Integrating rules, disintegrating markets: the end of national discretion in European banking?

ABSTRACT

This article examines the integration of prudential banking regulation and supervision in the European Union. It demonstrates that incremental integration in the two decades preceding the sovereign debt and banking crisis produced a regulatory process that afforded member states wide discretion when it came to implementing European rules at the national level. Decentralized implementation was an aggravating factor in the partial disintegration of the Single Market in financial services since 2008. The recently agreed ‘Single Rulebook’ and ‘Single Supervisory Mechanism’ reduce the scope for home bias in the implementation of European rules, thereby shifting the regulatory process in banking closer towards an idealized ‘end-point’ of outright supranational governance. However, change is taking place within an increasingly complex regulatory environment that is shaping the outcome of reform. One consequence is that national discretion in some important prudential areas will persist.

KEY WORDS:

• Banking union
• financial integration
• historical institutionalism
• supranational governance

INTRODUCTION

The architecture of banking governance in the European Union (EU) is currently undergoing significant reform. Two of the most important innovations agreed to date are the Single Rulebook, whereby prudential standards are subject to ‘maximum harmonization’ across the Union, and the Single Supervisory Mechanism (SSM), which will see the European Central Bank (ECB) assume strong powers to supervize euro area banks from late 2014. The latter is a core element of the proposed ‘banking union’, which is designed to break the pernicious interdependence between banks and sovereigns in the euro area. Together, these reforms also represent the latest steps in a decades-long shift towards more integrated banking regulation and supervision in the EU.

What is driving this supranationalization of banking governance? Member states made the Single Supervisory Mechanism a prerequisite for the other key element of the banking union (Quaglia 2013: 24), namely, a common resolution framework with EU-level funding arrangements for future bank bailouts. Yet the shift to more supranational regulation and supervision cannot be explained as merely the quid pro quo for so-called ‘burden sharing’. The decision to establish a Single Rulebook was taken in 2009, three years before policy-makers first settled on the idea of the banking union. Furthermore, while the SSM is the first step towards the banking union, its stated aim is to ensure coherent and effective implementation of the Single Rulebook ‘unfettered by other, non-prudential considerations’ (Council Regulation 1024/2013, Recital 12). In other words, the SSM aims to eliminate conflicts of interest in the implementation of European rules that have given rise to cross-border supervisory divergences, competitive distortions and financial instability. In part, then, the SSM is an effort to complete the Single Market (cf. Enria 2013a).¹

The drive for deeper integration in this area is puzzling because the financial services policy field as a whole had already been subject to significant integrative reforms in the late 1990s and early 2000s. The European Commission's 1999 Financial Services Action Plan (FSAP) led to a series of reforms that aimed to eliminate regulatory barriers to cross-border transactions and establish deep, liquid capital markets in the EU (Commission 1999). Almost simultaneously, the ‘Lamfalussy’² framework – established in the securities sector in 2001 and extended to banking and insurance in 2004 – put in place a complex four-level system of committee governance, ‘incorporating’ informal networks of national regulators and supervisors within the formal policy-making process (Eberlein and Newman 2008).

The Lamfalussy framework was credited with expediting the policy-making process in finance (Commission 2007a). Recognizing this, scholarship on EU financial integration suggested policy-makers had established a supranationalized regulatory process, centred upon technocratic deliberation between national officials co-operating in formalized transgovernmental committees (Mügge 2006, 2010; Posner 2007, 2010; Quaglia 2008a). This article suggests this scholarship overstated the discontinuity between the outcome of these reforms and the more intergovernmental arrangements they ostensibly superseded (cf. Grossman and Leblond 2011). Like Posner (2007, 2010) and Thatcher and Coen (2008), it suggests integration in this policy field has occurred incrementally, with existing governance arrangements deeply constraining the speed and direction of subsequent reforms. Yet whereas Posner (2007: 152) concluded that a gradual accumulation of actions at the EU level over several decades had set the stage for a ‘regulatory revolution’ in the 2000s, this article suggests the FSAP and Lamfalussy reforms should themselves be coded as significant, but ultimately non-transformational, adaptations of a decentralized regulatory process that provided scope for discretion, diversity and home bias in the implementation of rules at the national level.

In the context of the financial crisis, this decentralized regulatory process reached the point of exhaustion: national authorities began adopting ‘beggar-thy-neighbour’ policies, aggravating the trend of national retrenchment in European financial markets that began with the onset of the financial crisis in 2008. Financial disintegration produced higher, more differentiated, borrowing costs for households and corporations. It also disrupted the monetary policy transition mechanism in the euro area (ECB 2012). A more complete Single Market in banking thus became a necessary, though not sufficient, condition for preserving the wider project of Economic and Monetary Union.

