‘Don’t Crunch My Credit’: Member State Governments’ Preferences on Bank Capital Requirements

Across Europe, banks remain, to this day, the main suppliers of finance to the European economy, but also a source of systemic risk. As such, regulating them requires that policymakers find an appropriate balance between restricting their risk-taking behaviour and increasing lending to support economic growth. However, the ‘varieties of financial capitalism’ that characterize national banking sectors in Europe mean that the adoption of harmonised capital requirements has different effects across countries, depending on the country-specific institutional setting through which banks provide lending to the national economy. This article conducts a new analysis of Member State governments’ positions in the post-financial crisis reform of the EU capital requirements legislation, expanding the scope of previous studies on the topic. Here, I examine in detail the positions of Member States on a wider set of issues and for a broader set of countries than the existing literature. Building on the varieties of financial capitalism approach, I explain these positions with regard to structural features of national banking sectors. I find that Member State governments’ positions reveal a general agreement with the proposed increase of bank capital requirements, while seeking targeted exemptions and preferential treatment that they deem necessary to preserve their domestic supply of retail credit.


Introduction
On 27 March 2020, the Basel Committee on Banking Supervision, the international standard-setter for banks' capital requirements, announced the deferral of implementation deadlines of the Basel III frameworkadopted in response to the global financial crisis of 2008-to ensure "that banks and supervisors are able to commit their full resources to respond to the impact of Covid-19" (Basel Committee on Banking Supervision, 2020). As for the EU transposition of the final elements of the international standards, it seems that the European Commission (EC) has put on ice the legislative proposal amending the Capital Requirements Directive (CRD) and Capital Requirements Regulation that it was supposed to issue in June 2020.
This reaction to the emerging economic fallout of Covid-19 suggests policymakers, first, consider that banks should play an important role in fostering economic recovery, and second, fear that the planned tightening of capital requirements may be incompatible, in the short term, with said bank support of the real economy. How EU Member States face this perceived short-term trade-off is of particular importance in the context of Economic and Monetary Union. Economic and Monetary Union reforms in recent years (see e.g., Rehm, 2021) relied on the assumption that banking regulation-notably capital requirements-would, by reining-in banks' excessive leverage, contribute to reducing the interdependence that tied together banks and public finances in a vicious circle and wreaked havoc on several Member States during the sovereign debt crisis (Merler & Pisani-Ferry, 2012). Nevertheless, the EU's transposition of the early parts of Basel III (known as CRD-IV) was criticised for watering down the international framework (Véron, 2013). Previous international political economy accounts of the negotiation have attributed much of this dilution to some Member States' demands for limiting the increase of capital requirements in order to protect the competitiveness of their respective banking sectors, but also to preserve shortterm economic growth (see e.g., Howarth & Quaglia, 2013, 2016a.
This article pursues two objectives. The first is to provide a new examination of Member State positions on the CRD-IV reform, analysing a larger sample of countries than previous studies and delving into the technical detail of positions on a series of issues, some of which have not been examined before. Analysing the responses of fifteen Member States' national authorities to three EC preparatory consultations, I find that, instead of clear general preferences for tighter rules on bank capital or conversely, a general forbearance, each Member State's requests for preferential treatment focus on very specific instruments and institutions amid a general agreement with the necessity to increase bank capitalisation.
The second objective is to explain these particular positions. Important literature on Basel III and CRD-IV suggests that the lobbying of the banking industryin particular by large, international banks-significantly shaped the debate on post-crisis capital requirements (e.g., Lall, 2012;Young, 2014). However, while many of the requested changes did benefit large banks, Member State positions and the wish list of international banks differ in important ways. International political economy, in turn, suggests that the 'varieties of financial capitalism'-that is, the country-specific institutional settings that characterise banking sectors-that coexist in Europe mediate Member State preferences on financial regulation (Story & Walter, 1997). Among relevant factors, previous studies have notably highlighted the role of bank capitalisation levels and bank-industry ties (Howarth & Quaglia, 2016a) and different degrees of foreign ownership (Spendzharova, 2012) in national banking sectors. Here I argue that, while these are relevant factors, in order to account for the detailed amendments the Member States requested, we must also consider the qualitative composition of banking sectors and the types of instruments on which retail lending relies.
The next two sections present the analytical framework (2) and methodological approach (3) of the article. I then examine Member State positions on CRD-IV, highlighting the conflictual issues and suggesting variables that explain these conflicts (4). I then discuss these findings in terms of 'varieties of financial capitalism' (5) and conclude (6).

