Central Counterparties are Too Big for the European Securities and Markets Authority (Alone): Constructive Critique of the 2019 CCP Supervision Regulation

Once seen as the solution to the systemic risk of over-the-counter derivatives, central counterparties (CCPs) have become a source of concern for the financial stability of the European Union (EU). Distress of a single CCP might be transferred to clearing members and others CCPs, thereby endangering EU monetary and fiscal stability. However, the European Market Infrastructure Regulation (EMIR) has left fiscal responsibility as well as day-to-day supervision of EU-based CCPs (EU-CCPs) at the national level, albeit cross-border colleges of authorities can overturn key decisions of national supervisors. The Commission concluded that these supervisory arrangements needed to be reformed because they hamper the development of a single coherent EU approach to CCP supervision. Although the Commission had proposed to centralise EU-CCP supervision in the European Securities and Markets Authority (ESMA), the final 2019 CCP Supervision Regulation seeks to improve EMIR’s ineffective collegial framework by establishing an internal CCP Supervisory Committee within ESMA. Nevertheless, the continuing lack of loss mutualisation would keep on hindering the development of a single approach to EU-CCP supervision. Moreover, the new regulation adds another layer of complexity to the EMIR framework and clashes with the European monetary constitution. If an agreement on loss mutualisation was reached, EU-CCP supervision should be centralised in two EU bodies along objective lines: while the European Central Bank should be tasked with EU-CCP prudential supervision, ESMA should be responsible for EU-CCP conduct of business supervision. This framework would improve EU-CCP supervision and fit in with the current EU governance at once.


Introduction
Marx and Engels seemed to be vindicated by the dynamics of the over-the-counter (OTC) derivatives 1 market during the Global Financial Crisis. In their Manifesto, they compared capitalism to 'the sorcerer who is no longer able to control the powers of the netherworld whom he has called up by his spells'. 2 Indeed, the opaque network of credit derivatives which concentrated systemic credit risk into a few interconnected institutions was seen by policymakers as one of the main propellants of the Global Financial Crisis. 3 Still, the leaders of the Group of Twenty (G20) did not despond. In fact, in 2009 they pledged to introduce a clearing obligation for standardised OTC derivatives in order to contain credit risk within Central Counterparties (CCPs). 4 Since then, the volume of derivatives transactions cleared by CCPs has significantly increased worldwide. 5 However, the growing importance of these financial market infrastructures 6 (FMIs) has meant that they have become a source of systemic risk themselves. 7 CCPs are now seen as too big to fail. 8 Accordingly, there is an ongoing effort to develop an effective CCP recovery and resolution regime to avoid enormous bailouts. 9 Furthermore, the Brexit saga has reignited the debate over third-country CCP regimes and locations. 10 Against this backdrop, this article focuses on a specific topic: the arrangements for the supervision of European Union-based CCPs (EU-CCPs). In particular, this paper will conduct a legal-institutional analysis of the 2019 CCP Supervision Regulation, 11 which was approved after two years of intense negotiations. This reform seeks to improve the supervisory arrangements contained in the 2012 European Market Infrastructure Regulation 12 (2012 EMIR). To this end, in 2017 the Commission proposed to centralise certain supervisory tasks in a new CCP Executive Session within the Board of Supervisors of the European Securities and Markets Authority (ESMA). 13 However, co-legislators have significantly watered down such a proposal by the Commission: instead of the decision-making CCP Executive Session, a merely internal CCP Supervisory Committee is going to be established within ESMA to foster supervisory convergence and cooperation between the increasing number of authorities involved in EU-CCP oversight. It is the intention of co-legislators that the 2019 CCP Supervision Regulation would secure swift decision-making, clarify areas of competences between authorities, and lead to a coherent and consistent supervisory approach to CCP supervision across the EU.
This article will argue that the 2019 CCP Supervision Regulation would not achieve such praiseworthy objectives. The development of a truly European approach to CCP supervision would be undermined by the continuing absence of loss mutualisation between Member States. Moreover, the new supervisory architecture would add another layer of complexity and confusion to the previous framework. If an agreement on loss mutualisation could be reached, this paper suggests instead to centralise EU-CCP supervision in two EU bodies along objective lines, at least in the euro area. While EU-CCP prudential supervision should be conferred upon the European Central Bank (ECB) within the Single Supervisory Mechanism (SSM), ESMA's Board of Supervisors should be responsible for EU-CCP conduct of business supervision. As opposed to the 2019 CCP Supervision Regulation (as well as to the initial proposal of the Commission), this framework would improve EU-CCP supervision and fit in with the current EU supervisory architecture at once. This paper belongs to the legal-institutional literature which examines the governance of EU financial supervision after the crisis-era reforms. 14 Nevertheless, the analysis draws heavily upon the related political economy literature to explain the dynamics behind EU institutional reforms. 15 Moreover, this article relies on the legal and economic literature on CCPs to explain their inner workings and risks. 16 This paper is organised as follows: after this brief introduction, Sect. 2 considers how the presence or the absence of loss mutualisation between Member States has shaped the current EU supervisory architecture. Section 3 examines how fiscal considerations also informed the peculiar supervisory arrangements of the 2012 EMIR regime. Section 4 presents the shortcomings of such arrangements, and their reform as proposed by the Commission and then approved by co-legislators. Section 5 highlights three weaknesses of the 2019 CCP Supervision Regulation and advocates instead a centralised and objective-based framework for EU-CCP supervision. Section 6 concludes.

Setting the Scene
The optimal degree of centralisation for EU financial supervision has been at the centre of a long debate. On the one hand, it has been argued that decentralising supervision at Member State level would support regulatory innovation and local expertise as well as avoid the fiscal, political, legal, and accountability sensitivities associated with centralisation. 17 On the other hand, pan-EU supervisory authorities would bring the benefits of economies of scale, reduce regulatory capture and home biases. 18 Presently, two different types of financial supervisory arrangements coexist at the EU level. 19 The first one is the largely decentralised and sector-based European System of Financial Supervision (ESFS) recommended by the de Larosière Group. 20 This is a pan-EU network which comprises 21 : national competent authorities (NCAs), the European Systemic Risk Board 22 (ESRB), three sectoral European Supervisory Authorities 23 (ESAs) and their Joint Committee. The ESFS is the dominant model in EU financial markets supervision. 24 Even though the sectoral ESA-ESMA-is the pan-EU supervisor of trade repositories 25 and credit rating agencies, 26 the 'home country control' principle reigns supreme for the rest of financial markets supervision: NCAs of the Member State where intermediaries have obtained their respective authorisation remain primarily responsible for their dayto-day supervision. 27 However, ESMA has incisive quasi-regulatory and supervisory powers to foster supervisory convergence and cooperation between NCAs. 28 In principle, the decentralised ESFS also operates in the EU banking sector. 29 However, euro area countries have agreed to integrate further their banking supervisory arrangements by adopting a second framework. In particular, they have centralised banking prudential supervision 30 in the ECB within the SSM. 31 Nevertheless, conduct of business supervision remains a national competence. 32 The coexistence of these divergent supervisory arrangements has its origins in fiscal concerns.

From Maastricht to the SSM
The in-depth reason for this divergence is the asymmetry of the Economic and Monetary Union (EMU) agreed in Maastricht. 33 A perfect Monetary Union governed by an independent central bank was established. 34 However, EMU was explicitly designed to avoid fiscal transfers between Member States. 35 Accordingly, monetary financing and sovereign bailouts were (and still are) prohibited. 36 However, the Economic Union lacked credible mechanisms to deal with asymmetric shocks and inconsistent national economic policies. 37 The dangers of this system were laid bare during the sovereign debt crisis. When the crisis hit, the euro area did not have a set of financing options to rapidly sever the sovereign debt-bank 'vicious circle': 38 32 Colaert (2015), p 1592. As noted in ibid., p 1581, the International Organization of Securities Commissions (1990), p 5 has defined business of conduct regulation as 'principles of conduct which should govern the activities of financial services firms in protecting the interest of their customers and the integrity of the market'. 33 Schimmelfennig (2016), pp 490-496. 34 Lastra (2015), pp 69, 240-242. In particular, the Maastricht Treaty has established an European System of Central Banks (ESCB) comprising the national central banks of all Member States and the ECB, whose monetary decision-making bodies (the Executive Board, the Governing Council, and the General Council) govern the ESCB. However, the euro has not yet been adopted by all Member States. Therefore, according to the Treaty on the Functioning of the EU (TFEU), Art. 282(1) the definition and the implementation of the monetary policy of the euro area is conferred upon a smaller group: the Eurosystem, which comprises only the ECB and the national central banks of the Member States which have adopted the euro. 35 Dyson (2014), pp 612-613. 36 TFEU,Arts. 123,125,respectively. 37 Amtenbrink and De Haan (2003), pp 1079-1095; Dyson (2014), pp 630-631. 38 Merler and Pisani-Ferry (2012), p 11. 30 As noted by the de Larosière Group (2009), p 38 prudential supervision comprises two dimensions: on the one hand, micro-prudential supervision aims at 'supervis [ing] and limit[ing] the distress of individual financial institutions' via a vast array of controls on their capital and liquidity positions, business model, and governance; on the other hand, macro-prudential supervision is tasked with 'limit[ing] distress of the financial system as a whole in order to protect the overall economy from significant losses in real output'. While the two dimensions are strictly intertwined, in this paper prudential supervision refers to micro-prudential supervision where not specified otherwise. 31 Regulation (EU) No. 1024 OJ L 287/63 (2013 SSM Regulation). The SSM is an institutional framework which allocates prudential supervision of euro area banking groups according to their systemic importance as determined by proxies (non-euro area countries could ask to join). The largest banking groups are directly supervised by the ECB with the assistance of NCAs. On the other hand, NCAs remain responsible for the day-to-day prudential supervision of 'less significant' credit institutions, albeit under the oversight of the ECB. Two key supervisory tasks are centralised in the ECB regardless of the significance of the intermediary: the granting of a banking licence and the authorisation of a proposed acquisition of qualified holdings. Non-euro area Member States can request to join the SSM. The Supervisory Board is the de facto head of SSM decision-making, see infra n. 356; Moloney (2014b); Pizzolla (2018). the deterioration of public finances of certain euro area countries negatively affected the balance sheets of their banks, and vice versa. 39 In turn, this self-propelled circle severely hindered the monetary policy of the Eurosystem. 40 As a response, the Eurosystem had to employ unconventional monetary measures 41 to save the Single Currency, thereby becoming the de facto lender of last resort of the euro area. 42 Meanwhile, unprepared governments had to arrange temporary and legally contentious rescue packages. 43 Member States finally concluded an international treaty establishing a permanent 'fiscal backstop' to mutualise losses: the European Stability Mechanism 44 (ESM). Immediately after being established, the ESM lent € 39.5 billion to Spain to recapitalise its banking system. 45 Hence, the sovereign debt-bank circle was temporarily severed through significant fiscal transfers and the adoption (and the announcement 46 ) of unconventional monetary measures. However, these extraordinary responses shattered the EMU's macroeconomic fiscal and monetary foundations. 47 Therefore, it was clear that the EMU needed long-term reforms. 48 One of these has been the centralisation of banking prudential supervision in the ECB within the SSM. 49 The establishment of the SSM was considered as the sine qua non for future fiscal transfers via the ESM by Member States such as Germany, the Netherlands, and Finland. 50 From the perspective of those countries, supervisory centralisation was needed because the decentralised model of the ESFS was seen as an insufficient guarantee against the moral hazard of recipient Member States. 51