The Single Rulebook and the SSM reduce the scope for home bias in the implementation of European rules, thereby shifting the regulatory process in banking closer towards an idealized ‘end-point’ of outright supranational governance. However, change is taking place within a complex regulatory environment that is shaping the path of reform and ensuring vestiges of the old decentralized regulatory process will persist. Existing organizations have been converted to serve novel purposes and new rules have been layered upon longstanding legal principles. This will enable national authorities to continue to exercise discretion in some important prudential areas.

The article proceeds as follows. The next section establishes an analytical framework capable of distinguishing between transformational and adaptive change in the empirical case at hand. The third and fourth sections examine the regulatory process in banking through two episodes of integrative reform in the decades preceding the sovereign debt and banking crisis. The article then discusses the exhaustion of the decentralized governance structure during the crisis period, before examining, in the sixth section, reforms that have occurred in the past five years. The conclusion highlights implications for future research on the determinants of European integration in this area.

Categorizing institutional change: transformation or adaptation?

A challenge for scholars of post-crisis financial reforms is to distinguish between transformative changes that radically overhaul existing institutional systems and adaptive changes that preserve the essential characteristics of current arrangements (see Moschilla and Tsingou 2013). This section first delineates the institutional system of interest before outlining a historical institutionalist conceptual apparatus capable of classifying change in that system as either transformative or adaptive.

The empirical case examined below is the legal and organizational framework for prudential banking regulation and supervision in the EU. The article examines this framework as a multi-phase regulatory process. This includes the ‘high’ politics of agenda setting and negotiations over primary legislation and the ‘low’ politics of technocratic (ostensibly less political) secondary rule making and day-to-day rule application by supervisors. Attention to the low politics of the regulatory process is important because putting rules into practice is an imperfect process in which space exists between the ideal behaviour prescribed by rules and real life activity under them (cf. Streeck and Thelen 2005). The gap between de jure and de facto behaviour can sow the seeds of future rule changes, as policy-makers update rules to take account of divergent, devious or otherwise unforeseen implementation. Focusing on low politics is especially important in banking, where ostensibly arcane divergences in the interpretation of a rule can have outsized consequences for bank profitability and financial stability. In the empirical case at hand, attention to the low politics of the regulatory process directs us to examine the extent to which European institutions provide scope for national discretion in the transposition, interpretation and enforcement of rules.

How can change in this multi-phase regulatory process be classified as either transformative or merely adaptive? Historical institutionalism offers conceptual insights that illuminate how institutional systems change over time. Historical institutionalist scholars generally agree that existing institutions limit possibilities for abrupt path-altering change because they have self-reinforcing properties and ‘lock-in’ effects that make them resistant to wholesale dismantlement or rapid displacement (Pierson 2004). One strand of historical institutionalist thinking suggests transformative change often emerges incrementally, as new institutions gradually become more prominent and old institutions slowly slip into obsolescence. Streeck and Thelen (2005) identify displacement, layering, drift, conversion and exhaustion as five common modes of incremental institutional change. Three of these modes are particularly apt to describe the process of change that has taken place in the empirical case discussed below (cf. Thatcher and Coen 2008). Layering occurs when new institutions emerge alongside existing institutions. In this scenario, new institutions do not directly challenge old ones, but may gradually supplant them over time. Conversion occurs when old institutions are redirected to new purposes, for example, when existing bureaucratic agencies take on new competencies. Finally, exhaustion occurs when institutions permit or promote behaviour that sows the seeds of their own destruction, leading rule-makers to update or replace existing institutions to account for the changed circumstances.³

Recognizing such patterns helps us to identify elements of difference and continuity in processes of institutional change. It also encourages attention to sources of change that are ‘endogenous’ to the operation of the institutional system itself. However, distinguishing between transformative and adaptive change in a given empirical setting also requires an empirically tailored conceptual apparatus for assessing the substantive character (or ‘direction’) of the changes observed.

The incremental changes discussed below involve a common denominator: they have all involved transfers of sovereignty from national to supranational authorities. To categorize these changes, it is useful to envisage a continuum spanning from national to intergovernmental to supranational politics (cf. Stone Sweet and Sandholtz 1997). At the national pole, rule-making and public administration occurs autonomously. It is unconstrained by formal or informal intergovernmental co-operation and reflects the interplay of domestic political, cultural and economic concerns. At the intergovernmental mid-point of the continuum, national governments and bureaucrats co-operate within formal multilateral organizations and institutionalized bargaining processes at the EU level. National, rather than supranational, authorities remain the key players, and they pursue preferences that derive from domestic societal contests. At the supranational pole, any rules of consequence originate from the EU level. Political actors with authority deriving from supranational institutions are capable of constraining the behaviour of member states, even within multilateral organizations. Relevant societal interests are organized on a transnational basis. As Figure 1 demonstrates, different phases of the regulatory process have been located at different points on this continuum at different moments in time. The following sections trace the impact of different episodes of integrative reform on the rightward progression of each phase of the regulatory process along this continuum.

