Before the global financial crisis (GFC) there had been some discussion of the way the regulation of banks in Europe might evolve. Some academic contributions had set out what a functional system might look like in the future. Mayes et al. (2011), for example, laid out a simple scheme based on US experience, where banks which wished to operate across borders should register as European companies, subject to a single regulatory scheme, under a new regulator – not the European Central Bank (ECB) – and should banks fail they would be resolved by a European equivalent of the Federal Deposit Insurance Corporation (FDIC), a European Deposit Insurance Corporation. Treatment of purely national banks would remain a national responsibility. However, such ideas were regarded as something for the distant future, with integration expected to take the form of steadily closer cooperation among national authorities and increasing harmonisation of regulation led by the Committee of European Banking Supervisors (CEBS) based in London. (Mayes et al. had assumed that CEBS would become the European level single supervisor.)
All this has now changed, and in the course of five years the European Union (EU) has implemented a major programme of legislation, which has been labelled as ‘banking union’. Although the current plans will not be fully operational until 2024, parts of it have come into force already. This speed of action has come at a price. The so-called ‘union’ is not complete – most importantly only the euro area is fully covered by the measures – and it does not have the neatness of the comprehensive scheme one might have expected. The reason is simple: such a comprehensive agreement could not have been negotiated among the member states if unanimous approval were required. Any changes which required amendments to the EU treaties would have entailed that unanimity.
The EU authorities and the member states are to be congratulated on their ingenuity in getting round these constraints. But the result has many flaws. The questions that are addressed in this book are: Will it work well despite these flaws? What changes can be made – within the bounds of political feasibility – that can improve it?
However, we need to start with a broader consideration of the pressures in the European and indeed global financial system that banking union seeks to address. Some of these are still with us, and it is extremely difficult to write this chapter with the uncertainty over the future over Greece hanging over us. If any of these spill over into a new crisis in European banks, then there will be no time for the new recovery and resolution arrangements to mature and for there to be any judgement about how well they could operate once the new authorities involved, particularly the Single Resolution Board, have been equipped for their task.
In this chapter, therefore, we set out a brief outline of banking union, followed by an analysis of the short-run pressures it faces. We then consider the longer-run difficulties that banking union will have to cope with before providing an assessment of how well it may be able to meet these challenges. These sections provide an introduction to the remaining chapters in the book, whose contribution we explain in a final section.
An outline of banking union
As explained in more detail in chapter 2 by Thomas Huertas, banking union is a label being applied to a group of three main actions by the EU in response to the GFC and the discussions on the way forward for banking regulation by the G20/Financial Stability Board and the Basel Committee for Banking Supervision.
The first action is increased harmonisation and the implementation of the Basel 3 rules for capital adequacy, through the capital adequacy directive and regulation, and through the increasing work of the European Banking Authority (EBA), which was formed out of CEBS. This harmonisation is notable for its move away from simply requiring minimum standards (which member states could exceed if they wished) to trying to set common standards that all must follow across the whole EU. Thus, although the main point of this action is to make banks more resilient to future shocks, it goes beyond this towards creating a more homogeneous system than is actually necessary for a single market to operate.
The second action is the creation of the Single Supervisory Mechanism (SSM), based on the ECB, to try to achieve a common high standard of supervision, especially for the largest banks that run across borders. The ECB is only supervising the largest banks directly, and the smaller banks and the non-banking activities will still be subject to the national authorities. This is not a genuine SSM, as only the euro area countries have to participate. Other countries may participate if they choose, but they cannot be full members as they are not represented on the ECB’s Governing Council. This division into SSM and non-SSM countries is unfortunate as the banks in the different categories of course interact. The reason for the division is not that there is something inherently different in banking terms by being a member of the euro area. It is simply that there was not the political will among the member states to create a new single supervisor that would cover the whole EU. The only way forward that did not involve a treaty change, which would have failed to achieve the necessary unanimity, was to exercise the option for the ECB to be assigned supervisory responsibility for banks under Article 127(6) (Treaty on the Functioning of the European Union). This conflates monetary policy and financial stability and potentially creates a conflict of interest for the ECB across its two roles as well as possible damage to its independence. This latter point is taken up below.
For the large majority of banks, which do not have substantial operations across borders, there is no particular need for an SSM, but for cross-border banks there are three obvious reasons. First, it could help lower the costs of banking services if banks only have to face a single regulatory regime and not a different one in every jurisdiction in which they operate. Second, it could help reassure host countries that the supervision of both branches and the operations of the parent bank are being conducted not just to a satisfactory standard but to a standard they can influence as they are part of the mechanism. Last, that reassurance can make it easier for countries to accept the pooling of resources in the resolution of large cross-border banks.
