8 December 2004
Speech by Callum McCarthy

  1. In a recent paper by Tomasso Padoa-Schioppa, he argued for two developments – the establishment of a single rulebook for financial services across the EU and for effective supervisory convergence. Today I would like to consider the latter of these broad subjects. I will leave the question of a single rulebook for another occasion. Let me simply say that before embarking on the task of establishing a single rulebook it would be important to have a realistic understanding of the scale of the task, and be clear that the benefits would in fact outweigh the costs – an assumption for which there is, to my knowledge, no evidence whatsoever.

  2. It is with some relief therefore that I turn to what I regard as the less contentious problem of how we should effectively regulate international financial services firms. None of us here needs reminding of the reality of international business: the continuing and increasing importance of financial services firms operating cross border. This is true in any Member State; it is particularly true in the UK, where as a deliberate and I would argue highly successful act of public policy we have an approach more open to foreign investment in financial services than any other large EU country, and probably than any G10 country. We now have a major retail bank in the UK owned by a Spanish parent – a European cross border first for a major EU country. But the growing importance of cross border activities is not just a UK phenomenon: the US banks are very large players in Europe; in London, Deutsche Bank is a very large influence in both equity and debt markets; German and Austrian banks in particular account for a very large proportion of the financial systems of several countries that have recently acceded to the EU; EU banks – ABN Amro, Paribas, RBS are obvious examples – are expanding in the US.

  3. I need not emphasise how this spread of international activity is putting pressure on the management of the banks and their regulators. Managements have to identify and control the risks and conflicts which occur in acute and difficult form when running a 24 hour risk book in different cultural, regulatory, legal and political regimes – a mammoth task which strains the competence of even the most able and responsible bank, and which will overwhelm any which has global ambitions which are not matched by the necessary competence and standards. Regulators also have a responsibility to ensure that management and control structures are equal to this task and that local requirements are being met – a considerable challenge when the business is run on global lines. As business has become more global in its scope and the regulatory task has become more complex, a natural tension has developed between the need for national regulators to have an ever closer understanding of major firms and the desire on the part of the firms themselves for more streamlined regulation that mirrors the way they organise themselves. Regulators are coming under increasing pressure to do what they can do to remove unnecessary, duplicative or even inconsistent regulatory requirements arising from international firms having to deal with multiple regulators round the world, or across the EU.

  4. It is not surprising that the response to this has been demands from the firms for regulatory convergence. This is sometimes equated to the development of the lead supervisor concept and for a redefinition of home and host supervisory powers. All of these are variations on a theme. Let me say right away that I am a strong supporter of the principle of reducing the supervisory burden on firms. We need always to remember that regulation poses costs, that it needs to be rigorously justified and that we need to be constantly alert to any redundancy or inefficiency in what we do. Having said this, however, we need to think very carefully about what is meant by terms such as "lead supervision" to make sure that, in trying to meet the legitimate demands of firms, we do not do more harm than good. Those of us involved in the supervision of internationally active firms have long recognised and attached specific duties to the 'home' supervisor. This traditionally refers to the supervisor that has overall or consolidated responsibility for a global firm or group and is typically that of the country in which the head office is domiciled or which is the principal place of business. In recent discussions, however, a distinctly hard edged concept of the 'lead supervisor' has emerged. In its simplest form, the idea is that any financial institution should be supervised, for prudential purposes at least, pretty much exclusively by the home country (or lead) supervisor – with the 'host' regulators in other countries losing any scope for challenging or forming an opinion different to the analysis and decisions of the home regulator. On this basis, HSBC, which tends to operate globally in the form of subsidiaries would be subject exclusively to prudential supervision by the FSA as its home regulator in the UK. Host supervisors throughout Europe (and in some extreme variants of the model, throughout the world) would cede to the FSA any responsibility for assessing the quality of management, controls, governance or financial soundness in subsidiaries in their jurisdictions. This would put HSBC's operations on a par with those of Deutsche Bank in the UK, where it operates as a passported branch and which is supervised, in respect of prudential issues solely by BaFin in Germany. (The FSA in London, where Deutsche's trading decisions can have such an effect on the British markets for both debt and equity products, has no part to play in assessing the capital adequacy of the Deutsche group, whose British activities are carried on by a branch of the parent). Most of the discussion of the lead supervisor concept has taken place within Europe. But a logical extension of this would clearly be a global variant. US broker dealers and other institutions would be regulated solely by one from the range of US regulatory authorities which now share responsibilities – the Federal Reserve Bank of New York (or one of the other regional Federal Reserve Banks), the Office of the Comptroller of the Currency, the Securities and Exchange Commission, or the Office of Thrift Supervision; and, if I go back a decade or more, BCCI was in fact supervised by the Luxembourg authorities with results which are still subject to continuing legal action.