Figure 1 Episodes (E) of incremental supranationalization

Note: E₁ = Single Market; E₂ = Lamfalussy; E₃ = de Larosière; E₄ = Banking union.

The Single Market in banking

Although financial integration had been an ambition of European policy-makers since the Treaty of Rome, the first major episode of integrative reform began after the Commission issued an ambitious White Paper on ‘Completing the Internal Market’ in 1985 (Commission 1985). This document set out three fundamental principles for the Single Market in financial services: mutual recognition meant each country would recognize the adequacy of the regulatory and supervisory arrangements of every other country; minimum harmonization meant all member states would adopt a set of common basic regulations; and home country control meant the task of supervizing cross-border firms would lie with the authorities of the countries in which the firms were headquartered.

These principles were instituted in the banking sector with the adoption of the Second Banking Directive, the Own Funds Directive and the Solvency Ratio Directive in 1989. In keeping with the principle of minimum harmonization, this legislation required transposition into domestic rulebooks before having legal effect and it permitted member states to embellish – or ‘gold-plate’ – the rules as they saw fit. For example, the Own Funds Directive set out the maximum range of financial instruments that could be included in the calculation of a bank's ‘Original’ and ‘Additional Own Funds’ (known more commonly as ‘Tier 1′ and ‘Tier 2’ capital) and set limits on how much of each item could be included in the total calculation. Member states were then free to permit fewer items to be included and/or to impose more restrictive limits on the relative quantities of each item (Directive 89/299/EEC: Article 2(2)). Similarly, the Solvency Ratio Directive set an overall ratio of 8 per cent capital to banks' risk-weighted assets, but allowed national authorities to apply a higher ratio if they so chose (Directive 89/647/EEC: Article 10).

These minimum standards failed to set an inviolable baseline for capital adequacy regulation. This is because they afforded member states numerous opportunities to impose less stringent – or ‘tin-plated’ – standards as well. For example, the Own Funds Directive stipulated that any funds could count as Additional Own Funds if (1) they were freely available to absorb normal banking risks, (2) they had been previously disclosed in internal accounting records and (3) their amount had been determined by the banks' management, externally audited, and previously supervized by the relevant authorities (Directive 89/299/EEC: Article 3(1)). Similarly, the Solvency Ratio Directive contained numerous loopholes including inter alia exemptions for banks specializing in interbank lending and public debt markets, options to allow home country control of subsidiaries of foreign banks and national discretion in relation to asset risk-weights in several areas (see Directive 89/647/EEC: Articles 1(2), 3(4,6), 7, 8).

During this period there already existed a number of mechanisms for promoting co-operation between national supervisors and consistent implementation of European rules. In 1972, supervisors from the then six-country European Economic Community formed a Groupe de Contact to facilitate information exchange on cross-border banks and enable mutual learning about different national supervisory practices. Additionally, from 1977, a Banking Advisory Committee (BAC) composed of officials from member states' finance ministries, central banks and supervisory authorities worked alongside the Commission on regulatory matters. The BAC was mandated to devise ratios for solvency, liquidity and profitability; to advise the Commission over relevant legislative proposals; to assist with domestic transposition of banking directives; and to advise the Commission in the use of its implementing powers (Schüler 2003).

Both of these groups were informal. The Groupe de Contact was a forum for discussion only; it was ‘specifically not intended to lead to any decisions’ (Goodhart 2011: 21). While the Second Banking Directive and the 2000 Codified Banking Directive gave legal recognition to the BAC's monitoring functions, neither gave it any legally binding powers. What influence it did wield arose from the secretive and informal nature of its composition, which favoured personal contact and mutual trust (Dragomir 2012: 214).

As Figure 1 illustrates, the high-politics of the regulatory process were largely intergovernmental during this episode of reform. It has been documented elsewhere that the timing of the Single Market Programme was attributable largely to the convergence of key member states' preferences (Britain, Germany and France) in favour of market integration and liberalization (Moravcsik 1991; Story and Walter 1997; Underhill 1997; although see Jabko 2006). Likewise, as several authors have further highlighted, intergovernmental disputes nevertheless undermined the integrative power of the resulting directives, with legislative texts ‘fudged’ in the final drafting, such that areas of national divergence were swept under the carpet (Brown 1997; Hall 1997: 694).