This takes us to the third action, the establishment of a regime for being able to resolve all banks, whatever their complexity, should they get into difficulty, at minimum cost to the real economy and without the need for a taxpayer-financed bail-out. This regime has been established in two parts:
- a Bank Recovery and Resolution Directive (BRRD), which requires all member states of the EU (and indeed, all members of the European Economic Area (EEA)) to have in place a regime which allows early action by the authorities should a bank get into difficulty – so that it can, if possible, be turned round before failure – and, should a failure become inevitable, to have all the tools necessary to effect a resolution;
- a Single Resolution Mechanism (SRM), which matches the SSM and is run by a new Single Resolution Board (SRB) in Brussels, independent of the ECB and responsible for implementing such resolutions especially of the large cross-border banks.
As part of the BRRD, the resolution authority in each country is required to have a resolution fund at its disposal, amounting to at least one per cent of banks’ covered deposits, which can be used to help finance resolutions. In the case of the SRM, these funds will be progressively mutualised as they are built up over 2016–2024.
Thus, the SRM, like the SSM, is incomplete and, since many of the most important cross-border banks lie in the UK, as does Europe’s most important financial market, this deficiency is substantial. Unlike the US, the EU has not created a single deposit insurance/guarantee organisation, which can act as the resolution authority. That is not such a significant concern, as, with depositor preference as part of the BRRD, it is not so likely that these funds will be drawn on in resolution.
The BRRD, however, is a major step towards enabling an immediate resolution of any bank in a manner which keeps the functions that are essential to the stability of the financial system operating without interruption. Previously, the regulatory environment in most countries meant that there could not be intervention until insolvency, and then the available insolvency procedures precluded uninterrupted operation unless the government either purchased the bank outright or recapitalised it enough to allow it to continue. In other words, they had to bail it out. With the new tools in the BRRD and the ability to ‘bail-in’ creditors – either write down their claim or convert their debt into equity such that the bank becomes adequately capitalised again – the hope is that bail-outs by the taxpayer will not be needed. The problem is that many of these procedures are untested, particularly with a large bank. Hence, we do not know whether they will work. Exploring this is the purpose of this book.
There are two parts of banking union that have not (yet) taken place, but might have been expected. One is a directive relating to the structure of financial institutions. This was proposed by the Commission in January 2014, but it has not as yet moved further, partly as a result of opposition from the industry, for which it would be costly, and partly because the member states have different views about what should be done. Belgium, France, Germany and the UK, for example, have implemented their own legislation already, and the idea of making banks change their structures twice in short order with the associated high compliance costs is not welcome. The problem is partly that the large banks are very complex, and it may prove impossible to make them readily resolvable, as required by the BRRD, unless their structures are simplified. However, beyond this, banking activities can be divided in two ways: first, according to risk; and second, according to the importance of the activity for maintaining financial stability. Part of the argument is over whether more risky activity should be separated so that it does not bring down the core activities should it make major losses. (It should not be forgotten, though, that the covariances among different sources of risk are at least as important as the risks evaluated individually.) The argument is also, however, over whether a banking group should be saved in its entirety by the resolution process or whether the bail-in should only relate to the parts which are vital to the smooth running of the financial system as a whole.
There are several possible answers to this, and the future of banking union will remain somewhat uncertain until this is agreed. The most important distinction is whether one should move to resolve the group as a whole (single point of entry, SPOE) or to resolve the vital parts separately in each jurisdiction, but coordinated by the lead authority (multiple point of entry, MPOE). SPOE is simpler and involves a single jurisdiction whereas MPOE involves coordination, which has not been achieved in the past. SPOE, however, may involve a larger bail-in and may be beyond the resources of small countries with large banks.
The second omission from banking union is the unification of deposit insurance/deposit guarantees. While the EU has a common minimum guarantee of €100,000 per depositor per bank, it has not gone as far as implementing a single EU-wide scheme, mainly because they are too different to harmonise easily. Given the BRRD has required the creation of resolution authorities in each member state and has instituted depositor preference, the issue of the use of insurance funds in a crisis is no longer likely to be very important. So, although this means that the EU does not look like the US where the deposit insurer, the FDIC, covers all banks and is the resolution authority, it is not in itself a problem.