Between Meroni and NCAs: ESMA
EU financial markets supervision did not experience a level of centralisation similar to the SSM. From a purely legal perspective, the ascent of a single European financial markets supervisor has been blocked by constitutional limitations. The establishment of the SSM could rely on a ready-made EU institution and on TFEU Article 127(6). 52 This provision empowers the Council to 39 Bank for International Settlements (2011), pp 29-30. 40 ECB (2012a), pp 70-73. 41 Bini Smaghi (2009), p 3 notes that while central banks have traditionally governed the money supply indirectly by fixing official interest rates, unconventional monetary measures can be broadly described as 'those policies that directly target the cost and availability of external finance to banks, households and non-financial companies' such as quantitative easings. For an overview of how the ECB's monetary policy changed during the crisis, see Constâncio (2018), pp 3-6. 42 De Grauwe (2016), pp 173-175, 210-213. 43 See, for example, Ruffert (2011), pp 1778-1787. 44 Lastra and Louis (2013), pp 103-106. 45 Howarth and Quaglia (2016), p 172. 46 As the Outright Monetary Transactions, see ECB (2012b). 47 See the seminal work of Tuori and Tuori (2014), especially pp 181-192. 48 Outlined in Van Rompuy (2012). 49 Ibid., p 4. 50 Howarth and Quaglia (2016), p 90. 51 Dyson (2014), p 367. 52 Glöckler et al. (2017Glöckler et al. ( ), pp 1147Glöckler et al. ( -1148 confer specific prudential supervisory powers over credit and financial institutions (apart from insurance undertakings) upon the ECB. In the financial markets, the only feasible solution was to create a new EU agency-ESMA-on the basis of TFEU Article 114. 53 This provision enables co-legislators to take measures to foster the integration of the Internal Market, including by establishing agencies. 54 However, such 'agentification' entails the restrictions of the Meroni case law, 55 which limits the extent to which executive powers can be delegated to EU agencies. 56 In one of its earliest rulings, the Court of Justice was asked to assess whether the High Authority of the European Coal and Steel Community could legally delegate some of its treaty-based powers to manage the ferrous-scrap market to two 'Brussels agencies' (formally two cooperative companies incorporated under Belgian law). The powers of such agencies were contested by Meroni-a steel company required to make contributions to the ferrous-scrap equalisation mechanism-because, inter alia, they had not been envisaged by the institutional framework of the Treaty of Paris. The Court stated that the allocation of powers enshrined in the Treaty had to be considered a fundamental guarantee for the undertakings dealing with the then Community, which had been founded on a constitutional 'balance of powers'. Accordingly, treaty-based bodies cannot delegate their executive powers (provided that they have them in the first place) unless they can be 'subject to strict review in light of objective criteria determined by the delegating authority'. 57 Consequently, Meroni prohibits the delegation of discretionary powers which 'make possible the execution of actual economic policy'. 58 While doubts have been expressed over the scope-and the very meaning-of such a dated ruling under contemporary legislation, Meroni has exercised a decisive influence over EU agency-design ever since. 59 Nevertheless, the de Larosière Group expressly recommended that ESAs should have been empowered to challenge national supervisory decisions. 60 Therefore, ESMA has been endowed with potentially discretionary powers over NCAs and financial market participants: for example, ESMA can overrule NCAs should a breach of EU law occur, should the Council determine the existence of an emergency situation, or when NCAs irreconcilably disagree on the application of EU law. 61 However, these powers are in tension with the tight constraints of Meroni. 62 53 Moloney (2014a), pp 960-964. 54 As stated by the Court of Justice with regard to the establishment of the then European Network and Information Security Agency (ENISA) in Case C-217/04 United Kingdom v. European Parliament andCouncil, ECLI:EU:C:2006:279. 55 Case 9-56 Meroni v. High Authority, ECLI:EU:C:1958:7. 56 Lenaerts (1993), pp 40-49. 57 Case 9-56 Meroni v. High Authority, ECLI:EU:C:1958:7, at p 152. 58 Ibid. 59 See e.g. Chamon (2011);Griller and Orator (2010). 60 De Larosière Group (2009), p 56. 61 2010 ESMA Regulation, Arts. 17, 18, 19, respectively. If NCAs fail to abide by ESMA's decisions, ESMA can ultimately take action vis-à-vis the regulatees. However, such powers are designed as an extrema ratio and subject to strict conditionality, tight procedural constraints, and require the involvement of EU institutions or specific empowering legislative acts in order to respect the 'Meroni doctrine'. 62 Moloney (2014a), p 995.
To solve this conundrum, ESMA has been equipped with exceptional governance arrangements. 63 While in a typical EU agency the technocratic component is separated from the oversight and representative internal bodies, 64 ESMA's governance arrangements assign a decisive role to NCAs in ESMA's governing body, the Board of Supervisors. 65 Since they are the only voting members of the Board of Supervisors, NCAs control and supervise ESMA's decision-making process. As a consequence, ESMA's effectiveness and legitimacy are built upon NCAs' technical capacity. 66 Nonetheless, ESMA's supervisory powers have still had to be subject to highlydetailed conditions and procedures to respect the 'Meroni doctrine'. 67 As ESMA has lamented, 68 these constraints undermine the effectiveness of its action. 69 Moreover, Meroni renders ESMA's powers prone to legal challenges. 70 For example, the United Kingdom asked the Court of Justice of the EU to assess whether ESMA's short selling intervention powers were compatible, inter alia, with Meroni case law. 71 In particular, if NCAs fail to act adequately against extreme cross-border threats to the functioning and the integrity of EU financial markets, the financial stability of the Union, or a part thereof, the 2012 Short Selling Regulation empowers ESMA to restrict, prohibit, and order to disclose or notify short positions instead of NCAs. 72 Before acting, ESMA should consult with the ESRB and other relevant authorities 73 as well as duly take into account whether its intervention addresses the threat, negatively affects the efficiency of the markets, and creates the risk of regulatory arbitrage. 74 The reasoned decision to impose or renew such measures should be notified to NCAs with at least one day's notice (if possible) 75 and published on ESMA's website. 76 ESMA should review the decision at least every 3 months. If measures are not renewed, they expire automatically. 77 Nonetheless, the United Kingdom argued that such short selling intervention powers could allow ESMA to take discretionary decisions with relevant economic and financial implications against the 63 As noted in Chiti (2009Chiti ( ), pp 1427Chiti ( -1428Everson (2012), pp 17-18. 64 Bergström (2015). 72 2012 Short Selling Regulation, Art. 28(1-2). According to Art. 28(11), ESMA's measures prevail over those-if any-previously taken by NCAs. 73 Ibid., Art. 28(4). 74 Ibid., Art. 28(3). 75 Ibid., Art. 28(5-6). According to Art. 28(8), ESMA must also notify NCAs of the measures effectively taken. 76 Ibid., Art. 28(7). According to Art. 28(9), measures take effect at the moment of their publication on ESMA's website unless specified otherwise. 77 Ibid., Art. 28(10). principles of the 'Meroni doctrine'. The Court disagreed: given the procedural and substantive constraints attached to such temporary powers, 78 the residual margins of discretion left to ESMA have been 'precisely delineated and amenable to judicial review in light of the objectives established by the delegating authority' 79 in compliance with the 'Meroni doctrine'.

Piercing the Veil of Meroni
While the Court of Justice of the EU ultimately upheld the legality of ESMA's powers, it is worth noting that the United Kingdom itself started to question the short selling intervention powers on the basis of Meroni only at a late stage of the negotiations. 80 This fact reveals the in-depth reason why ESMA was not designed to be the single supervisor of EU financial markets: the concerns of many Member States over the split between supervisory competences and fiscal responsibility. 81 These concerns were often hidden behind Meroni. 82 During ESA negotiations, loss mutualisation was opposed by a wide and diverse coalition led by the United Kingdom. 83 This coalition included countries such as France, Germany, Ireland, and Luxembourg which years later consented (if not demanded) to establish the SSM. 84 In this alliance there were also many central and eastern European States, which were firmly against any form of supervisory centralisation. 85 Accordingly, ESMA's supervisory powers are both de jure and de facto limited when the fiscal interests of Member States come into play: a 'fiscal safeguard' limits ESMA's powers to act in emergency situations and to conclude binding mediations between NCAs. 86 Moreover, sovereign debt and sovereign credit default swaps 87 (CDSs) are not covered by ESMA's short selling intervention powers. 88 Similarly, ESMA's product intervention powers to restrict or prohibit financial 78 The conditions and criteria of the 2012 Short Selling Regulation are even more specified in Commission Delegated Regulation ( 87 CDSs are credit derivatives functionally similar to insurance. The 2012 Short Selling Regulation, Art. 2(1)(c) defines a CDS as 'a derivative contract in which one party pays a fee to another party in return for a payment or other benefit in the case of a credit event relating to a reference entity and of any other default, relating to that derivative contract, which has a similar economic effect'. 88 Ibid., Art. 28(1)(b). activities, practices, financial instruments, and entering derivatives positions 89 entail no significant fiscal risk. 90 Finally, the regulatees subject to ESMA's direct supervision-credit rating agencies and trade repositories-pose limited fiscal risk. 91 Therefore, the current landscape of EU financial supervision has been heavily influenced by the presence or absence of loss mutualisation. 92 The establishment of the ESM ultimately led to the centralised SSM and its single 'supervisory manual'. 93 On the other hand, given the continuing absence of loss mutualisation between Member States, NCAs have remained responsible for financial markets supervision within the decentralised ESFS. Against this backdrop, the supervisory arrangements of the 2012 EMIR regime are somewhat exceptional. Central clearing poses unique risks to fiscal and monetary stability. Therefore, the decentralised model of the ESFS has had to be adapted.