However, the influence of supranational actors cannot be discounted. The substantive nature of the Single Market in financial services is scarcely explicable without reference to the European Court of Justice, which set the stage for mutual recognition with its well-known ‘Cassis de Dijon’ ruling in 1979. As the originator of legislative proposals, the Commission also exercised significant influence in setting the agenda for specific legislative acts (cf. Quaglia 2010). Furthermore, the 1986 Single European Act extended qualified majority voting in the Council to matters relating to the internal market, thus helping to shift the negotiation phase somewhat closer to the supranational pole.

Turning to the low politics of the regulatory process, both secondary rule-making and supervision during this period remained significantly closer to the national end of the continuum. The legislative principles of minimum harmonization and mutual recognition confirmed a clear separation between centralized rules and decentralized implementation by national authorities. EU-level mechanisms for promoting uniformity in day-to-day supervision remained unintrusive, comprising only informal dialogue and information sharing through the Groupe de Contact. While intergovernmental co-operation in relation to secondary rule-making had a basis in European law, it too placed few constraints on national regulators' room for manoeuvre.

The Lamfalussy era

In 1999 the Commission re-launched financial market integration with the FSAP. This resulted in an outpouring of new legislation and, indirectly, to the convening of a ‘Committee of Wise Men’, chaired by Baron Alexandre Lamfalussy, with a mandate to propose improvements in the European architecture for rule-making and implementation in securities markets.

Although prudential supervision was excluded from this Committee's mandate (Council 2000), discussions over greater participation of EU institutions in banking supervision took place under the aegis of a Council working group on financial crisis management (Economic and Financial Committee 2001). This issue was subsequently vaulted into the public spotlight in early 2002, when the ECB's then President Wim Duisenberg and its head of banking supervision Tomasso Padoa-Schioppa made public their interest in expanding the ECB's supervisory role (Crooks 2002). The prospect of ECB-led supervision was promptly rebuffed in the Council under strong opposition from the British and German governments. Both countries were advocating a single supervisor model of financial supervision at the time and consequently opposed greater central bank participation in banking supervision (Barber 2002). Nevertheless, a compromise struck over the course of 2002 saw the Council agree to facilitate co-operation between national financial supervisors by extending the Lamfalussy framework from securities to the banking and insurance sectors.

This led to the establishment in 2004 of the Committee of European Banking Supervisors (CEBS) in London. The CEBS succeeded the Groupe de Contact, although the latter continued to exist as the CEBS's main working group.⁴ The CEBS was composed of one voting representative from each country's national supervisory authority and one non-voting representative from each country's central bank. Representatives from the ECB and the Commission participated as observers. As a ‘Level 3’ committee in the Lamfalussy framework, the CEBS had a mandate to provide advice to the Commission in respect of so-called ‘Level 2’ legislation (that is, detailed technical rules for implementing primary ‘Level 1’ legislation). The CEBS was also tasked with producing so-called ‘Level 3’ measures (non-binding guidelines, recommendations and standards) to encourage uniform domestic implementation and application of Level 1 and Level 2 legislation. A further task was to facilitate co-operation and information exchange between supervisors in matters relating to cross-border firms.

The CEBS chalked up some successes in its standard-setting capacity, issuing 12 sets of Level 3 guidelines by 2006. However, its tangible impact was not as substantial as initially hoped. In 2007, a review of the Lamfalussy system found that national supervisory regimes continued to differ along a myriad of dimensions including the frequency of on-site and off-site bank inspections, the predilection for principles-based versus rules-based approaches, the range of tools available to remedy instances of non-compliance and the attitude of supervisors towards bailing out troubled banks (Commission 2007b).

The CEBS's inability to produce more substantive convergence is at least partly attributable to the legacies of the informal network it succeeded. The CEBS relied heavily on the ‘buy-in’ of its members. Decisions were taken by consensus, even where its founding charter offered the possibility of using qualified majority voting. The Commission identified this consensus culture as promoting lowest common denominator solutions and a preference for generalized principles over prescriptive rules (Commission 2007b). In 2008, the option to resort to qualified majority voting was extended to all categories of decision-making, a comply-or-explain mechanism was introduced to encourage compliance with non-binding measures and a mediation mechanism was established to help resolve disputes. Yet the CEBS's ability to act independently in the supranational interest continued to be hindered by the interplay of divergent national preferences. While pre-crisis co-operation between supervisors was considered to be generally constructive and characterized by expert deliberation (author interview 30 October 2012; cf. Quaglia 2008a), areas of divergence were not always discussed openly at senior levels, leaving unresolved differences to be tackled by lower-level officials (author interview 1 March 2013). With the onset of the financial crisis, the tendency for supervisors to represent national interests increased (author interview 30 October 2012).