Analytical Framework
International political economy has long framed policymakers' preferences on banking regulation as a 'dilemma' between two conflicting goals: financial stability through strict capital requirements and international competitiveness through reducing the cost of regulation of national banks (Kapstein, 1989;Singer, 2004). The economic downturn that followed the 2008 financial crisis added short-term economic growth to the list of concerns: Policymakers perceived that "trade-offsperceived or real-might still have to be made and notably between financial stability and economic growth because, ceteris paribus, banks need to deleverage-and thus shrink their lending-to improve their capital position" (Howarth & Quaglia, 2016a, p. 206). There is however no consensus among economists about the relation between capital requirements, credit supply, and economic growth, and while in the short-to-medium term higher capital requirements are expected to increase the cost of credit for borrowers (Macroeconomic Assessment Group, 2010), higher capitalisation levels are likely to bring net long-term benefits in terms of GDP growth (Admati & Hellwig, 2013).
Here I assume that policymakers were aware of these debates as well as of the short-term costs and long-term benefits associated with higher capital requirements, but still perceived that a trade-off needed to be made between the long-term objective of a resilient banking sector and the short-term objective of maintaining a steady flow of credit to fight off the post-financial crisis recession. Whether, on a particular issue of banking regulation, Member States favoured one or the other depends, I argue, on the structural features of their national economies and banking sectors and the extent to which the proposal was likely to affect the supply of credit to the national real economy, particularly SMEs and households. This analysis then builds on the 'varieties of financial capitalism' approach (Story & Walter, 1997) and seeks to complement previous accounts of the CRD-IV negotiations.
In their respective works, Quaglia (2013, 2016a) and Spendzharova (2012) have put forward three explanatory factors to account for Member State positions. Howarth and Quaglia explain the conflict between the Franco-German tandem and the UK on the level of minimum capital ratios in terms of systemic patterns of bank capital (different levels and composition) and bank-industry ties (degrees of real economy reliance on bank credit). Spendzharova, focusing on Central and Eastern European (CEE) countries, shows how the predominance of foreign ownership in those countries' banking sectors made their governments fearful of foreign banks depleting local subsidiaries in order to repatriate funds to the home country in case of trouble.
Following a similar approach, I argue that, in order to account for the specific exemptions and preferential treatments the Member States requested, we must also take into account the types of banks that dominate each country's banking sector and the particular instruments on which they rely to supply credit to the real economy. Indeed, banks of different sizes (small local banks vs. large banking groups) and legal forms (joint stock vs. cooperative, mutual, savings, and public banks) which rely on different sets of financial instruments would be affected in very different ways by the Basel III rules. Where each country stands in relation to these structural factors is then likely to shape in important ways how their common double preference for stability and growth translates into positions on specific policy proposals. This is not to say that Member State positions are fully determined by economic and banking sector structures-the different levels of politicization (Högenauer, 2021), as well as different sets of valuebased ideas (van Loon, 2021) of financial regulation issues across Member States, also contribute to shaping positions-but that these largely determine the material interests at stake in capital requirements. The analysis presented in this article should thus be seen as a complementary contribution to the fruitful research agenda on national preference formation about international financial regulation. The next section will detail which countries constitute the sample, as well as the CRD-related issues chosen for analysis. Section 4 will then outline, for each of the six selected issues, the positions adopted by Member States.

Methodological Approach
The focus of this article on Member State governments is justified, I believe, first by the central role that governments play in the policymaking process for capital requirements at the international and European level, and second by the fact that they remain, ultimately, responsible for macroeconomic stabilisation. I choose to extract Member State governments' positions on reform proposals from the written comments they submitted in response to three public consultations conducted by the EC in 2008EC in , 2009EC in , and 2010. These documents have the advantage (when compared to collecting positions through interviews or a review of press coverage) that they emanate directly from the national representatives involved in the negotiation, offering a detailed view of positions which have not been subject to any posthoc reinterpretation. Furthermore, because they all respond to the same set of EC questions, they facilitate the crosscountry comparison of positions on a given set of issues.
The period 2008-2010 corresponds to the preparatory works for the EC's 2011 CRD-IV proposal, which the EC conducted in parallel to the elaboration of the Basel III standards. In this article, I limit the analysis to six broad issues: definition of capital, large exposures, liquidity standards, leverage ratio, treatment of mortgage loans, and supervisory arrangements. These constitute only a subset of all the issues consulted during the period but were selected for the potential of conflict among Member States on the degree of stringency vs. leniency and the degree of harmonisation vs. national discretion that the new framework should permit.