CCPs as a Solution: Mutualisation of Credit Risk
The Global Financial Crisis revealed that systemically important financial institutions were heavily exposed to each other through an opaque network of OTC derivatives contracts, notably CDSs. 94 The concentration of systemic credit risk in a limited number of interlinked institutions had dramatic consequences such as the bailout of AIG. 95 The failure of one of the major players of the CDS market would have jeopardised the risk management of all its counterparties with unforeseeable knock-on effects. 96 On the other hand, the $9 trillion OTC derivatives portfolio of Lehman Brothers was smoothly liquidated by the CCP LCH.Clearnet. 97 Indeed, CCPs are FMIs whose main function is to centralise the management of the credit risk of their clients, 98 the clearing members. 99 Pirrong (2009), p 2. 99 2012 EMIR, Art. 2(1)(14) defines a clearing member as 'an undertaking which participates in a CCP and which is responsible for discharging the financial obligations arising from that participation'. In turn, clearing members might also clear on behalf of their own clients, which are defined by 2012 EMIR, Art. 2(1)(15) as every 'undertaking with a contractual relationship with a clearing member of a CCP which enables that undertaking to clear its transactions with that CCP'. clearing members are novated 100 in favour of the CCP, which thus becomes 'the buyer to every seller and the seller to every buyer'. 101 Consequently, CCPs stand at the centre of the cleared transactions between their clearing members. This makes multilateral netting possible, 102 which reduces credit and operational risk. 103 Moreover, the concentration of trade in CCPs imposes trade standardisation, thereby enhancing market liquidity and transparency. 104 Crucially, if a member has defaulted, the CCP would be able to guarantee the performance of open contracts for the benefit of the other clearing members so as to prevent cascades of defaults. 105 However, this is possible only if CCPs do not themselves fail by effectively absorbing the shock of defaults of clearing members. Therefore, CCPs must have robust default management processes. These can be described as a set of legal devices, 106 which ultimately allocates uncovered losses among CCPs and clearing members via a mutually financed default fund. 107 In particular, CCPs do not ordinarily assume market risk because they run a 'matched book': following novation, a CCP's position with a contract 'seller' is offset by an opposite position with a contract 'buyer', and vice versa. 108 However, if a member defaulted, a CCP would need to close out its now unmatched positions. 109 In order to return to a matched book, the CCP would firstly try to liquidate the defaulter's portfolio through an auction between the other clearing members. 110 However, if this was not sufficient or possible, 111 losses would be allocated according to the 'default waterfall'. 112 This is the pool of resources available to CCPs to progressively share losses with clearing members. The first step of the default waterfall is to gather the collateral and the contribution to the default fund of the defaulter. 113 To mitigate the moral hazard of the CCP's shareholders, the second step is their 'skin in the game', i.e. part of the CCP's own capital. 114 If the skin in the game is not enough, losses are mutualised among non-defaulting clearing members through their contribution to the default fund. 115 As an extrema ratio before its failure, CCPs might call additional resources from the remaining members or tear-up contracts. 116 The mutualisation of credit risk via CCPs seemed to be an appropriate response to the systemic dangers of the OTC derivatives market. 117 In 2009, the G20 made the pledge that all standardised OTC derivatives should have been cleared through CCPs by the end of 2012. 118 Since then, this commitment has progressively been implemented worldwide. 119 In the EU, this so-called clearing obligation was introduced by the 2012 EMIR. 120

CCPs as a Danger: Mutualisation of Systemic Risk
The introduction of the clearing obligation and its progressive enlargement 121 have had transformative effects on important segments of the EU OTC derivatives market. For example, only 25% of interest rate derivatives were centrally cleared in the EU before the 2012 EMIR entered into force. 122 This figure stood at 63% at the end of 2018. 123 Most of these contracts are cleared by EU-CCPs. 124 These data demonstrate the success of the reform. However, while the amount of contracts cleared has steadily risen, the number of EU-CCPs is stuck at thirteen. 125 Moreover, not all asset classes are cleared by all EU-CCPs: for example, just two EU-CCPs are authorised to clear credit derivatives such as CDSs. 126 Furthermore, EU-CCPs are increasingly interconnected with one another either directly 113 2012 EMIR, Art. 45(1-2). 114 Ibid., Art. 45(4). 115 Ibid., Art. 45(3). 116 Elliott (2013), pp 8-9. 117 Ferrarini and Saguato (2015), pp 583-587. 118 G20 (2009), para. 13. 119 For the most recent report on the implementation of the clearing obligation, see Financial Stability Board (2019b), pp 10-12, 25-28. 120 2012 EMIR, Art. 4. 121 Ibid., Art. 5 states that the Commission issues the regulatory technical standards determining the classes of OTC derivatives subject to the EMIR clearing obligation. Such draft regulatory technical standards are developed by ESMA (after consulting the public and the ESRB) either on its own initiative or following a new grant or extension of EU-CCP authorisation. For the full list of the derivatives to the EMIR clearing obligation, see ESMA (2020). 122 European Commission (2017a), p 16. 123 ESMA (2019a), p 13. 124 Ibid., p 21. 125 If the withdrawal of the UK from the EU is taken into account, see ESMA (2019b), pp 1-2. 126 Ibid., pp 4, 6. via interoperability arrangements 127 or indirectly via their clearing members. 128 In particular, only a limited number of systemically important institutions can meet the strict requirements to be a clearing member. 129 For example, the Commission noted that '24 globally systemically important banks (G-SIBs) are members of [the German-CCP] Eurex, BNP Paribas is a member of at least five EU CCPs'. 130 The fact that systemic credit risk is concentrated in a limited number of interlinked CCPs has become a source of concern for EU macro-prudential stability. Financial and liquidity distress experienced by one single EU-CCP could be transferred to others through interoperability arrangements 131 and common clearing members. These could be asked to mutualise losses via the default waterfall, thereby deteriorating their financial position. 132 This would happen when they can 'least afford them' because the default waterfall's exceptional solidity would be endangered only when a systemic crisis is already occurring. 133 Indeed, a 'CCP-bank nexus' could be triggered. 134 The outright failure of a CCP 'would deprive market participants of certain basic functions, such as trade processing, thereby entailing the shutdown of entire markets'. 135 Moreover, disruptions of clearing services could hinder the 'smooth functioning of euro area payment systems. This could lead to increased demand for central bank liquidity and possible challenges in implementing the Eurosystem's single monetary policy. In addition, such disturbances can impair the functioning of financial market segments that are key to the transmission of monetary policy'. 136 Accordingly, there is a growing awareness that CCPs cannot fail. 137 Since a CCP resolution framework is still under development, 138 public funds would be employed to rescue failing CCPs. 139 In the past, governments and Central Banks already had 127 ESMA (2016a), pp 11-12. 2012 EMIR, Art. 2(1)(12) defines an interoperability arrangement as an 'arrangement between two or more CCPs that involves a cross-system execution of transactions'. As noted by ESRB (2019), p 6, interoperability arrangements allow clearing members to have access to more netting opportunities without seeking membership of multiple CCPs. 128 Alfranseder et al. (2018), pp 29-33. 129 European Commission (2017a), p 22. The first line of defence of CCPs is limiting their membership to those financial institutions that can meet stringent capital, liquidity, and operational requirements. E.g. see the requirements of Eurex Clearing (2019). 130 Ibid., p 21. 131 ESRB (2019), pp 25-26. 132 Domanski et al. (2015), pp 68-69. 133 Pirrong (2013) Tucker (2014), p 14; Gracie (2015), p 2; European Commission (2016), pp 4-5. 138 Financial Stability Board (2019a), pp 6-8. Resolution can be defined as special administrative procedures-alternative to normal insolvency proceedings-designed to avoid the externalities stemming from the bankruptcy and bail-outs of financial institutions. Accordingly, authorities have access to an array of tools (e.g. the transfer of assets and bail-in) to ensure the protection of the critical economic functions of distressed financial institutions without costs for the taxpayers by allocating losses to shareholders and uninsured creditors. See Financial Stability Board (2014). 139 Singh (2011), pp 9-10. to intervene to rescue CCPs. During the crash of 1987, the Federal Reserve had to provide exceptional liquidity support to the US-CCP CME to avoid its failure. 140 In the same year, the Hong Kong Government had to employ HKD 2 billion to try to avoid the failure of a Hong Kong CCP. 141 Today figures are even more appalling: in 2018 the total notional amount of interest rate derivatives cleared by LCH Group's SwapClear surpassed USD 1 quadrillion. 142 However, the EU-CCP home Member State would not be the only one to bear the costs stemming from CCP bail-outs. Indeed, CCPs' losses would be mutualised among clearing members via the default waterfall. In turn, their respective Member States would 'need to provide these participants with financial support if they are faced with large payment obligations due to CCP losses'. 143 Unfortunately, CCP failures are not unheard of. 144 These FMIs face a wide range of risks, 145 notably credit and liquidity risks. 146 Indeed, CCP solvency can be threatened by defaults of relatively small clearing members. 147 As recently as 2018, an EU-CCP needed to be bailed-in by its members after the failure of an individual to meet its margin calls. 148