The most important banking sector legislation of this episode of reform was the 2006 Capital Requirements Directive (CRD). The CRD incorporated the Second Basel Capital Accord (Basel II) into EU law. While it aimed to minimize regulatory diversity across the EU, it included at least 152 options and discretions that provided flexibility for national authorities (CEBS 2008). Although many of these provisions had little economic significance, some were a source of competitive distortion. Examples included flexibility in relation to the definition of ‘credit institutions’ (and hence the scope of application of the rules); and the precise definition of Original and Additional Own Funds (de Larosière 2009). Flexibility also stemmed from the choice of legislative instrument: the use of directives meant that, as before, European rules needed to be transposed into national rulebooks before having legal effect. The CRD was thus layered on top of the Single Market framework. While it aimed to enhance regulatory uniformity, it did not fundamentally challenge the principles of minimum harmonization and mutual recognition. Scope for diversity between national regulatory regimes therefore remained.

The CRD also reinforced the discrete and autonomous role of supervisors within the regulatory process. In line with Basel II it revolutionized permissible methodologies for calculating capital adequacy requirements, permitting banks with sophisticated risk management capabilities to use internal credit-risk models to risk-weight their assets. Under this regime, the business of authorizing banks' internal models – ‘model validation’ – became a central element of prudential supervision. The CRD contained few legally binding provisions concerning this activity. By their nature, banks' internal credit-risk models are idiosyncratic, relying on information and processes peculiar to the individual banks themselves. As Blochwitz and Hohl (2011: 251) argue, supervisors’ ‘validation techniques need to be as individual as the rating system they are used for’. While the CEBS adopted guidelines on model validation in 2006 (CEBS 2006a), in practice national supervisory authorities developed different approaches in this area (author interview 31 May 2012).

The CRD (and Basel II) further reinforced the autonomy of national supervisors by including the ‘Supervisory Review Process’ as one of the three pillars of the capital adequacy framework. This process (‘Pillar 2’) requires supervisors to review banks' internal risk management and control functions to ensure their commensurability with the underlying financial risks that the banks are taking. As with model validation, the CRD left scope for national divergence in relation to Pillar 2, as recognized by the CEBS: ‘different supervisory authorities will use different types of evaluation processes … [with] differences in the emphasis on qualitative versus quantitative judgments and the degree of automation within a system’ (CEBS, 2006b: 26).

How supranational was the regulatory process during this episode of reform (see Figure 1)? Others have shown that financial lobbying became more Brussels-centric and favourable to market integration over the course of the 1990s and 2000s (Grossman 2004; Mügge 2010; Quaglia 2008b). This helped supranationalize the agenda-setting phase of the regulatory process, although the ECB's thwarted attempt to expand its supervisory powers is an indication that large member states could keep controversial issues off the negotiating table when it suited them. Negotiations over legislation relating to the Single Market had become more supranational after the 1992 Treaty of Maastricht, which enhanced the European Parliament's powers vis-à-vis the Council in the legislative process. Nevertheless, the key legislative acts in banking continued to be shaped by intergovernmental disagreements, with member states representing the positions of their national financial industries (Christopoulus and Quaglia 2009).

The incremental nature of the FSAP and Lamfalussy reforms ensured their impact on the ‘low politics’ of the regulatory process in banking was relatively slight. Organizational reforms converted the existing network of national supervisors into a formal committee, but the legacies of informal consensus-based co-operation inhibited this Committee's capacity to produce tangible supervisory convergence in practice. New directives expanded the scope and complexity of European rules, but were ‘layered’ upon the legal principles of the Single Market, thereby reinforcing the separation between centralized rules and decentralized implementation.

Exhaustion: the consequences of decentralized implementation

Policy-makers and practitioners have long recognized decentralized implementation to be a disincentive to financial integration (Commission 2005; DLA Piper 2009). However, this disincentive was not enough to prevent an overall trend of integration in European banking in the decade preceding the crisis. Interbank markets and capital-market-related activity were subject to the greatest internationalization, attributable largely to their proximity to the single monetary policy and associated infrastructural developments such as the integration of Large-Value Payments Systems (ECB 2013). Integration in corporate banking also increased, with cross-border bank lending to non-financial firms more than doubling between 1997 and 2009, albeit from a low base. Even retail banking – where cross-border expansion is inherently most difficult – showed signs of integration, with saving and borrowing rates converging across different countries and international merger and acquisition (M&A) activity increasing.