  5. I have spelt out some of the range of examples because it is useful to be reminded of the range of complexities which arise in practice associated with the beguilingly simple concept of home regulator. I will leave aside the question of change of country of incorporation – though it is not a trivial question: HSBC has publicly stated that, as a matter of policy, it considers the question of country of incorporation on a regular basis. But each of the other examples gives rise to real questions which need examination. My purpose is to raise some of the questions which are too often ignored in the discussion of home supervisor, and to suggest some answers to them. Let me emphasise again that it is not my purpose to reject here any of the proposals currently around for greater integration in supervisory practice, but rather to make a meaningful contribution to the debate by moving it from the realm of theoretical concept and slogan into the world of a practical reality. In Europe, the Lamfalussy process, with its four level approach, provides the necessary framework for doing this.

  6. Let me start with a question which all regulators should be concerned with, and which all sensible regulators worry about. It is the question of the legitimacy of their actions: how do they justify the use of the often substantial powers which are granted to them? The FSA's answer to this question has many facets. We start with the recognition that regulation imposes costs. These need to be rigorously scrutinised and set against the expected benefits of regulation in protecting participants in financial markets from various forms of market failure. We are transparent in setting out our proposals, both generally through the annual cycle of plan and report on the actual achievements against the plan, and in particular through repeated consultation on individual policy proposals and our reaction to the comments we receive on those proposals; the obligation to demonstrate proportionality in our actions; and – critically – our accountability to Parliament for the discharge of the powers and duties which Parliament has given us, made manifest in the repeated meetings we have with Parliamentary Committees. This Parliamentary accountability matters as a means of maintaining the legitimacy of our operations. Yet how would we achieve this on a simplistic interpretation of the supremacy of the home regulator? How, for example, would Parliament respond to an answer about a major entity in the UK – perhaps one which had run into difficulties, perhaps difficulties which had caused severe knock on effects – when the FSA had no information about the entity, and had made no decisions about it? Would it be good enough to say that all enquiries should be addressed to the regulator in some other country, perhaps in the EU, perhaps elsewhere? This is not a theoretical question: I have already explained that all questions on the capital adequacy of Deutsche Bank are questions for BaFin, not for the FSA. In other countries the position is even more stark: in Hungary, for example, around 75 percent of the banks' assets are accounted for by banks owned outside Hungary. Around 50 per cent of assets under management in that country are also managed by subsidiaries of firms from outside Hungary. Yet there exists a Hungarian financial regulator, the Hungarian Financial Supervisory Authority, which is charged under Hungarian law, and is certainly expected by the man or woman in the street in Budapest, to deliver a reasonable level of financial regulation. A similar position exists in other accession countries: Nordea, for example, is a bank of arguably systemic importance in Estonia, and it would be hard to believe that the Estonian regulator could simply address all questions raised with him to the Swedish authorities. There are other examples which can be adduced. The general point however is that there are questions of legitimacy for the regulatory organisation which have to be answered, and for which a simplistic approach to the powers of a home supervisor provide no answer. These questions are important, and affect the democratic accountability of the institutions. Any approach to home regulator powers – or to supervisory convergence – has to take them into account.

  7. There are other questions which also need to be addressed if we are to progress the concept. We should recognise that there are substantial differences between the legal powers granted to different regulatory organisations in different countries. Even within the EU there is no consensus as to whether regulation should be carried out on an integrated basis, so that banking, insurance and securities trading are all subject to a single regulatory body as occurs in both Germany and the UK, or whether there should be some different combination, such as a functional separation of prudential and conduct of business as in the Netherlands, or a sectoral approach with different and separate regulatory bodies, as occurs in France and Spain. Nor is there any agreement as to whether there should be unified or separate regulation of prudential questions and of conduct of business questions. Accordingly there is no common basis of legal powers even within the EU. And beyond the EU the position is even less clear, as my examples of the wide variety of US regulators who could make a claim for the title of lead regulator shows. This absence of common powers matters: if, for example, there are limitations on the ability of a home regulator to obtain information there are consequently severe problems in relying on that regulator as the home regulator for normal supervisory purposes. It also matters in more extreme circumstances: the powers of the FSA and of the SEC which were exercised in the context of the recent Shell decision are powers not available to the Dutch regulator. It would be unattractive if the sanctions which were available to be used were restricted to those in the home regulator's country. I strongly support the recommendation made by CESR in its recent discussion document that there should be an effort to consolidate and unify the regulatory powers available to regulators in different member states. Of course, realistically, this will take some time to achieve, even if the will exists to do so. Although that recommendation was made in the context of securities regulation, I believe it has a wider applicability.