The Supervisory Architecture of the 2012 EMIR Regime
CCP failures date back to a time when CCPs were not so systematically important as today. When the clearing obligation was introduced, the EU decided to move away from the prior mostly self-regulatory regime. 149 EU-CCPs are now subject to a granular and ever-growing EU rulebook. The 2012 EMIR regime implements and expands the standards agreed in international forums. 150 In turn, EMIR's prudential, organisational, and conduct of business requirements 151 are specified by a vast set of administrative rules 152 and ESMA's soft law. 153 Nevertheless, EMIR has left fiscal responsibility for EU-CCPs at the national level. 154 Member States opposed the suggestion of the de Larosière Group to 140 Bernanke (1990), p 148. 141 Davison (1988), p 1. 142 LCH (2018), p 5. 143 Coeuré (2018), p 3. 144 Hills et al. (1999), pp 129-130; Bignon and Vuillemey (2018 centralise EU-CCP supervision 155 because of the above-mentioned concerns over the split between supervisory duties and fiscal responsibility. 156 However, the decentralised model of the ESFS has been adapted to take into account the systemic risk posed by CCPs. 157 Accordingly, EMIR has mandated collegial procedures for some important supervisory decisions: the granting and extension of authorisation, and the approval of interoperability arrangements. 158 In particular, an enhanced form of the college of supervisors must be established by the home NCA 159 within 30 days from the application for authorisation. 160 In addition to the chairing home of the NCA, colleges originally comprised 161 : ESMA; NCAs of the clearing members established in the three Member States with the largest contributions to the default fund; NCAs of the trading venues served by the EU-CCP; NCAs of other EU-CCPs with which interoperability arrangements have been established; NCAs of linked central securities depositories; the members of the ESCB responsible for the applicant and the interlinked EU-CCPs; and the central banks of issue (CBIs) of the most relevant Union currencies of the financial instruments cleared. The establishment and the functioning of the college are based on written agreements between its members. 162 Unlike other fields of EU financial markets supervision where the home country control principle reigns supreme, 163 the home NCA cannot unilaterally grant or refuse the authorisation (and its extension), but must wait for the college's opinion. 164 To this end, the college must receive a report on the risks posed by the applicant from the home NCA. 165 Detailed voting thresholds are set to take into account the cross-border implications of the decision. 166 If the college's opinion matches with the conclusion of the report, the decision of the home NCA is adopted. 167 However, if college members (excluding the authorities of the home State) jointly issue a negative opinion, the home NCA cannot grant the authorisation. The matter could only be referred to ESMA for binding mediation. 168 If only two-thirds of the members express a negative opinion, any authority concerned can ask ESMA to conclude 155 De Larosière Group (2009) 165 Ibid., Art. 19(1). 166 Ibid.,Art. 19(3). This provision has been amended by the 2019 CCP Supervision Regulation to take into account the establishment of the new CCP Supervisory Committee, and the dual role of the ECB as supervisor and CBI, see infra, Sect. 4.3. 167 Ibid., Art. 17(4). However, the home NCA could still refuse to grant the authorisation if it does not agree with the college's positive assessment. 168 Ibid. binding mediation. 169 If the home NCA does not comply with these EMIR requirements (and more generally with EU law), ESMA is empowered to take direct action against NCAs and financial market participants. 170 Mutatis mutandis, the same procedure applies for the approval of interoperability arrangements. 171 Once the CCP is authorised, college members must keep on exchanging information and cooperating closely, especially in emergency situations. 172 For example, any member can ask the home NCA to review whether an EU-CCP remains in compliance with authorisation requirements. 173 Similarly, the home NCA must take into account the reservations of the other college members before withdrawing the authorisation. 174 College members might also delegate tasks to others. 175 For example, one NCA delegated ESMA to oversee the compliance of its CCP with EMIR prudential requirements until the end of 2017. 176 Moreover, colleges must issue opinions on significant changes to CCP risk models, which must also be validated by ESMA. 177 Indeed, the 2012 EMIR regime had assigned a significant role to ESMA even before the recent reform was approved. ESMA must coordinate colleges and NCAs to build a single 'supervisory culture and consistent supervisory practices'. 178 Accordingly, ESMA has been a (non-voting 179 ) member of all EMIR colleges, whose workings it must facilitate. 180 Moreover, ESMA has adopted a set of supervisory convergence measures, which include opinions on ECB's role in colleges 181 and on common indicators for the exercise of certain NCAs' powers. 182 Furthermore, EMIR requires ESMA to conduct pan-EU CCP stress tests and (at least annually) NCA peer reviews. 183 Nevertheless, EU-CCP supervision has mostly remained in the hands of home NCAs. 184 Moreover, the 2012 EMIR framework has not harmonised supervisory and enforcement powers: NCAs must only have supervisory and investigatory powers that are necessary for exercising their functions. 185 As ESMA's peer reviews 169 Ibid. 170 Ibid., Art. 17(5). 171 Ibid., Art. 54. For the recent extension of EMIR collegial decision-making see infra, Sect. 4.3. 172 Ibid., Arts. 18(4), 23-24. 173 Ibid., Art. 20(4-5). 174 Ibid., Art. 20(6). 175 Ibid., Art. 18 (4) (6). 184 Ibid., Recital 51. 185 Ibid., Art. 22(2). Moreover, appropriate administrative measures must be in place for non-compliance with EMIR requirements (Art. 22 (3)).
have revealed, these supervisory arrangements have been suffering from significant shortcomings.

Cracks in the 2012 EMIR Supervisory Architecture
The supervisory arrangements of the 2012 EMIR regime have been the result of a political compromise. As seen above, the macro-prudential risks posed by CCPs have led to the creation of EMIR colleges. On the other hand, the decentralised model of the ESFS has been retained because fiscal responsibility has remained at the national level.
ESMA's peer reviews have shown that this hybrid supervisory architecture has been under strain. While home NCAs seem to take into account the inputs from other members during the authorisation process, afterwards colleges risk becoming mere forums for exchanging information and not an 'effective supervisory tool'. 186 Furthermore, over the years there have been cases where home NCAs did not share necessary information with other college members in a proactive and timely manner. 187 In turn, these reacted either by abstaining or by voting against the college's opinion. 188 In other instances, home NCAs did not consider it necessary to inform other NCAs of their decisions. 189 Moreover, colleges comprise a cumbersome number of authorities with different mandates: over twenty members in certain instances. 190 This fact has inevitably meant that there are different degrees of participation in college discussions. 191 Overall, home NCAs have retained their often idiosyncratic approach to supervision and enforcement. 192 In light of these peer reviews, the Commission concluded that the supervisory arrangements of the 2012 EMIR framework were inadequate to address the growing macro-prudential risk of EU-CCPs. Regardless of EMIR colleges and ESMA, NCAs' diverging supervisory approaches 'create risks of regulatory and supervisory arbitrage, jeopardising financial stability and allowing for unhealthy competition '. 193 Moreover, effective decision-making has been compromised by the large number of authorities which sit in the colleges. 194 Furthermore, the 2012 EMIR regime did not sufficiently reflect the risk of misalignments between CCP supervisory actions and monetary policy imperatives. In the colleges, central banks sit alongside supervisory authorities without a clear division of competences. 195 Such concerns were shared by the European Parliament and the Council. 196 However, co-legislators profoundly disagreed with the Commission on how to improve this sub-optimal state of affairs: while the former have decided to retain an EMIR college-based framework, the latter proposed to centralise EU-CCP supervision at the EU level.

The 2017 CCP Supervision Proposal
In the opinion of the Commission, the centralisation of EU-CCP supervision in ESMA would have significantly improved the 2012 EMIR framework as well as fostering the broader Capital Markets Union agenda 197 (CMU). The centralisation of EU-CCP supervision in ESMA would have streamlined decision-making, clarified competences between central banks and supervisory authorities, and led to a single approach to  In particular, a new CCP Executive Session would have been established within ESMA's Board of Supervisors. 199 This new body would have had five permanent members: 200 a voting Head and two voting Directors, a non-voting representative of the ECB, and a non-voting representative of the Commission. Moreover, the CCP Executive Session would have also included non-permanent, CCP-specific members: 201 one voting representative of home NCAs and one non-voting representative of CBIs. However, CBIs could have vetoed certain home NCA draft decisions on prudential requirements (e.g. liquidity risk controls and collateral requirements) on the basis of monetary stability concerns. 202 The CCP Executive Session would have substituted the Board of Supervisors in the tasks already conferred upon ESMA by the 2012 EMIR framework. 203 (2015) launched an action plan to build a deeper CMU and thus reduce the dependency of the European economy on bankbased finance. While initially the main focus of this initiative was regulatory reform (e.g. of the prospectus regime), later the European Commission (2017b), p 10 stated that strengthening EU supervisory arrangements was 'at the heart of efforts to build CMU given the central role that it plays in accelerating market integration and creating single market opportunities for financial entities and investors '. 198 European Commission (2017a), pp 57-59. 199 2017 CCP Supervision Proposal, ESMA Regulation, Art. 6(1a). 200 Ibid., Art. 44a(1)(a). 201 Ibid., Art. 44a(1)(b). 202 Ibid., EMIR, Arts. 21a(2), 21b. 203 Ibid., ESMA Regulation, Art. 44b(1)(c). representative). 204 Moreover, EMIR colleges would have been chaired by the Head of the CCP Executive Session instead of the home NCA. 205 Most importantly, home NCAs would have needed the consent of the CCP Executive Session to adopt decisions on: 206 granting, extending, and withdrawing CCP authorisation; capital requirements; a review and evaluation of the CCP's regulatory compliance; a review of qualified shareholders; outsourcing; a review of CCP risk models; the approval of interoperability arrangements; and access to the CCP and to trading venues. In these instances, consent would have been deemed as given, unless the CCP Executive Session had put forward objections or amendments to the draft decision of home NCAs. 207 If amendments had been proposed, the home NCA would have been bound by them. If the CCP Executive Session had objected, the draft decision would have been rejected. The home NCA could have made a final appeal against the decision to ESMA's Board of Supervisors. 208 The CCP Executive Session's decisions would have been adopted under a simple majority vote (the Head would have had the casting vote). 209 Moreover, the CCP Executive Session could have demanded specific supervisory actions from NCAs 210 as well as information directly from regulatees. 211 If home NCAs had not complied with the decisions of the CCP Executive Session, this body could have exercised direct supervisory powers towards financial market participants. 212 More generally, an amendment to the Proposal would have conferred the powers of the Board of Supervisors to conclude binding mediations and to take actions against NCAs for breaches of EU law on CCP matters upon the CCP Executive Session. 213 Therefore, the establishment of the CCP Executive Session would have centralised EU-CCP supervisory decision-making in a supranational body. If the 2017 CCP Supervision Proposal had been passed in its original form, the home country control principle would have been significantly weakened in key areas of EU-CCP supervision. However, the final outcome of the legislative process to reform the 2012 EMIR framework has been rather different.