From 2008, these trends went into reverse. The proportion of banks' assets held in non-home country branches and subsidiaries declined, cross-border M&A activity dried up and the spreads between borrowing costs in different countries widened. Interbank lending declined by approximately 30 per cent between 2008 and 2012 (ECB 2013). To be sure, the patchwork of national regulations and the fragmented nature of supervisory authority were not the primary cause. The decline in interbank lending was above all a manifestation of the central vulnerability in the European financial system in recent years, namely, the vicious cycle of weak growth, fiscally weakened sovereigns and teetering domestic banks. Nor would a level playing field in European banking have prevented the mispricing of euro area sovereign debt or the inflation of asset price bubbles during the ‘great moderation’.

Having said that, decentralized implementation of European rules played an important role in both the genesis and propagation of the crisis. In 2011 the Commission noted that regulatory divergences were a source of legal uncertainty and had led to insufficient transparency and comparability of firms' financial situations (Commission 2011). Particularly egregious were variations in the definition of Tier 1 capital, which had caused both markets and regulators to lose confidence in this measure as a means of comparing banks' prudential positions. A further problem was that liquidity risk management – a core element of prudential regulation – remained to all intents and purposes beyond the scope of the CRD altogether. Though mutually recognized, liquidity regulations differed across the member states, reflecting different arrangements for central bank credit, deposit insurance, collateral policies and national regulators' risk appetite. Taken together, these divergences distorted competition, raised administrative costs and were a source of financial instability in their own right.

Arguably an even more serious problem was the misalignment between the incentives facing national authorities and the interests of the Union as a whole. During the crisis, markets and supervisors pressured banks to raise capital relative to risk-weighted assets. The easiest means by which banks could achieve this was to restrict lending and divest assets (i.e., to deleverage). Simultaneously, member states encouraged banks to maintain lending to domestic markets. In some cases, pressure was overt. For example, the UK Treasury openly exhorted banks to provide more loans to domestic small businesses as a part of its ‘Project Merlin’ initiative. In other cases, pressure was exerted behind the scenes. Notably, governments in some southern European member states are widely understood to have used ‘moral suasion’ to encourage banks to load up on domestic sovereign debt (Wiedmann 2013). This so-called ‘financial repression’ contributed to the fragmentation of the Single Market, as banks deleveraged more quickly overseas than domestically. It also reinforced the interdependence between banks and sovereigns.

Misaligned incentives also played a role in prolonging the banking crisis. Supervisors in countries with troubled banking sectors turned a blind eye to banks that chose not to write down non-performing loans, resulting in zombie banks continuing to operate with insufficient capital, unable to provide new lending to the real economy (ESRB 2012). Such ‘supervisory forbearance’ can reflect legitimate financial stability considerations: large-scale write-offs can result in bank insolvencies, which in turn can provoke self-perpetuating spirals of credit contraction, fire sales and falling asset prices. However, in the context of a generalized failure of trust between national supervisors during the most acute phases of the financial crisis, there has been a tendency to shield national champions from revealing losses, which could put them at a competitive disadvantage vis-à-vis their international competitors (author interview 1 March 2013).

From DE Larosière to the Banking Union

In recent years, policy-makers have reacted to the pathologies of decentralized governance by further supranationalizing the regulatory process. In 2008, the Commission convened a High-Level Group on Financial Supervision, chaired by Jacques de Larosière, to propose reforms to the supervisory framework. Building on the High-Level Group's recommendations, a new European System for Financial Supervision came into operation on 1 January 2011. It comprised three micro-prudential supervisory authorities, which converted the ‘Level 3’ Lamfalussy committees in the banking, securities and insurance sectors into executive agencies of the Commission. It also comprised a new European Systemic Risk Board (ESRB) responsible for macroprudential oversight.

The conversion of the CEBS into the European Banking Authority (EBA) marked another incremental step towards more supranational secondary rule-making (see Figure 1). Whereas the CEBS could only issue advice to the Commission in respect of ‘Level 2’ implementing legislation, the EBA is mandated with actually drafting technical standards, which the Commission may subsequently endorse. If the Commission decides to reject or amend an EBA standard, it must allow the EBA an opportunity to redraft the measures. Since objections by the Commission are rare, the EBA has in practice become the de facto producer of binding technical regulations for the EU banking sector.