In chapter 6 Rosa Lastra argues that there is a third omission, that of a lender of last resort (LOLR) at the European level. However, the LOLR function exists in each of the member states and in the euro area. The LOLR function effectively involves two actions. The first is the expansion of market liquidity when an institution gets into difficulty and the pessimism this generates threatens to spread round the rest of the financial system, turning a problem into a disaster. The second is simply to lend to the troubled institution itself, assuming it has appropriate collateral, when the market has decided that it is too great a risk to justify lending to that institution.1 In non-euro countries the central bank performs both functions, but within the euro area the ECB performs the first and the national central banks (NCBs) perform the second, normally with the ECB’s prior agreement, but in emergency they can lend first and clear it with the ECB afterwards. The ECB insists not just on solvency of the institution concerned but that any programme of public support can be credibly repaid within a few years. This chapter by Rosa Lastra identifies legal problems, which can exacerbate or even cause economic problems. It also highlights the distinction, clear in principle but inevitably blurred in practice, between common law (primarily the law of the English-speaking world) and civil (or Roman) law. The essence of the difference, baldly stated, is that under the former one can do anything that is not forbidden, while under the latter one can do anything that is allowed. This distinction appears in her discussion of recent US changes and of a restriction on the ECB. She notes that the Federal Reserve’s freedom for action in a crisis has been restricted by the Dodd-Frank Act (2010); but the ECB ‘is not competent’ – that is to say, is not allowed – to provide liquidity assistance to individual banks. (Professor Lastra notes that it has imposed this restriction on itself by its interpretation of the ECB Statute.)
The ECB indubitably has clear authority to provide liquidity to the market as a whole; indeed, doing so involves exactly the same action as does an easing of monetary policy. The only difference is the reason and possibly also the scale. But responsibility for individual banks lies with NCBs. This is identified as a crucial gap in banking union, for, she argues:
While prudential supervision was at the national level, it was perhaps logical to assume that the adequate expertise and information to assess the problems of banks within their jurisdictions . . . But . . . the ECB should be . . . lender of last resort for all those institutions it now supervises.
The practical importance of this was illustrated fairly recently: the problems of dealing with Northern Rock in the UK were exacerbated by supervision not being undertaken by the Bank of England.
A further and more general problem is that there is no European ‘fiscal backstop’; that is to say, there is no European body that can provide capital should it be needed to maintain banking stability. There is a body which can recapitalise the ECB should it experience substantial losses. But that body comprises national governments, which by having fragile fiscal positions may have caused problems in their banking systems and retain responsibility for their NCBs should they need recapitalising. This last observation leads to a further question.
Is this union?
Before going on to show how the chapters in this volume fit into and develop the above framework, it is necessary to consider what can be meant by the term ‘banking union’.
The plans that have been described as banking union by their designers embody an implicit definition. A banking system, spread over areas and countries, is a banking union if it has a common regulatory framework and a common regulator. That implicit definition is troublingly incomplete, and it has curious implications. It says, for example, that Britain did not have a banking union until 1979 – until then British banks were regulated only by the same set of company laws that regulated other firms. There was no specific banking regulation. Also, the definition is not only odd in some of its implications; it omits, or perhaps presumes as following automatically from having a single regulator, an important aspect of what should be integral to such a union.
Certainly one reason to promote a banking union (and the same reason applies to the promotion of a capital markets union) is to increase the efficiency of resource allocation. How can this be done in the present context?
Suppose that the interest rate charged for an identical loan to identical borrowers in different parts of the euro area differed. Would that comprise a capital market imperfection, which could be eliminated by a capital market union? Lance Davis (1963) would say so: he measured the approach to a national capital market in the US by the decline in differences in interest rates among regions. But as Stigler (1967) points out: ‘No-one would dream of using this criterion for wheat or automobiles’. Stigler’s point generalises as follows. Price differences between the same good at different points, which are less than the transport costs between them, make the movement of the goods uneconomic. The goods will not be transported, and the remaining price difference is evidence of efficiency not of inefficiency. Similarly, a misallocation of capital is created, not removed, if interest rates move together without the genuine removal of what was keeping them apart. Moving them together by regulation, for example, would be damaging.2
Now what does this imply for the meaning of banking union? Take the example of the UK. That country has had for many years now a genuine banking union. That has come about through a system of nationwide banks. Not all banks are nationwide and many fewer banks were nationwide when interest rate differences first virtually vanished across the UK, but there is a banking union created by the ability individuals have to borrow and lend where they wish and for banks to lend and borrow in any part of the UK they wish, without regulatory impediment or barriers to entry in any part of the country (at any rate for an institution which is already in the banking system). Both supply and demand can move freely across the nation.
A genuine, efficiency improving, banking union would be one which led to such a situation in the EU. Can what has been described as a banking union be expected to do that? It is hard to see how. Centralising regulation for large cross-national banks is indeed likely, as...