The 2019 CCP Supervision Regulation
Co-legislators have adopted an opposite approach from the 2017 CCP Supervision Proposal to address the shortcomings of the 2012 EMIR regime. Instead of centralising EU-CCP supervision at the supranational level, the 2019 CCP Supervision Regulation extends and proceduralises EMIR college-based arrangements. 214 Accordingly, the competences of the colleges have been increased: 215 they are now required to express their opinion also on qualified shareholdings 216 and approvals of significant outsourcing agreements of risk management activities. 217 Moreover, active participation in the meetings is encouraged. 218 To prevent colleges from becoming mere occasions, each member is empowered to add points to the agenda of the meetings 219 as well as to request the insertion of specific recommendations on EU-CCP's risk management in the collegial opinion. 220 On the other hand, any significant deviation from the college's opinion must be thoroughly explained by the home NCA in its final decision. 221 Consequently, the written agreements regulating the functioning of the colleges must be improved by inserting granular rules on, inter alia: the procedures for determining the agenda as well as the frequency of the meetings; the format, the scope, and modalities for exchanging information between members; and the congruous timeframes to allow college members to analyse the documentation sent by home NCAs. 222 Additionally, the 2019 CCP Supervision Regulation seeks to enhance the co-ordinating role of the colleges by broadening their membership. 223 Subject to the home NCA's consent, NCAs of any clearing member as well as CBIs of any EU currency can participate in the meetings of the college 224 in a non-voting capacity. 225 Notably, the composition of colleges has been adapted to take into account the major institutional novelty of the 2019 CCP Supervision Regulation: ESMA's new CCP Supervisory Committee, 226 whose Chair and two independent members must alternatively take over ESMA's non-voting seat in the colleges. 227 Such CCP Supervisory Committee is designed as the hub where information and best practices are going to be collected and shared across the network of NCAs, colleges, and CBIs to promote CCP supervisory coordination and convergence. 228 The 214 See 2019 CCP Supervision Regulation, Recitals 22-24. 215 Ibid., EMIR, Art. 18(1). 216 Ibid., Art. 32(1). 217 Ibid., Art. 35(1). 218 Ibid., Recital 24. 219 Ibid., EMIR, Art. 18(4). 220 Ibid. Art. 19(1a). The request must be endorsed by the majority of the college, albeit separately from the vote on the whole opinion. CBIs are entitled to adopt recommendations pertaining to their own currency. 221 Ibid., Art. 19(4). 222 Ibid., Art. 18(5). 223 Ibid., Art. 18(2) obliges home NCAs and ESMA to publish an updated list of all college members on their website to increase transparency accordingly. 224 Ibid., Art. 18(2)(ca, i). Requesting NCAs and CBIs must provide evidence that the EU-CCP's distress could impact the financial stability of their Member State or the currency they issue, respectively. The home NCA must explain the eventual refusal of the request in writing. 225 Ibid., Art. 19(3). 226 Ibid., Art. 24a(1). 227 Ibid., Art. 18(2)(a). 228 Ibid., Arts. 23a(1). ESMA must particularly focus on 'supervisory areas which have a cross-border dimension or a possible cross-border impact'.
Chair and the two independent members of this permanent internal body are going to be appointed by ESMA's Board of Supervisors-subject to the confirmation of the European Parliament 229 -for five-year terms (renewable once). 230 Alongside the CCP-specific NCA, these three independent 231 experts would be the only voting members 232 of the CCP Supervisory Committee. 233 However, CBIs can request to take part in the CCP Supervisory Committee's discussions on Union-wide stress tests and market developments relating to EU-CCPs, 234 albeit in a non-voting capacity. 235 Moreover, the Chair can invite the remaining members of EMIR colleges as non-voting observers. 236 The tasks of the CCP Supervisory Committee include promoting exchanges of information and discussions among all the authorities involved in EU-CCP supervision. 237 Furthermore, the CCP Supervisory Committee is entrusted with conducting NCA peer reviews and coordinating EU-CCP stress tests on (at least) an annual basis. 238 On the other hand, NCAs and EMIR colleges have to promptly provide the CCP Supervisory Committee with all the information necessary to fulfil its tasks. 239 Nevertheless, the main mechanism supporting the supervisory convergence role of the CCP Supervisory Committee is the introduction of a new ex ante consultation procedure. 240 On the one hand, home NCAs are obliged to submit a wide set 229 Ibid., Art. 24a(5). In particular, the Chair and the independent members are selected on the basis of their merits, expertise, and gender balance following an open selection procedure. ESMA's Board of Supervisors must submit the shortlist of selected candidates to the European Parliament. In turn, the European Parliament would either approve or reject the candidates after a hearing. 230 Ibid. However, if the Chair or any independent member ceased to be fit for office or committed serious misconduct, the Commission should advise-subject to the European Parliament's approval-the Council to remove them. However, both the Council and the European Parliament can inform the Commission that in their opinion the conditions for dismissal are met. In turn, the Commission must reply to such notification. 231 Ibid., Art. 24a(6). On the other hand, Art. 24e clarifies that the Chair and the independent members are accountable to two EU institutions: the Council and the European Parliament. Both may invite the Chair and any independent member to make a statement. Such a statement must be preceded by a written report on the main activities of the CCP Supervisory Committee. Moreover, the European Parliament is entitled to request them to provide any type of information on an ad hoc and confidential basis, apart from confidential information relating to a specific CCP. 232 Ibid., Art. 24a(2)(a-c). If a Member State appoints multiple NCAs, each of them is entitled to send a representative to the CCP Supervisory Committee. However, all the representatives from a single Member State can cast only one vote. 233 Ibid., Art., 24a(4) clarifies that the CCP Supervisory Committee should meet on the initiative of any voting member. However, the CCP Supervisory Committee must be convened at least five times a year. 234 For the role of CBIs in relation to third-country CCPs, see infra, n. 261. 235 2019 CCP Supervision Regulation, EMIR, Art. 24a(2)(d)(ii). Membership shall be granted automatically upon a written request to the Chair. 236 Ibid., Art. 24a(3). 237 Ibid., Art. 24a(7)(c). Furthermore, the CCP Supervisory Committee must specifically discuss all the opinions of the colleges (Art. 24a(7)(d)). 238 Ibid., Art. 24a(7)(a-b). 239 Ibid., Art. 24a(7). Moreover, Art. 21(3) empowers ESMA to request NCAs to allow its staff to participate in their on-site inspections of EU-CCPs. Furthermore, NCAs may forward any information gathered therefrom to ESMA. 240 See ibid., Recital 15. of draft decisions to ESMA before their formal adoption. 241 Indeed, this obligation covers most supervisory areas not related to the 'default waterfall': access to CCP and trading venues, the granting and extension of CCP authorisation, record keeping requirements, a review of qualifying shareholdings, conflict of interest management, outsourcing agreements, fair treatment requirements towards clearing members and clients, and the approval of interoperability arrangements. 242 On the other hand, the 2019 CCP Supervision Regulation permits home NCAs to voluntarily submit any other draft decisions to the CCP Supervisory Committee. 243 In either case, the CCP Supervisory Committee should assess whether the home NCA's draft decision impinges on the consistent application of specific EMIR requirements. 244 If so, the CCP Supervisory Committee must provide home NCAs with its detailed opinion within twenty days from the receipt of the draft decision. 245 However, while home NCAs must duly take into consideration the opinion eventually issued and inform ESMA of their subsequent conduct, 246 the content of the final decision remains at their 'full discretion'. 247 Similarly, colleges are free 'to determine the content of their opinion at their own discretion' 248 despite any contrary advice by the CCP Supervisory Committee. Nonetheless, if such consultation procedure or the other workings of the CCP Supervisory Committee have revealed the persistence of idiosyncratic supervisory practices, ESMA's Board of Supervisors is empowered to issue the necessary guidelines, recommendations, or opinions on its own initiative 249 or at the request of the CCP Supervisory Committee. 250 Furthermore, the CCP Supervisory Committee could always issue opinions on the Board of Supervisors' draft decisions on the application of EMIR by NCAs, 251 save for those on breaches of EU law and binding mediations. 252 Hence, unlike the Commission's original proposal, the 2019 CCP Supervision Regulation leaves the home country control principle intact. Notably, however, the reform is essentially aligned with the 2017 CCP Supervision Proposal on two important aspects: the absence of loss mutualisation between Member States and the third-country CCP supervisory arrangements. With regard to the former, co-legislators and the Commission did not want to challenge the original choice of the 2012 241 Ibid., EMIR, Art. 23a(2). 242 Ibid. 243 Ibid. 244 Ibid., Arts. 23a(3), 24a (7). 245 Ibid., Art. 23a(3). However, Recital 15 clarifies that the CCP Supervisory Committee should refrain from issuing such an opinion where no diverging views have emerged within its members. 246 Ibid., Art. 23a(4). 247 Ibid., Recital 15. 248 Ibid. 249 Ibid., Art. 24a(8), but see also Art. 23a(3). 250 Ibid., Art. 24a(9)(a). The Board of Supervisors should duly consider the request and provide an appropriate response. 251 Ibid., Art. 24a(9)(b). For example, Recital 16 states that the CCP Supervisory Committee may issue its opinion on 'draft technical standards or draft guidelines developed by ESMA in the area of authorisation and supervision of CCPs '. 252 I.e. 2010 ESMA Regulation, Arts. 17, 19. EMIR regime to leave fiscal responsibility at the national level. 253 Accordingly, neither the decision-making CCP Executive Session nor-a fortiori-the internal CCP Supervisory Committee were designed to overrule home NCAs' decision on fiscally sensitive areas such as the default waterfall.
The other similarity between the 2017 CCP Supervision Proposal and the final text agreed by co-legislators concerns the introduction of a new third-country CCP supervisory framework. The reason behind such an agreement is the decision of the United Kingdom to withdraw from the EU: EMIR supervisory arrangements have had to be adapted to take into account the fact that vast amounts of transactions denominated in Union currencies are going to be cleared by third-country CCPs. 254 Like the 2017 CCP Supervision Proposal, 255 the 2019 CCP Supervision Regulation seeks to improve third-country CCP supervisory arrangements by establishing a sliding scale of supervisory and regulatory regimes. Accordingly, ESMA is empowered to determine whether a third-country CCP is (or is likely to be) systemically important. 256 If it is not (Tier 1), the third-country CCPs would be subject to an equivalence regime. 257 If it is judged to be systematically important (Tier 2), a thirdcountry CCP would need to abide by the EU regime 258 under the supervision of the CCP Supervisory Committee. 259 Supported by a third-country CCP college, 260 the CCP Supervisory Committee will prepare the draft supervisory and enforcement decisions 261 on third-country CCPs for their final adoption by ESMA's Board of Supervisors. 262 The external implications of EMIR reform are outside the scope of this article. 263 However, the new third-country CCP supervisory framework also has internal repercussions for the division of competences between ESMA and the ESCB. 264 In 253 European Commission (2017a), p 57 was particularly explicit: home NCAs would 'retain primary responsibility for those supervisory tasks which could potentially impact on national fiscal responsibilities such as the authorisation/withdrawal of authorisation of the CCP, the default waterfall, default procedures, and recovery plans'. 254 European Commission (2017a), p 11; Mersch (2018). 255 2017CCP Supervision Proposal, EMIR, Arts. 25-25n. 256 2019. In turn, Arts. 24a(10), 24d specify that such a decision is taken by ESMA's Board of Supervisors on the basis of the preliminary assessment of the CCP Supervisory Committee. ESMA's Board of Supervisors must also consult with the ESRB and CBIs before making the final determination. 257 See ibid., Recital 32, Art. 25(2)(e). 258 Ibid., Art. 25(2b)(a). Nevertheless, Art. 25a allows Tier 2 CCPs to apply to ESMA for comparable compliance according to Art. 25a. 259 Ibid., Art. 25b. 260 See ibid., Art. 25c. Contrary to the EU-CCP colleges, the third-country college does not have any decision-making power. Instead, it should facilitate information exchange among its members, which are: the permanent members of the CCP Supervisory Committee, NCAs of EU-CCPs, NCAs of EU-based clearing members, NCAs of EU-based trading venues served by the third-country CCP, NCAs of EUbased interlinked CSDs, and all the members of the ESCB. 261 See ibid., Arts. 25f-25k, 25p-25q, and Annexes III-IV. CBIs of the financial instruments cleared or to be cleared by a third-country CCP can request to participate in the preparation of such draft decisions, albeit in a non-voting capacity (Art. 24a(2)(i)). 262 Ibid., Arts. 24a(10), 24d. 263 On the topic, see Ferrarini and Trasciatti (2018). 264 For the different layers of the European monetary governance, see supra, n. 34. particular, the CCP Supervisory Committee must consult with the CBIs of all Union currencies of the financial instruments cleared (or to be cleared) 265 by a Tier 2 CCP before submitting its draft decisions on margin and collateral requirements, liquidity risk controls, settlement arrangements, and the approval of interoperability arrangements. 266 While the original proposal of the Commission envisaged that CBIs could veto such draft decisions on the basis of monetary stability concerns, 267 the 2019 CCP Supervision Regulation states that ESMA is not bound by the eventual objections or amendments put forward by any CBI. 268 However, the CCP Supervisory Committee must duly explain eventual deviations therefrom to the CBI concerned and ESMA's Board of Supervisors. 269 Moreover, the 2019 CCP Supervision Regulation empowers each CBI to impose a specific number of additional obligations which Tier 2 CCPs must meet in order to be recognised. Accordingly, a Tier 2 CCP might be requested to provide the CBI with information that is otherwise unavailable from ESMA as well as to cooperate during stress tests. 270 Additionally, a Tier 2 CCP might be asked to open (or notify the intent to open) an overnight deposit with the CBI. 271 Finally, if 'temporary systemic liquidity risks affecting the transmission of monetary policy or the smooth operation of payment systems' have materialised, any CBI might impose additional conditions on collateral and margin policy, 272 settlement and interoperability arrangements, and liquidity risk control of a Tier 2 CCP. However, co-legislators have not given CBIs complete carte blanche. In fact, such requirements are designed to be exceptional and non-permanent: they can be applied only for up to six months (extendable once) and cannot cover 'other areas of prudential supervision' than those explicitly mentioned. 273 Moreover, these measures cannot result in atypical burdens, but they must be in line with the requirements which supervisory authorities can already apply under EMIR to ensure the 'efficiency, soundness and resilience' of CCPs. 274 Furthermore, an ex ante consultation procedure applies: the CBI must explain to ESMA, the third-country CCP college, and all the other relevant members of the ESCB why imposing or extending such exceptional requirements is going to be necessary and proportionate to achieve their monetary policy tasks. 275 In turn, ESMA and the other members of the ESCB can propose amendments to the draft decision of the CBI. 276 In the case of an extension, the European Parliament and the Council must be informed as well. 277 Notably, the power to impose the relocation of Union-denominated business in the EU is not included among the additional conditions which can be applied to Tier 2 CCPs by CBIs. In fact, the 2019 CCP Supervision Regulation makes it clear that the order to open an overnight deposit 'should not amount to an obligation to relocate all or part of the clearing services of the CCP'. 278 Indeed, relocation is envisaged as a last resort measure, 279 which is not up to CBIs to decide autonomously. Only if a recommendation of ESMA's Board of Supervisors deems a Tier 2 CCP (or some of its services) to pose unmitigable risks can the Commission deny its recognition; in this case, the only alternative for a Tier 2 CCP would be to relocate in the EU. 280 CBIs can only put forward amendments or objections to the draft recommendation. 281 However, CBIs' amendments or objections are binding only in relation to the currency they issue and 'not to the recommendation as a whole'. 282 Such an important but constrained role of central banks in CCP oversight is the final outcome of a spectacular inter-institutional querelle, which has exposed the first of the three weaknesses of the 2019 CCP Supervision Regulation.