As with its predecessor, the EBA must also promote co-operation and convergence in the day-to-day application of European rules by supervisors. Here, the EBA has few more powers than its predecessor.⁵ It has continued initiatives already underway within the CEBS, including staff training and secondment programmes and efforts to improve the efficiency of ‘colleges of supervisors’ (where supervisors from different jurisdictions co-operate in supervizing international banks). Latterly, the EBA has begun work on a ‘European Supervisory Handbook’ containing methodologies and best practices for day-to-day supervision. Overall, however, its work on supervisory convergence remains at an early stage.

The High Level Group also called for the ‘maximum harmonization’ of core prudential rules. This was partially realized with the adoption of the 2013 ‘CRD-IV Package’, which implemented the Third Basel Accord (Basel III) into EU law. The CRD-IV Package consists of two pieces of legislation. The Capital Requirements Directive IV (CRD-IV) contains rules concerning banks' rights of establishment and the responsibilities, powers and tools of prudential supervisors. As before, the CRD-IV must be transposed into national laws before having legal effect. The Capital Requirements Regulation (CRR) contains measures relating to Own Funds, including definitions of capital and methodologies for calculating capital charges. It also contains requirements on large exposures, liquidity ratios, supervisory reporting and public disclosure. In line with the maximum harmonization approach, the CRR does not require transposition into domestic rulebooks.

Maximum harmonization under the CRR diminishes opportunities for gold- and tin-plating. It is yet another step towards a more supranational regulatory process, even if intergovernmental disputes led to some rules being watered down (Howarth and Quaglia 2013). However, the CRD-IV Package did not alter the basic governance framework in which centralized European rules are implemented by national supervisors. Moreover, the legislation contains more than 140 provisions that invite national authorities to exercise discretion (Enria 2013b). While many of these provisions are necessary to account for diverse nature of national banking systems, others could become a source of competitive advantage and even financial instability.

The most important example is the flexibility granted to national authorities in relation to macroprudential policy. The CRD-IV requires national authorities to set a countercyclical capital buffer and a systemic risk buffer, both of which can be up to 2.5 per cent of banks' risk weighted assets. The CRR also permits national authorities to impose a range of other macroprudential tools, such as higher risk weights for real estate assets. There are strong financial stability arguments in favour of macroprudential policies being applied in a country-specific manner (Kincaid and Watson 2013). Inevitably, however, national flexibility could be used for non-prudential ends, such as protecting national champions or ‘trapping’ capital within the home jurisdiction. The CRD-IV Package provides the Commission and the ESRB with limited oversight powers in respect of member states' use of these instruments, although it is unclear whether the latter has either the institutional independence or analytical resources necessary to be effective in this role (McPhilemy and Roche 2013). The ECB will also be involved in macroprudential decision-making for countries participating in the SSM (see below). Still, as currently constituted, European rules provide member states with considerable flexibility in these areas.

The regulatory process in banking is undergoing further transformation with the creation of the banking union. Under the SSM, the ECB will have overall responsibility for the supervision of some 6,000 euro area banks.⁶ For approximately 130 ‘significant’ banks, the ECB will assume direct responsibility for all key supervisory tasks, including inter alia granting and withdrawing authorizations, ensuring compliance with relevant union law and carrying out the Supervisory Review Process (Regulation 1024/2013 Articles 4, 6). ‘Joint Supervisory Teams’, composed of officials from the ECB and national authorities, will supervize each significant bank. Supervisory decisions will be prepared by a Supervisory Board consisting of ECB officials and one senior supervisor from each national authority. Ultimate decision-making responsibility will rest with the ECB Governing Council. While national competent authorities will supervize less significant banks (under a system of ECB oversight), the ECB will have the option to take over the supervision of any bank at any time.

One prominent policy-maker described the SSM as a ‘greater pooling of sovereignty than signing up to the Euro’ (Bowles 2013). While this may be an overstatement, the SSM greatly diminishes opportunities for home bias in banking supervision. It may therefore be regarded as the most important single reform of the regulatory process in banking to date. Having said that, some cross-national divergences are likely to persist. Obviously, non-euro area countries that choose not to participate in the banking union could continue to adopt divergent supervisory practices. The EBA has been tasked with promoting uniformity between participating and non-participating countries, in particular by publishing the aforementioned Supervisory Handbook. However, it is unclear why the EBA should be any more successful in producing supervisory convergence than it has been in the past. Its powers in relation to supervisory convergence remain non-binding. More generally, the scarcity of its resources relative to the ECB and the Bank of England (which will constitute twin poles of supervisory power in the emerging regime) may impede its ability to achieve substantive convergence between them.