15 EU Member States are analysed (see Table 1). The selection includes all the Member States whose government (Treasury department) submitted an answer to at least one of the three consultations. 14 out of the 27 EU Member States provided comments at the time, but of these, I excluded Slovenia and added Italy and Spain. The 2008 Slovenian response did not address any of the substantial issues raised by the consultationonly one minor technicality-and could not be used to extract positions. Although there were no Italian or Spanish government responses available, I include these two countries using responses by their respective central banks as a proxy, supplemented by a review of finance ministers' public statements. With these additions, the selected countries constitute a representative sample of EU Member States, including in particular both large and small banking sectors with a variety of banking sector structures. We should note that the absence of published comments from a government does not imply that it takes no position: A government may have required that its comments not be published or may have used another way, other than the consultation, to convey its views on the proposals (e.g., Council meetings). For reasons of comparability across different methods for collecting positions, I chose to limit the analysis to countries for which responses were available, Italy and Spain constituting the only exceptions which were partly compensated for by their central banks' responses.
To analyse positions, I first extracted from each document the sections devoted to each of the six issues and summarised them. The Supplementary File provides the reader with this summary of each country's position for each of the six issues. In a second step, I applied a "constant comparative method" (Glaser & Strauss, 1967, pp. 101-116) to identify similarities and differences across responses, thereby identifying recurrent themes and oppositions. The result of this process is presented in Section 4.

Definition of Capital
At its core, the Basel framework defines how much of a bank's assets (its various investments and the loans it distributes) must be funded through financial instruments that contractually are able to absorb potential losses arising from borrowers defaulting or bad investments both during the life of the bank ('going-concern') and in case of failure ('gone-concern'). These loss-absorbing instruments constitute banks' 'capital.' Regulatory capital is broader than the equity held by its shareholders, and also include a series of debt securities. Defining bank capital then implies listing the instruments that are sufficiently loss-absorbent to be part of the capital base, which in Basel III, is divided into three buckets: common equity tier 1 (CET1), the most loss-absorbent and broadly corresponding to common shares or equivalents; additional tier 1, which includes debt instruments that can be written-down to absorb exceptional losses on a goingconcern basis; and tier 2, which includes debt securities to be written down only in case of failure. Furthermore, 'prudential adjustments' have to be made to amounts of eligible instruments to account for particular situations that may make part of the capital base unavailable to absorb losses.
On eligibility criteria, the most recurrent theme regarded the limitation of CET1 to common shares. Pre-crisis, Member States could adjust the CRD rules to local specificities in the national transposition, thus definitions of core capital varied importantly across countries. The harmonisation on a common shares model would significantly affect banking sectors where non-joint stock banks (the various forms of banks whose core capital is not composed of traditional public listed shares, notably cooperatives, mutuals, savings banks and a number of public banks) are important actors, since these banks would have to either change their legal structures to meet the new requirements or disappear. The countries calling most forcefully in defence of non-joint stock banks' capital instruments were, unsurprisingly, those where non-joint stock banks represent a large part of the banking sector: Austria and Germany above all, followed by Finland and France. In 2016, more than half of the Other Systemically Important Institutions (O-SIIs)-that is, domestic systemically important banks-in those countries were either public banks (e.g., several German Landesbanken), or the central institutions of cooperative and savings banks (e.g., Austria's Raiffeisen Bank International, France's Groupe Crédit Mutuel, or Germany's DZ Bank, see Table 2), which shows their importance not only in terms of their size but also in terms of their centrality in the domestic economy. Illustrating the cost of harmonisation, Germany also made a plea for temporarily maintaining the possibility to include in tier 2 cooperative bank members' uncalled commitments which until then had been allowed under German law but excluded under Basel III and which constitute an important part of German cooperative banks' capital. By contrast, those countries that have no non-joint stock bank among their O-SIIs did not insist on the issue and merely mentioned the need to make the criteria compatible with different legal structures. The 2010 reform of the important Spanish Cajas sector, which transformed them into jointstock banks largely explains why Spain did not voice concerns on this issue. On prudential adjustments, the full deduction of 'minority interests' (capital instruments held by minority shareholders of a banking group subsidiary) was opposed by a diverse set of countries: Austria, Czechia, Denmark, Finland, France, Hungary, Italy, Slovakia, and Spain. The deduction would affect banking groups by reducing the contribution of subsidiaries to groups' 'consolidated' (i.e., aggregate) amounts of capital. For Austria, France, Italy and Spain-home to several internationalised banking groups-important amounts of minority interests (see Table 3) reflect a strategy to raise capital for the group through subsidiaries. Considering the generally low levels of bank capitalisation in those countries, minority interests were then to constitute an important resource to meet the increased capital requirements. Similarly, France-the land of the bancassurance model of financial conglomerates-forcefully opposed the deduction of investments in insurance subsidiaries which would also have impacted the capital ratios of all its major banking groups (International Monetary Fund, 2011). By contrast, the UK's large banks, being better capitalised than their continental peers (HSBC, Lloyds and Barclays all had above 10% of CET1 capital at end-2010, to be compared to 8.1% for France's Société Générale, 7.8% for Italy's UniCredit and 7.1% for Spain's Santander; European Banking Authority, 2011), did not need to rely on minority interests. Czechia, Hungary, and Slovakia, in turn, are in this debate hosting the subsidiaries raising minority interests (see Table 5) and highlighted in their comments the risk that the deduction would create an incentive for groups to undercapitalise local subsidiaries.