'Whatever It Takes'
Given the importance of CCPs for both monetary and financial stability, preventing dangerous interest misalignments between supervisory authorities and central banks has been one of the main objectives of the EMIR reforming effort. Accordingly, the 2019 CCP Supervision Regulation aims at clarifying the spheres of competences of central banks-especially the ECB 283 -in relation to CCPs. However, there seems to be tension between the supervisory arrangements of the 2019 CCP Supervision Regulation and the monetary constitution of the EMU. 275 Ibid., Art. 25(2b). 276 Ibid. CBIs must keep on sharing information and cooperating with ESMA and the other relevant members of the ESCB on a continuous basis. 277 Ibid. 278 Ibid., Recital 33. 279 Ibid., Recital 39. 280 Ibid.,EMIR,Art. 25(2c). 281 Ibid.,Art. 24b(3). Such draft recommendation is prepared by the CCP Supervisory Committee in cooperation with the ESRB. 282 Ibid.,Art. 25(2c). 283 Ibid., Arts. 18(2)(c), 19(3) state that if the ECB sits in the colleges as both the CBI and clearing member supervisor, it can express two votes.
As mentioned above, 284 the Monetary Union is founded on an independent central bank. 285 In particular, the definition of the monetary policy of the euro area is reserved to the ECB's Governing Council. 286 This body enjoys a wide margin of discretion to accomplish its mandate. 287 As seen above, 288 CCPs' distress might have significant repercussions for monetary stability. Accordingly, the ECB has shown a keen interest in CCPs since it came into operation. As early as 2001, the ECB developed a CCP oversight policy. 289 In 2011, this framework was updated to include a CCP 'location policy'. In particular, the ECB could have compelled 'off shore' CCPs to relocate their euro-denominated business in the euro area if certain thresholds were exceeded. 290 Now, as then, a great amount of euro-denominated transactions were cleared by CCPs established in the United Kingdom, thus outside the euro area. 291 Consequently, the United Kingdom brought three cases against this ECB policy before the General Court. The judges ultimately held that the Treaties 292 do not confer explicitly or implicitly an 'autonomous regulatory competence in respect of all clearing systems' but only of 'payment clearing systems' upon the ECB. 293 However, the Court also noted that the Treaties envisage a simplified procedure to amend the Statute of the ECB. If the ECB had considered an amendment 'necessary for the proper performance' of its tasks, it could have recommended to do so to the co-legislators. 294 The ECB issued this recommendation after the presentation of the 2017 CCP Supervision Proposal. While the ECB had welcomed the Commission's initiative, 295 in its opinion the 2017 CCP Supervision Proposal did not adequately reflect the risks posed by CCPs to the monetary stability of the euro area. 296 As noted earlier, 297 the 2017 CCP Supervision Proposal envisaged that CBIs could only veto a supervisory decision taken by ESMA. For the ECB, this regime would have not sufficiently addressed the growing systemic importance of CCPs and the Brexit-related risks. 284 Supra,Sect. 2.2. 285 See TFEU Arts. 130,283(3), which aim at insulating the ECB, national central banks, and members of their respective decision-making bodies against external pressures via a wide-ranging independence guarantee. 286 IV Protocol (ECB's Statute), Art. 12.1. 287 See Case C-62/14 Peter Gauweiler and Others v. Deutscher Bundestag, ECLI:EU:C:2015:400. TFEU Art. 127(1) sets a clear hierarchy of objectives which the ECB is bound to pursue. The first and foremost objective is to maintain price stability. Subject to that, the ECB should also support the 'general economic policies in the EU'. 288 Marjosola (2015); Wolfers and Voland (2015). 294 Ibid., pp 108-109, referring to TFEU, Art. 129(3). 295 ECB (2017a), p 1. 296 ECB (2017b), pp 2-3. 297 Supra,n. 267. Accordingly, it recommended to amend its Statute to empower the Eurosystem 'to adopt binding assessments and require remedial action' on its own. 298 Moreover, the ECB should have been given additional powers-outside the EMIR frameworkover CCPs clearing significant amounts of euro-denominated transactions.
Consequently, co-legislators agreed to amend the ECB's Statute. However, following the advice of the Commission, 299 they added significant restrictions on the recommended amendment. The ECB would not have had any new authority over EU-CCPs. Instead, the ECB would have received only a specific number of additional powers over systematically important third-country CCPs. These circumscribed exceptional powers should have been exerted within the framework of the revised EMIR. Accordingly, they should have been exercised only over third-county CCPs deemed systematically important by ESMA. 300 Notably, the ECB could have not ordered third-country CCPs to relocate their euro-denominated business in the euro area. As seen above, these tight constraints were transposed almost verbatim in the final text of the 2019 CCP Supervision Regulation.
The Governing Council unanimously considered these limitations unacceptable and withdrew its recommendation. 301 Exercising its powers within the constraints set by secondary legislation would have deprived the ECB of the 'discretion to adopt measures which are necessary to carry out the Eurosystem's basic tasks, and would thus violate the ECB's functional independence, established in the Treaties'. 302 Moreover, the ECB noted that it could have exercised these powers only over thirdcountry CCPs deemed to be Tier 2 by ESMA. This fact would 'limit the scope of the ECB's competence to address risks to monetary policy or the smooth functioning of payment systems, in a way that infringes its independence, as the ECB's competence would be based on decisions taken by another authority (ESMA) on the basis of criteria which are distinct from the ECB's primary mandate'. 303 Furthermore, the ECB pointed out that both the amendment to the Statute agreed by co-legislators and the text of what would become the 2019 CCP Supervision Regulation would have required central banks to ensure the 'efficiency, soundness and resilience of clearing systems for financial instruments' when they exercise their powers. 304 For the ECB, these provisions 'may imply that the ECB must prioritise the objectives of the prudential supervisory framework when carrying out its own tasks, notwithstanding its central banking objectives under the Treaties'. 305 Therefore, the framework of the CCP Supervision Regulation does not seem to take adequately into account the monetary implications of central clearing. Nevertheless, the next section will argue that the CCP Supervision Regulation would also fail to sufficiently meet the just-mentioned objectives of prudential supervision. 298 ECB (2017b), p 3. 299 See European Commission (2017c). 300 ECB (2019a), pp 3-4. 301 See ECB (2019b). 302 ECB (2019a), p 3. 303 Ibid., p 4. 304 Ibid. 305 Ibid.