Even within the SSM, certain divergences could persist. As noted above, the Single Rulebook contains many options and discretions, most importantly in the area of macroprudential policy. The ECB may impose stricter macroprudential requirements than those set in participating countries, but it cannot require a member state to lower or remove a macroprudential measure (Regulation 1024/2013 Article 5). Thus, ECB-led Joint Supervisory Teams could be forced to apply different rules to the banks under their jurisdictions, depending on where those banks are headquartered (Enria 2013b).⁷ Ultimately, these possibilities arise because the SSM is being superimposed upon a regulatory framework that even now is composed of mutually recognized, but ultimately different, national jurisdictions.

Conclusion

Prior to the financial crisis, scholarship on European financial integration underplayed the scope for – and potential significance of – national discretion in the implementation of European rules. This article has demonstrated that from the introduction of the Single Market framework through to the onset of the financial crisis, the ‘low politics’ of implementation remained less integrated than the ‘high politics’ of agenda setting and negotiation. Though it was well understood that divergent national rulebooks and different styles of supervision could distort competition and act as administrative barriers to cross-border integration, few scholars, policy-makers or practitioners recognized the adverse implications that such diversity would have for financial stability, especially in the context of a financial crisis.

Figure 1 demonstrated that change in this policy field has occurred incrementally, with supranationalization of different phases of the regulatory process occurring at different moments in time. It is worth noting that the rightward movement of each phase of the regulatory process has in reality taken place in an even more incremental manner than depicted in this figure. The various episodes of reform identified in this article were not discrete events, but rather open-ended cycles of agenda-setting, regulatory reform and behavioural change. The preceding analysis indicates that a two-way ‘vertical’ interdependence exists between different phases of the regulatory process. Negotiations over primary legislation set the parameters for uniformity or divergence at the implementation stage. Simultaneously, the consequences of (divergent) implementation feed back into the agenda-setting phase, preparing the ground for future rounds of negotiations over primary rules.

With the benefit of hindsight, the reforms initiated in the late 1990s and early 2000s may be regarded as a staging post in the transformation of the regulatory process in banking. While the Single Rulebook and SSM have shifted the regulatory process further towards to supranational pole of the continuum proposed above, this article has suggested that national authorities will continue to exercise discretion in important areas of secondary rule-making, such as macroprudential regulation. Based on the foregoing analysis, it is difficult to predict whether the idealized end-point of outright supranational banking regulation will ever be reached. A potentially fruitful avenue for future research into this question would involve turning from the vertical interdependences between different stages of the regulatory process to the horizontal interdependences between prudential regulation and related public policy functions at the national and EU levels. Some such functions, including rules relating to state aid and facilities for lending as a last resort in the euro area, are already supranational competencies.⁸ Others, such as bank resolution and crisis management, are in the process of passing to the EU level, albeit in a piecemeal fashion characterized by ‘variable geometry’. Others still, including deposit insurance and conduct of business regulation, appear set to remain in hands of national authorities for the foreseeable future. Mapping the balance of national and supranational competencies in these areas – and their linkages with prudential regulation and supervision – is important for understanding what is at stake in contemporary contests between national and supranational authorities in European banking.

ACKNOWLEDGEMENTS

I would like to thank Andre Broome, David Howarth, Michele Chang and the four anonymous reviewers for their constructive input and criticisms. This work was supported by the Economic and Social Research Council (grant number ES/I022589/1).

Additional information

Biographical note

Samuel McPhilemy is a doctoral researcher at the University of Warwick, UK.

Notes

1 Another reason why the SSM cannot be explained as merely the quid pro quo for burden sharing is that the former was agreed in the absence of any firm agreement on the latter. At the time of writing, member states have agreed to only a highly limited form of shared resolution funding.

2 Named after Baron Alexandre Lamfalussy, chairman of the 2001 ‘Committee of Wise Men’, which proposed the framework.

3 Following Streeck and Thelen (2005), institutions are conceived here in a deliberately narrow sense as formal rules specifying permissible and impermissible behaviour.

4 Additionally, a new ‘European Banking Committee’ succeeded the Banking Advisory Committee in 2005. Composed of national finance ministry officials, this ‘Level 2’ committee assists the Commission in adopting implementing legislation.

5 In ‘emergency situations’, the EBA may exercise enhanced powers, including the power to direct national authorities and individual firms. These powers have never been used because no authority has ever deemed it prudent to declare an emergency.

6 Other member states may participate by concluding close co-operation agreements.

7 As foreseen in Regulation 1024/2013, Article 4(3).

8 Seikel (2013) examines the role of the Commission's Directorate General for Competition in spurring regulatory integration.