enforcing Level 1 sectoral regulation (and completing it via regulatory and implementing rules) is sound policy in compliance with the Treaties and such delegated discretion, being placed in a highly regulated context, is by definition circumscribed by sufficient criteria which per se prevent it from transcending into policy. Being clear on this would also help the co-legislators in focussing on what really matters, namely core policy choices (there should be a halt to rulemaking in the field, where, as Karel Lannoo (2018) noted, it increased by a factor of at least 12 over the last 10 years). It is now time, not only in the US, for proportionate simplification, to take account of the fact that the ESAs are fully accountable to the Parliament, the Council and the Commission and are subject to a comprehensive system of quasi-judicial and judicial review which has proven stringent.

Going forward

I would like to conclude these legal remarks with a couple of final observations on ESMA’s new tasks under Articles 4-9 of the Omnibus Regulation and on funding. In principle, it seems to me that extending ESMA’s direct supervision on newly regulated market players (such as benchmark administrators or cross-border data reporting service providers) is fully in line with the approach already adopted in the past with CRAs and trade repositories. At the same time, this will leverage ESMA’s role as the central data repository and financial data supervisor, in preparation for all future challenges posed by FinTech supervision.

At the same time, also direct supervision of numbered cross-border prospectuses and special funds (EuSEF, EuVECA and ELTIF), whilst apparently duplicating supervision of the AIFMs (both from ESMA and the NCAs depending on the class of AIFs implicated), would also push the system of supervisors towards more day-to-day coordination as occurs, successfully, within the SSM. This impinges also on funding. My concern is that the reform is pragmatic but timid. Investing more in the ESAs, and in ESMA in the first place, is a necessity in the new and very challenging FinTech context in which the supervisory system is bound to live; the system must rapidly evolve into a data-intensive, new-generation tech-authority (SupTech). And this requires appropriate funding at every level of the hub and spokes, to take the ESAs, in due time, to their final destination and in the meanwhile maintaining the right direction of travel.

ECMI Commentaries offer views and insights into developments affecting capital markets in Europe. They are produced by specialists associated with the European Capital Markets Institute, which is managed and staffed by CEPS. Unless otherwise indicated, the views expressed are attributable only to the authors and not to any institution with which they are associated, and nor do they necessarily represent the views of the ECMI.


D’Ambrosio, Raffaele (2018), “The maintenance of the structural and functional parallelism among ESAs”, Presentation, CIRSF Annual International Conference 2018 – Lisbon, June 6.

De Larosière, Jacques (2009), Report of the High-Level Group on Financial Supervision in the EU, Brussels, 25 February.

Juncker, Jean-Claude et al. (2015), “Five Presidents report: Completing Europe’s Economic and Monetary Union”, 22 June.

Lannoo, Karel (2018), “MiFID II and the new market conduct rules for financial intermediaries: Will complexity bring transparency?”, Policy Brief, ECMI, May.

Moloney, Niamh (2014), “EU Securities and Financial Markets Regulation”, 3rd edition, Oxford EU Law Library.

[1] This concerns the ECB letters of end March 2017 and the ECB position on AML and corporate governance of May 2018 and not the NPL addendum.

[2] See the Joint statement  of ESMA and EBA Statement of the EBA and ESMA on the treatment of retail holdings of debt financial instruments subject to the BRRD, 30 May 2018.