He Who Pays the Piper Calls the Tune
Fostering the development of a 'coherent and consistent' 306 European approach to EU-CCP supervision is another objective of the 2019 CCP Supervision Regulation. For co-legislators, the establishment of a CCP Supervisory Committee would address the need to counter supervisory idiosyncrasies by reinforcing 'ESMA's ability to promote convergence in the supervision of CCPs'. 307 However, the 2019 CCP Supervision Regulation does not address the primary source of such inconsistencies, namely the absence of loss mutualisation between Member States.
During the negotiations of the 2012 EMIR regime, Member States decided to allocate EU-CCP fiscal responsibility at the national level. As seen earlier, 308 this choice has shaped EMIR supervisory arrangements, which have left EU-CCP supervision mostly in the hands of home NCAs. The decentralised ESFS was just adapted by creating the inadequate EMIR colleges. As a necessary consequence, NCAs have not developed a single approach to EU-CCP supervision. In fact, some members of the Board of Supervisors have shown some apprehension for the supervisory convergence role played by ESMA under the 2012 EMIR regime 309 and for any further centralisation. 310 The 2019 CCP Supervision Regulation would not change this state of affairs. As in the 2017 CCP Supervision Proposal, 311 fiscal responsibility remains at the national level. Accordingly, the new ex ante mandatory consultation procedure has been designed so as to ensure 'that that the final content of the respective decision would remain at the full discretion of the CCP's competent authority', 312 regardless of any objection by the CCP Supervisory Committee. Furthermore, such a procedure does not cover home NCAs' measures on the default waterfall, which is the most sensitive aspect of EU-CCP supervision. As seen earlier, 313 the mutualisation of losses via the default waterfall is central clearing's main benefit and greatest danger at once. Nevertheless, home NCAs would remain totally responsible for the supervision of the main source of the macro-prudential risk of EU-CCPs as under the 2012 EMIR framework. As ESMA's peer reviews have shown, this could lead to idiosyncratic supervisory practices.
Indeed, past experience has demonstrated that it would be impossible to create arrangements ensuring a single supervisory approach without a prior agreement on loss mutualisation. As noted above, 314 the centralisation of banking prudential supervision in the ECB was only possible after the establishment of the ESM. States via the ESM. Accordingly, loss mutualisation has led to what appears to be a solid centralised framework based on a single supervisory manual. 315 Quite revealingly, the Commission acknowledged that a truly single CCP supervisor would be required to 'address effectively the need for supervisory convergence'; by its own admission, this choice was not possible because 'supervisory and fiscal responsibilities would not be fully aligned'. 316 Therefore, it is questionable that the CCP Supervisory Committee could effectively deliver an European supervisory approach to counter the macro-prudential risk of CCPs. However, it could be contended that the 2019 CCP Supervision Regulation would at least streamline supervisory decision-making. As the next section will argue, this is not the case.

Duplications
One of the main reasons for revising the 2012 EMIR regime was simplifying its decision-making. Paradoxically, the centralised framework envisaged by the initial 2017 CCP Supervision Proposal of the Commission would have added another layer of complexity to the EMIR supervisory architecture. 317 The centralisation of certain supervisory decisions in the CCP Executive Session would not have totally abolished the 2012 EMIR supervisory arrangements. Home NCAs would have retained a decisive role, especially in relation to the most fiscally sensitive tasks. Moreover, EMIR colleges would have continued to exist because of the cross-border, crosssectoral importance of EU-CCPs. Indeed, their unwieldiness would have been increased by the addition of the permanent members of the CCP Executive Session. Furthermore, the 2017 CCP Supervision Proposal would have left final decisionmaking to ESMA's Board of Supervisors. This was a forced choice. As mentioned above, ESMA's governance arrangements points to the Board of Supervisors as the ultimate location of decision-making in order to respect the EU constitutional 'balance of powers' as defined by the Meroni doctrine. Indeed, the supervisory powers of the CCP Executive Session would have needed to be tightly circumscribed and proceduralised to avoid the legal uncertainties associated with Meroni. 318 However, ESMA itself has repeatedly lamented that the formal and substantial constraints attached to its powers undermine its supervisory action. 319 Nevertheless, the 2017 CCP Supervision Proposal did not seem to address this issue. 320 Instead arrangements. 321 Even ESMA's Chair himself publicly declared that he did not support the reform put forward by the Commission. The 2017 CCP Supervision Proposal would have made 'ESMA's governance top-heavy and inefficient' and there would have been 'the risk of uncoordinated decisions, strategies, and communication' between the Board of Supervisors and the CCP Executive Session. 322 While it is difficult to see how the CCP Executive Session could have streamlined decision-making, the 2017 CCP Supervision Proposal might be interpreted as a failed attempt by the Commission to increase the supranational grip on ESMA. 323 The CCP Executive Session's Head and Directors would have been shortlisted by the Commission. 324 Together with the Commission representative in the CCP Executive Session, they would have increased the Commission's influence over regulates and NCAs. However, this intention of the Commission clashed with the realities of ESMA's governance arrangements. As already mentioned, 325 ESMA is built upon NCAs. Despite the risk of remaining 'national-minded bureaucrats', 326 the NCAs sitting on the Board of Supervisors have been able to develop a consistent and responsive approach to EU financial markets regulation and supervision. 327 Transplanting a hierarchical supranational bureaucracy into an NCA-based framework could have altered this equilibrium, especially since the CCP Executive Session could have forced NCAs to take fiscally sensitive decisions. 328 Co-legislators foresaw this risk of the initial proposal of the Commission. The final text of the reform could not be clearer: the three technocratic permanent members of the 'internal' 329 CCP Supervisory Committee will be shortlisted by ESMA's Board of Supervisors, 330 which however retains 'final decision-making power' 331 alongside NCAs, EU-CCP colleges, and CBIs. According to the rapporteur of the European Parliament, this framework 'would be a more proportionate option than this of a wholly new ESMA configuration such as the executive session proposed by the Commission'. 332 Indeed, even though the CCP Supervisory Committee 'should not constitute a precedent for the ESAs', 333 the 2019 CCP Supervision Regulation appears to be in line with the broader reform of ESMA's governance proposed by Professor Moloney to streamline ESMA's decision-making. She has suggested to establish a committee framework within ESMA. These committees would comprise members of the Board of Supervisors. They would be tasked with preparing draft decisions for later adoption by the qualified majority of the Board of Supervisors. Moreover, an independent 'Monitoring Board' would be created to constructively assist the Board of Supervisors in developing its supervisory strategies. 334 However, a closer look at the 2019 CCP Supervisory Regulation reveals that this piece of legislation shares just a superficial resemblance with such sensible adjustments. In fact, co-legislators have made EU-CCP supervisory arrangements even more burdensome. While home NCAs remain primarily responsible for EU-CCP supervision, the persisting college-based framework has been further complicated by the introduction of the CCP Supervisory Committee. Lacking decision-making powers, this internal body is designed as a hub where 'additional reaction' is given by a 'group of supervisors specialised and experienced in the supervision of CCPs'. 335 However, such experienced supervisors would already have the opportunity to exchange their views in EMIR colleges, which comprise all the authorities sitting in the CCP Supervisory Committee. Indeed, colleges have seen their competences extended to cover most of the same supervisory areas subject to the new ex ante mandatory consultation procedure. 336 Moreover, no enforcement novelties underpin such procedural duplication: the NCA-dominated Board of Supervisors remains the ultimate arbiter of which national supervisory idiosyncrasies must be sanctioned. 337 On the other hand, the 2019 CCP Supervision Regulation would accentuate the pre-existing flaws of EMIR colleges by extending their membership. As ESMA's peer reviews have exposed, the effectiveness of the colleges is hindered by the cumbersome number of authorities participating in the meetings, each with its own peculiar mandate, distinct expertise, and national sensitivities. 338 Nonetheless, the reform empowers NCAs of any clearing member as well any CBIs to ask to take part in the colleges. 339 While home NCAs must consent, peer dynamics and the 'comply-or-explain' mechanism backing the request 340 make an eventual refusal unlikely. Indeed, co-legislators have recognised that such enlargement of college composition could obstruct an 'appropriate, effective and swift decision-making process'. 341 In their opinion, such a risk could be prevented by depriving the members, on request, of the right to vote. However, their title of members would still empower them to participate in the college discussions, which would inevitably be time-consuming considering the increasing number of authorities sitting in the meetings. If this was not the case and the members on request passively received information without any active contribution to the college discussions, the reform would accelerate the 334  negative tendency of EMIR colleges to be mere forums for information exchange rather than 'effective supervisory tools'. 342 Therefore, far from simplifying supervisory decision-making, neither the centralised framework envisaged by the initial Commission proposal nor the 2019 CCP Supervision Regulation would streamline EU-CCP supervisory arrangements. In fact, maintaining the home country control principle alongside EMIR colleges would hamper the effectiveness of any central supervisory body, be it a CCP Executive Session or a CCP Supervisory Committee. Moreover, even admitting that it might be possible to improve the decision-making process under such conditions, there would still be the need for an European approach to EU-CCP supervision which could adequately take into account the monetary implications of central clearing. In order to achieve such objectives, a modest counterproposal is put forward in the next section.