Large Exposures
So called 'large exposures' are a bank's exposures to a single client or group of connected clients that could put the bank's solvency at risk in case of that client failing to repay. Limits on large exposures existed in the pre-crisis CRD to penalise such exposures but included a number of options for Member States to grant exemptions, in particular to intra-group (between entities of the same banking group) and certain interbank (between two independent banks) transactions. In 2008, the EC suggested strengthening the regime and consulted on withdrawing options and exemptions. Limits on intra-group transactions are especially relevant for banking groups, as they limit their freedom to shift capital and liquidities from one group entity to another. Limits on interbank transactions are crucial for decentralised banking networks (those where members of the network are independent of each other but share a brand and some central institutions, for example, the German Sparkassen) inasmuch as they impact liquidity management within the network as well as more generally for banks' daily liquidity management, since banks may need to borrow or lend large amounts on the interbank market.
Among the responding countries to the 2008 consultation, we find two overlapping groups supporting a more lenient regime. One was composed of the countries whose banking sector includes important decentralised banking networks and was eager to maintain exemptions for claims on central institutions of decentralised banking networks and on transactions where both parties are part of a joint risk-management or institutional protection scheme, which usually is the case of decentralised banking networks. The second group includes countries that are home to large banking groups and called for maintaining the options to exempt intragroup transactions between entities submitted to the same consolidated supervision. Austria is part of the first group; Denmark, France, Spain, and Sweden of the second; Germany and Finland are part of both. Furthermore, countries with highly concentrated banking sectors, France and Sweden, expressed concerns about liquidity management and a possible destabilisation of the interbank market unless further exemptions were made. Finally, Poland and Czechia joined Austria, Sweden, and Germany in welcoming the exemption for smaller transactions. Conversely, Czechia and Slovakia, two countries with foreign-dominated banking sectors (see Table 5), called for maintaining the national discretion to impose more restrictive limits on large intragroup transactions. This discretion was necessary, they argued, to prevent local subsidiaries from being exposed to the failure of group entities in other Member States. Sweden, conversely, strongly opposed such discretion, warning that national authorities could use it for ring-fencing at the expense of efficiency.

Liquidity Requirements
Liquidity standards were discussed in the 2010 consultation. Few countries had liquidity requirements in place before the crisis and there were none in international or European standards before Basel III and its transposition. Liquidity standards apply essentially on the assets side of banks' balance sheets: They require banks to hold reserves of 'liquid' assets, that is, assets that can be sold for cash immediately, even in times of crisis, without incurring any significant loss. While the liquidity coverage ratio aims to ensure that banks maintain a liquidity buffer sufficient to withstand a one-month-long market stress, the net stable funding ratio requires banks to match their long-term lending commitments with corresponding long-term funding sources.