The Counterproposal
The counterproposal is to centralise EU-CCP supervision in two EU bodies along objective lines. On the one hand, EU-CCP prudential supervision should be conferred upon the ECB. Accordingly, EU-CCPs should be subject to ECB supervision in relation to the matters already included in the SSM framework (e.g. authorisation, qualified holdings, capital and liquidity requirements, governance requirements, the review of risk models). 343 However, the scope of the SSM should be adapted to the peculiar mechanics and risks of central clearing. Therefore, the ECB should be empowered to supervise EU-CCPs' default waterfall, default procedures, and interoperability arrangements. 344 On the other hand, EU-CCP conduct of business supervision should be conferred on ESMA's Board of Supervisors. In particular, ESMA should ensure that CCPs act transparently, fairly, and professionally in light of the best interests of their clients. Therefore, ESMA should supervise EU-CCP conflict of interest requirements, participation requirements, record-keeping requirements, account segregation and portability requirements. 345 Given its acquired technical capacity, ESMA could also be involved in CCP stress testing together with ECB.
While the conferral of conduct of business supervision upon ESMA could be based on TFEU Article 114, entrusting the ECB with EU-CCP prudential supervision is somewhat more problematic. Given the failed attempt to reform the ECB's Statute, a possible legal basis could be the same provision which enabled the establishment of the SSM, namely TFEU Article 127(6). 346 It has been contended that the scope of this provision is not bound by the definitions provided by secondary 342  legislation. 347 However, it is worth considering whether CCPs can already fall within the definition of the 'financial institutions' named in this provision. EMIR suggestively, but imprecisely, defines CCPs as 'the buyer to every seller and the seller to every buyer'. 348 In turn, the EU's capital requirements regime defines 'financial institutions' as undertakings whose principal activity is, inter alia, 'trading for own account or for account of customers' in 'exchange and interest-rate instruments'. 349 These contracts form the majority of the OTC derivatives subject to EMIR clearing obligations. 350 Accordingly, it can be argued that CCPs are par excellence financial institutions. Their very raison d'être is to trade (to buy and to sell in the words of EMIR) derivatives including exchange and interest-rate instruments for the account of their customers (the clearing members). 351 Indeed, there might be grounds to include EU-CCPs even in the category of 'credit institutions' since the German CCP Eurex as well as the French CCP LCH SA are so authorised. 352 Adopting such a centralised and objective-based framework would bring the benefits that the 2019 CCP Supervision Regulation (as well as the initial proposal of the Commission) tries to achieve to no avail, but without its shortcomings.
First, as already recognised by some-including the Commission-, the full centralisation of EU-CCP supervision at the EU level would be necessary to develop an effective single supervisory approach. 353 In this regard, while pragmatically making good use of national expertise, the SSM has proved to be a solid and unitarian framework for prudential supervision. 354 In turn, ESMA has acquired unique knowledge in the field of conduct of business supervision thanks to its multiple roles already played in this field. 355 Second, the SSM framework already contains established internal mechanisms to conciliate supervisory actions with monetary policy. 356 358 Third, decision-making would be effectively streamlined. On the one hand, the SSM already contains relatively effective and 'Meroni-free' decision-making procedures. 359 On the other hand, ESMA's decision-making process-especially if enhanced by Professor Moloney's committee framework-would not be burdened by the constraints of the home country control principle and EMIR collegial procedures.
Fourth, the most fiscally sensitive supervisory decisions would be conferred upon an EU institution with relatively adequate accountability arrangements. While ESMA's accountability mechanisms are still developing, 360 the ECB has been able to build a constructive (although perfectible) dialogue with the representatives of the EU people. 361 Fifth, the eventual expansionist tendencies of the Behemoth represented by the ECB could be counterbalanced by ESMA. 362 Indeed, also the ECB has expressed its preference for a single counterparty in financial markets supervision. 363 Finally, this new supervisory framework would be a significant step towards the desired European 'Financial Union' 364 and the envisaged pan-EU 'Twin Peaks' 365 model, which could lead to responsive and consistent supervision across the Single Market. 366 Nevertheless, this reform would not be without its challenges, and a certain number of criticisms can be anticipated.
Firstly, it could be pointed out that such reform would require that Member States-at least those in the euro area-decide to mutualise losses. Indeed, the different outcomes of ESFS and SSM negotiations have confirmed that the 357 Ibid.,Art. 26(8). 358 Ibid.,Art. 25(5 Constâncio (2017). 364 According to Juncker (2015) (the Five Presidents Report), the EMU could not be 'genuine' as long as a Financial Union comprising both a completed Banking Union and an integrated CMU is not established. As seen supra, n. 197, the centralisation of supervisory tasks at the EU level has been one of the main elements of the CMU agenda. 365 Such is the nickname of any supervisory architecture which in principle entrusts conduct of business supervision and prudential supervision on two distinct authorities for the whole financial system (Bank for International Settlements 2018). Ferran (2015), pp 101-102 notes that the focus on the clear supervisory objectives of the Twin Peaks model is typically associated with more uniform and responsive supervisory practices than other possible institutional frameworks, namely sector-based systems (such as the ESFS) or single supervisory authorities for any given jurisdiction. 366 Colaert (2015), pp 1611-1613; Schoenmaker and Véron (2017), pp 5-9; Avgouleas and Ferrarini (2018), pp 69-72. centralisation of supervisory tasks is possible only after an agreement on loss mutualisation. However, the current institutional and political situation has proved to be particularly propitious for overhauls of the EMIR framework: Brexit has meant that the EU has lost the fiercest opponent of CCP supervisory centralisation. 367 Furthermore, linking EU-CCP supervision to the SSM could be questioned. It could be argued that the cross-border risks posed by EU-CCP require a prudential framework covering the whole EU from the beginning. Moreover, it could be contested that this reform would be another attempt by euro area countries to impose their objectives over the Internal Market project. 368 Indeed, establishing a pan-EU framework from the start would be the outcome desired by this paper. Nevertheless, the history of EU integration is marked by a pragmatic and incremental recycling of existing rules. 369 The counterproposal advanced here would be in line with that history. Given the great efforts made for establishing the SSM, creating a whole new prudential supervisory framework would be neither realistic nor efficient. Furthermore, a significant part of EU-CCP operations would still be supervised by a pan-EU authority such as ESMA, which would inject an Internal Market perspective into the framework. Moreover, it is worth noting that after Brexit only three EU-CCPs are currently established outside the euro area. 370 Therefore, it can be argued that EU-CCP supervision is mostly going to be an issue affecting the euro area. In addition, if non-euro area countries wanted to reap the benefits of this model, they could always ask for a close cooperation agreement under the SSM framework. 371 Employing TFEU Article 127(6) as a legal basis can be questioned too. Indeed, the exact scope of the 'specific tasks' mentioned by that provision has been the object of a long querelle between those in favour of an extensive reading 372 and those advancing a restrictive interpretation. 373 However, even admitting that the latter group is right, conferring EU-CCP prudential supervision upon the ECB remains ceteris paribus a more sensible solution than the supervisory arrangements of the 2019 CCP Supervision Regulation for the reasons listed above.
is not yet an established consensus on this issue. 375 Moreover, as noted above, the SSM already contains mechanisms to conciliate supervisory decisions with monetary policy. They seem to work smoothly. 376 The drafting of Article TFEU 127(6) has been influenced by another concern, which resurfaced when the SSM was established: the risk of concentrating too much power in the statutorily independent ECB. 377 However, this legitimate preoccupation must be reconsidered in light of EU-CCP industry. In particular, there are just thirteen CCPs established in the EU. 378 The ECB already directly supervises many of the euro area credit institutions which can meet the stringent requirements for being clearing members. Accordingly, the ECB is already part of EMIR colleges which supervise the few EU-CCPs. 379 Given these tasks already conferred upon the ECB, it can be argued that entrusting the ECB with EU-CCP prudential supervision would not significantly move the existing balance of powers. Moreover, as noted above, the ECB has been ready to refine its accountability arrangements for its supervisory tasks. 380

Conclusion
CCPs have gone rapidly from being a solution to a problem. There is a growing awareness worldwide that the concentration of large volumes of OTC derivatives transactions into few interconnected FMIs could have dramatic consequences in the future. Nevertheless, the risks posed by CCPs are particularly dangerous for the EU. Financial distress experienced by one EU-CCP could reverberate across the EMU with unpredictable knock-on effects. Indeed, the sovereign debt crisis is a testament to the fragility of the Union.
Therefore, the EU-CCP supervisory architecture must be built on robust foundations. EMIR supervisory arrangements have proved to be too burdensome and fragmented to effectively address the challenges of CCPs. The EU must be praised for recognising the dangers of inaction. Nevertheless, the 2019 CCP Supervision Regulation creates more problems than it solves. While NCAs might keep on maintaining their idiosyncratic supervisory practices, EMIR supervisory arrangements would become even more unwieldy. Furthermore, ESMA's new powers are in tension with the high level of independence enjoyed by the ECB as a central bank. 375 Ferran (2015), pp 113-115. As noted ibid., refraining from fixing interest rates which could cause bank failures is a traditional example of the trade-offs that central banks might face as supervisors. Anyhow, as seen supra, n. 356, the 2013 SSM Regulation establishes the principle of separation between the ECB's supervisory and monetary tasks. 376 Moloney (2019) The centralised and objective-based model proposed by this paper for EU-CCP supervision might seem more ambitious, but it would exploit the potential of the current EU supervisory architecture. The SSM has proved to be a solid and unitary framework for prudential supervision. The ECB has developed smooth decisionmaking procedures and relatively adequate accountability arrangements. On the other hand, this framework would put to good use ESMA's expertise in conduct of business supervision. Moreover, ESMA might represent the necessary counterbalance to possible expansionist tendencies of the ECB. Finally, the fruitful cooperative relationship developed by NCAs within ESMA's Board of Supervisors would not be endangered by an alien supranational bureaucracy as in the original proposal of the Commission.
Past experience has shown that such a reform might only be possible after an agreement on loss mutualisation between Member States. However, Member States should be wary of inaction. The decision made in Maastricht of having a Monetary Union without a truly complete Economic Union has had dramatic consequences decades later. However, Member States have proved to be ready to agree on ambitious supervisory reforms, at least when they are pressured to do so. The establishment of the SSM was possible only after a crisis which put into question the existence of the Union. The hope expressed by this paper is that an adequate reform of EMIR supervisory arrangements would not require such a trauma.