In relation to the liquidity coverage ratio, the main issue was listing the assets liquid enough to be included in the buffer, the so-called 'high-quality liquid assets' (HQLAs). Initial proposals essentially restricted eligibility to government bonds and stable deposits, a position supported by the UK and Estonia, but opposed by most other responding Member States (Austria, Czechia, Denmark, Finland, France, Germany, Hungary, Ireland, Italy, and Sweden), who called for the larger inclusion of additional assets. This conflict can easily be understood by looking at levels of liquid asset holdings across countries ( Table 4). The British and Estonian banking sectors were still in 2014 (earliest data published by the European Central Bank) the ones with the highest share of liquid assets in total banking sector assets (above 30%), and Estonia-unlike the other countries-already had tight liquidity requirements in place before the CRD-IV reform. Conversely, almost all the proponents of a more inclusive HQLA buffer (Austria, Czechia, Denmark, Finland, France, Germany, Hungary, Ireland, and Sweden) had ratios of liquid assets to total assets below 20%, and some far below (Austria). For those countries, a liquidity coverage ratio with a narrow HQLA definition would force banks to massively shift assets away from less liquid but productive assets, typically those funding the real economy.
For the pro-inclusion countries, HQLAs should notably include more covered bonds. Covered bonds are a particular form of securitisation where the pool of securitised assets is, in most cases, restricted to mortgage loans. These market-based assets developed at an exceptional rate across Europe since the early 2000s to supplement insufficient deposits in meeting mortgage lenders' funding needs (Johnson, Isgrò, & Bouyon, 2016, p. 7). Denmark and Sweden were the most vocal on this issue, stressing the stability of covered bonds through the financial crisis, their importance for (mortgage) banks' funding and the likely destabilising effect on covered bond and mortgage markets should they be excluded. In Denmark, the entire mortgage credit system-the defence of which in CRD-IV was "absolutely central" for finance minister Brian Mikkelsen ("Minister diskuterer," 2010)-relies on covered bonds. The fact that the EC specifically asked about covered bonds in its consultation is already evidence of their importance in European banking. As can be seen from Table 4, covered bonds constitute an important source of funding for banks, upon which they increasingly relied through the crisis years. Germany's stance on the issue should, for instance, be seen in light of the fact that their reliance on the stable Pfandbriefe market enabled German savings and cooperative banks to maintain lending levels through the crisis (Hardie & Howarth, 2013a). The Baselproposed cap on covered bonds in HQLAs would have depressed market demand for these assets, drying up an important source of refinancing for mortgage loans. By contrast, Spain's large covered bond market, which made possible the Cajas' frenzy of real-estate lending (Royo, 2013), was bound to adjust, which may explain the Banco de España's silence regarding their inclusion in HQLAs. Countries with important non-joint stock banking sectors also called for different types of preferential treatment for them. Germany called for the inclusion of "debt securities fully guaranteed by sovereigns or…securities of promotional banks under public ownership" (Bundesministerium der Finanzen, 2010, p. 2; see Supplementary File), that is, securities issued by its Landesbanken. Austria, Poland, and Slovakia, which all have cooperative or savings banks' central institutions among their systemically important institutions, asked for cooperative banks' deposits in their central institutions to be recognised as 'stable,' therefore contributing more to these institutions' stock of HQLAs.
On the net stable funding ratio, only Estonia defended a more conservative treatment than that proposed by the EC. All the other respondents to the 2010 consultation warned of its potentially destabilising effect on lending. Indeed, since it requires banks to balance the maturity of their liabilities and assets, it effectively forces banks to either reduce their reliance on short-term wholesale funding or limit their lending to and investments in long-term assets, notably loans to corporates and households, that is, the real economy. The UK authorities (HM Treasury & Bank of England, 2010, p. 7; see Supplementary File) thus warned that the ratio "could significantly disadvantage SME and retail loans relative to lending to large highly-rated corporates." It then sided with Austria, Germany, and Slovakia in calling for more favourable treatment of retail lending in terms of the stable funding required. Regarding the provision of stable funding, countries with important networks of independent cooperatives (Austria, Germany, but also Hungary, Poland, and Slovakia) called for preferential treatment of these banks' deposits with their central institutions. Calls for preferential treatment of covered bonds were also made, in particular by Austria, Denmark, France, and Germany.
The choice of the level of application (entity-level or consolidated level) and the proposal to shift the supervision of cross-border branch liquidity to the homecountry supervisor were two issues marked by opposition between CEE Member States plus the UK, and the other governments. While the latter supported shifting decision-making power to the home-country supervisor (supervising the group) on liquidity issues, the former insisted on preserving the freedom of the hostcountry supervisor (supervising a subsidiary) to impose the respect of liquidity coverage ratio and net stable funding ratio at the level of branches and subsidiaries. Observing the varying degree of foreign ownership in national banking sectors (Table 5) helps make sense of this divide: CEE banking sectors are characterised by a dominance of foreign banks, which own between a third and nearly all of total banking sector assets, and foreign banks constitute a major source of systemic risk in those countries, where they represent the majority of O-SIIs. The British banking sector is in a similar situation of exposure to foreign banks' systemic risk, with 12 out its sixteen O-SIIs being foreign-owned. By contrast, those countries supportive of home-country supervision are predominantly-Ireland being the exception-home to internationalised banking groups and little exposed to foreign banks.

Leverage Ratio
A leverage ratio requirement was a novelty introduced with Basel III: It is intended to act as a complement to risk-based capital requirements by setting a maximum nominal amount (not risk-weighted) of assets that a bank can acquire with its capital base. The most controversial issue was whether the new requirement should be a binding minimum (Pillar 1) or an indicator upon which supervisors could impose additional capital requirements if necessary (Pillar 2). A binding leverage ratio was expected to particularly affect undercapitalised banks, but the risk-insensitiveness of the measure was also expected to put relatively safer banking activities, notably traditional deposit-taking and retail lending, at a disadvantage: under the leverage ratio, they would 'cost' as much capital as riskier activities through yielding less income. The Swedish authorities (Regeringkansliet, Finansinspektionen, & Sveriges Riksbank, 2010, p. 4; see Supplementary File), for instance, thus considered it "important that a leverage ratio is not designed and calibrated so that it endangers the supply of mortgage credit to Swedish households." Of the respondents to the 2010 consultation, only the UK unambiguously argued in favour of a binding ratio. At the other extreme, France and Germany forcefully rejected the proposal, denouncing its likely unintended effects on bank lending. All the other respondents (Austria, Denmark, Estonia, Finland, Hungary, Poland, Sweden) argued for an introduction in Pillar 2. Beyond average leverage levels across countries (Table 6), understanding the opposition requires one to consider the parallel effect of proposals on the definition of capital, notably the deduction of minority interests and investments in insurance subsidiaries (see above) that would reduce the capital base of continental European banks more than that of their British competitors.

Treatment of Mortgage Loans
The 2009 and 2010 consultations contained proposals to reform the prudential treatment of mortgage loans (loans that are guaranteed by commercial or residential real estate) and in particular the conditions for granting them a preferential treatment under the form of a reduced 35% risk weight to part of the loan (i.e., only 35% of the amount would count towards the bank's riskweighted assets). The pre-crisis framework gave Member States an important degree of discretion to decide which loans could benefit from the preferential treatment. The EC proposed setting a harmonised condition under the form of a maximum loan-to-value ratio: The preferential risk-weight could be applied to the lent amount only up to a certain threshold relative to the value of the mortgaged real-estate property (40% in the 2009 proposal, 80% in 2010); the remaining amount would be applied a much higher risk weight (1,250%) in order to discourage lending to highly leveraged clients.
Respondents to the 2009 consultation unanimously rejected the proposed 40% loan-to-value ratio, denouncing its likely impact on mortgage credit supply. Indeed, the proposal would have led to most mortgage loans being more costly for banks (more regulatory capital), who would pass the extra cost to clients. In 2017, across the sample of countries, mortgage loans represented on average 42.82% of all bank loans and advances (European Central Bank's Statistical Data Warehouse), ranging from 18.35% (France) to 61.62% (Estonia). The emergence of mortgage lending in Europe since the 1990s owes a lot to favourable legislation (Johnson et al., 2016) and has become an essential instrument for home ownership. In 2017, more than a third of homeowners had a mortgage in Denmark, France, Finland, Sweden, and the UK, with CEE markets are quickly catching up (European Mortgage Federation, 2019, p. 40). The cost increase would then affect the masses, which may explain why even in a country like Spain-where a real-estate bubble brought about a banking crisiswas reluctant to increase requirements on all mortgages (Banco de España, 2009see Supplementary File). In 2009, the EC also suggested tightening specifically the treatment of mortgage loans denominated in a foreign currency. The issue was taken up only by the three responding CEE countries: Czechia, Estonia, and Hungary, who criticised the harshness of the proposals, whereas Austria welcomed them. Estonia and Hungary notably called to differentiate loans denominated in euros from loans in other currencies, the exchange rate risk being lower with the former.
The 2010 proposal for an 80% loan-to-value, more in line with industry practices, was more welcome. However, all respondents rejected the proposal to align the treatment of residential real-estate mortgages on that, more demanding, of commercial real-estate mortgages. The heterogeneity of European real-estate markets sparked calls from Denmark, Germany, Poland, Sweden, and the UK to maintain a certain degree of national discretion. The German government (Bundesministerium der Finanzen, 2010, p. 24; see Supplementary File) thus invoked the "particular importance of RRE [residential real-estate] financing" in its call to retain existing options. Only France, whose banks rely comparatively less on mortgages and which have large foreign retail activities, explicitly welcomed full harmonisation.

Supervisory Arrangements
Proposals regarding the degree of freedom granted to national authorities-legislator and supervisor-to adapt EU standards to banks active in their jurisdiction saw a clear opposition appear between 'home' and 'host' countries. The EC notably consulted in 2008 on 'colleges of supervisors' for cross-border banking groups. Czechia, Hungary, Poland, and Slovakia (the four 'host' countries; see Table 5) responded: First, by forcefully defending guaranteed rights for host-country supervisors to participate in colleges, against the proposal to leave the homecountry supervisor to decide on the composition and, second, they called for the limiting of colleges' decisionmaking powers, not to impinge on host-country supervisors' competences. Among these 'home' countries the positions varied: Austria and Finland agreed on the issue of composition, while France called for granting a strong decision-making role to colleges and an important role for the consolidating supervisor within them. In 2008, the criteria for designating branches of foreign banks as 'systemically important' were also discussed. The EC proposed additional rights for host-country supervisors, which Slovakia and Poland explicitly welcomed, although Poland called for a lower threshold (branch deposits to total banking sector deposits) for considering a branch as systemically relevant. Conversely, Germany opposed shifting additional branch supervision powers to hostcountry supervisors, and Finland and Sweden opposed any threshold lower than 5% of a national banking sectors' total deposits.
The 2009 and 2010 consultations furthermore suggested the removal of most of the existing options and national discretions in the CRD and the maxi-mum harmonisation of Pillar 1 requirements across the EU. This move to maximum harmonisation would deprive national authorities of the possibility to adapt European standards to local circumstances. France was the most vocal supporter of maximum harmonisation, which Denmark, Finland, and Germany also welcomed. Austria and Ireland equally supported the removal of options and national discretions, with the exception of real-estate. Conversely, Estonia, Hungary, Poland, the UK (which I consider as a 'host' country due to the importance of foreign O-SIIs in its banking sector), but also Spain and Sweden rejected maximum harmonisation for the sake of financial stability, doing so both individually (in their responses to public consultations) and collectively in a letter to Commissioners Michel Barnier and Olli Rehn (Djankov et al., 2011).
As with liquidity requirements and large exposures, CEE countries' and the UK's opposition to transfers of supervisory competence and reduction of national discretion appear motivated by the need to ensure against the systemic risk posed by the important operations of foreign banks within their jurisdictions. The link that CEE responses establish between national discretion and national responsibility for financial stability (e.g., Ministry of Finance  Spendzharova's (2012, p. 319) observation that these governments "were not apprehensive about transferring power to the supranational level per se. They did worry, however, about the fiscal and accountability consequences." Czech finance minister Miroslav Kalousek thus stated in May 2012: "There was a danger that the bank's regulator abroad would have more power over banks than the Czech supervisor….This could mean that parent banks could vacuum the Czech branches" ("EU: Na banky," 2012). Spain and Sweden, conversely, are among the countries least exposed to foreign banks, and their particular opposition to maximum harmonisation (but not to home supervision) finds its roots in their respective choice to increase capital requirements nationally to fight off domestic banking crises (in Sweden in the 1990s and in Spain with the Cajas from 2009).

Discussion of Results
We can already see governments' will to find a compromise between reducing bank leverage and preserving retail lending in the French and British attempts to impose retail-lending targets in exchange for bailouts (Jabko & Massoc, 2012;Macartney, 2014). A review of finance ministers' public statements around the time of the CRD-IV negotiation further reveals their fear that Basel III "risk[ed] threatening the financing of the economy" ("Christine Lagarde," 2010). Already in July 2009, Germany's Peer Steinbrück advocated a relax-ation of Basel II rules so that banks could increase lending to avoid a credit crunch ("Regierung und," 2009) and Austria's Josef Pröll reformed national taxes on banks to make retail lending a comparatively more attractive business ("Neuer Zwist," 2010). For Italy's Giulio Tremonti "Basel 3 [was] the direct way to produce a credit crunch" ("Banche: Tremonti," 2010) and Germany's Wolfgang Schaüble summed up the general mood stating: "We want a tightening of the rules [but] the financial sector must be in a position to continue to carry out its business" ("Highlights-Comments," 2010). As we could see in the previous section, this general will to find a compromise between strengthening financial stability and preserving lending however led Member States to adopt contrasted positions, which reflect the 'varieties of financial capitalism' (Howarth & Quaglia, 2013;Story & Walter, 1997) that persist in Europe.
Across the issues examined above, we could thus see the importance of the qualitative composition of national banking sectors-in terms of the legal form of banks that dominate them and whether they adopt the form of large, consolidated groups or decentralised networks-in shaping Member States' wish list. Indeed, the presence of (systemically) important cooperative, savings or public banks in countries such as Austria, Germany, France, but also Hungary or Poland is reflected in their insistence on exemptions and exceptions tailored to those particular types of bank, which have been shown to constitute important sources of finance for the local economies where they are established (Ayadi, Llewellyn, Schmidt, Arbak, & De Groen, 2010;Groeneveld, 2014). In June 2010, Austria's finance minister Josef Pröll explicitly linked his call for favourable treatment of cooperative banks to avoid a credit crunch ("Bankenabgabe kommt," 2010). Similarly, countries whose banking sectors are concentrated on a few large, internationalised national champions responsible for a major share of retail lending (e.g., France and Sweden) were keen to support these champions.
The particular instruments banks use to provide credit to corporates and households also appeared as key factors. The unanimous rejection of a sharp tightening of the treatment of mortgage loans reflected the importance of that particular form of credit in all sampled Member States, with those Member States where a majority of loans are mortgages (Estonia, Denmark, Sweden) making the most critical comments. Similarly, the strongest defence for covered bonds came from the countries where covered bonds markets are the most developed and stable (e.g., Denmark, Sweden, Germany). The particular defence of covered bonds may be interpreted in view of the fact that these instruments are specifically designed to support mortgage lendinghence help maintain lending levels-and were resilient through the financial crisis, so their inclusion would not jeopardize the pursuit of financial stability.
Finally, on issues related to the distribution of competences between home-and host-country supervisors and to harmonisation vs. national discretion, we can see a divide among Member States that reflects the varying importance of foreign bank operations across national banking sectors. The general reluctance of host countries (countries where foreign banks dominate the banking sector in terms of total assets or systemic importance; see Table 5) to give up national discretion reflect their exposure to the risk that foreign parents repatriate resources to the home country in times of crisis to benefit from nationally-oriented bailout schemes (Roubini & Setser, 2004), closing local subsidiaries or forcing them to deleverage rapidly, both resulting in a sharp decline of local credit supply. Host countries' insistence on national discretion can then also be interpreted as reflecting the general will to balance banks' contribution to the growth of the national economy with the systemic risk they represent.

Conclusions
In this article, I sought to examine the detailed positions of EU Member States on the post-crisis reform of capital requirements and to suggest factors that may explain these positions. In so doing, I have shown the importance of a series of structural features of national banking sectors (diversity of banking sector compositions, types of instruments used for retail lending, and varying degrees of foreign ownership) for Member States' assessment of policy proposals. I find that in most of the examined cases these factors explain the particular positions expressed by Member States. As such, my findings confirm the relevance of 'varieties of financial capitalism' (Howarth & Quaglia, 2013, 2016aStory & Walter, 1997) for our understanding of conflict between EU Member States on issues of financial regulation: the particular institutional setting on which each national banking sector relies to supply credit to the real economy mediates governments' double preference for stability and growth, resulting in sometimes conflicting positions.
Covering only a subset of EU Member States and CRDrelated issues, this analysis is necessarily limited and the explanation it provides for positions should be seen as complementary to other international political economy accounts. Further research is likely to uncover additional dimensions of Europe's 'varieties of financial capitalism' that shape Member State positions in important ways. Furthermore, since 2010, important events have occurred with major consequences for the setting of capital requirements. Banking Union, first, redistributed banking supervision and financial stability responsibilities, affecting perceived trade-offs between stability and growth (Epstein, 2017;Howarth & Quaglia, 2016b). Second, if after Brexit the UK adopts a deregulatory agenda on finance, the goal of promoting the competitiveness of their national champions may regain importance for the remaining Member States home to internationalised banks.