  8. There are also questions of political will. It is clear that the degree of independence of financial regulators is not uniform across the world, nor is it uniform even within the EU. To take two recent incidents, there has been the sudden change in the executive of the Dubai Financial Authority, which raised questions as to the independence of that authority; within the EU the change in the Hungarian securities regulator raised issues severe enough for the principle of independence to be raised with the Hungarian government by a number of institutions, including CESR. And it is clear that not all countries have the same approach to financial regulation: that is the problem represented in extreme form by a number of well known financial centres which have made a specialism of being involved in the less reputable aspects of finance. This problem, which is typically – but rather inaccurately – associated with offshore centres and often involves failures of policy in respect of transparency and compliance with basic standards, is being tackled. But, aside from these well known egregious cases, questions about implementation of international rules and standards, whether from global organisations such as IOSCO or from EU directives and regulations, remain and need to be tackled. We have to recognise that not all countries believe it is in their own interest to interpret a directive to a level consonant with the spirit intended, even if they comply with the letter.

  9. There is also a question of competence. Not all countries are able to devote the resources, or have the experience, needed to discharge the responsibilities of lead regulator in the hard edged form being proposed in some quarters. I think that it is generally recognised – at least in retrospect - that it was unrealistic to expect the Luxembourg authorities to discharge the very difficult task of supervising the worldwide activities of BCCI, and that in practice reliance upon them did not succeed. We should recognise that this is a real problem, which should not be glossed over.

  10. Finally, we should recognise that, even in a world of good supervision, there will be failures, and that there is a need for effective and even handed treatment of those affected by failure. There are two aspects of this. The first – connected to the question of democratic accountability which I discussed earlier – is the question of how to deal with a financial institution which is of relatively small importance in the home country but of major importance in the host country or countries. It is easy to envisage circumstances where a group wide problem would cause particular problems in the hosts countries, but the decision as to what should be done, and the costs of any action, would fall to the home country. Consider, for example, what would occur if there were a problem affecting Nordea, a bank incorporated in and regulated from Sweden, but arguably of greater importance to financial stability in both Latvia and Estonia than in Sweden. Were there to be a group-wide problem, the Swedish regulator, the Swedish central bank and the Swedish finance minister would be called upon to make the decision whether to mount a rescue and, were they to decide to do so in a way which made a call on public funds, it would be the Swedish taxpayer who would meet the costs of so doing. It is easy to see that the misalignment between those who would particularly benefit from the rescue and those who were required to pay for the rescue could be a source of acute difficulty. I should make clear that I use the example of Nordea to illustrate a general problem in a tangible form: I do not imply that I have any doubts about the particular institution. It is the general problem which concerns me.

  11. There is a further problem, which is that compensation schemes are not uniform. Deposit insurance in the US, for example, applies to deposits payable in the US. It does not extend to deposits in branches of US banks that are payable outside the US. This is a clear – but probably not widely understood – source of risk to some UK depositors. Within the EU the position on compensation, even after the compensation directives, is complex: there is, for example, at present no EU requirement for a compensation scheme for insurers; the minimum levels laid down in the compensation directives are lower than the practices which have been adopted in different member states; the coverage differs. The effects of these differences has been that national regulators have often thought it necessary either to impose their own individual requirements, or have approached the task of supervising firms operating within their country with a concern to ensure that, in the event of failure, those affected within their country are not treated less well than those elsewhere. And sometimes, it should be recognised, this proper concern has in practice been translated into a determination to safeguard the interests of a particular country, irrespective of the effects in other countries. To put it simply, US and Australian bank regulators have a legal duty to give preference to US and Australian depositors which inevitably affects their performance in a lead supervisor role. While there is a danger of this happening, it is clear that there will be a natural reticence on the part of national regulators to rely on the actions of the home regulator. This is not a question which can be solved by regulators acting with goodwill. Its origins lie in differences in different countries in the legal treatment of domestic and overseas creditors in the event of a liquidation. British bankruptcy law is very even handed as between UK and other creditors, but much bankruptcy law elsewhere is not, and a host regulator cannot but be mindful of this fact.

  12. My purpose has been to identify some of the real and practical difficulties which have to be overcome if we are to make progress in achieving the highly desirable end of regulatory convergence. I confess to some irritation with those who approach the subject with the self-confidence which comes all too easily from simply ignoring practical realities; and I despair of those who treat all attempts to discuss these issues as a form of job creation defensiveness on the part of central banks whose raison d'etre has been threatened by membership of the eurozone. Indeed, one of the reasons why I wish to discuss these questions is both a belief that they are important and the self-evident fact that this accusation is not one which can be levied at the FSA.

  13. Let me now turn to the answers to the questions which I have posed – for I think it important that we do find answers. My central answer is that we start from a recognition that not all financial institutions pose the same issues, and that we need to devise a taxonomy of financial institutions which enables us to make appropriate distinctions between financial institutions. Let me illustrate what I mean by this. There are many financial institutions whose activities in a host country are of not of major importance; they are not systemic or even, to use FSA jargon, 'high impact'. We should welcome and support the continuing enhancement of home regulator responsibility for those institutions. What I believe we should be aiming for here is not that host supervisors cede powers to home states but that they reach sensible agreements on limiting the circumstances in which it is appropriate to use those host powers in day to day supervision. But there are some financial institutions which are of systemic importance in a host country, or which, even if not of systemic importance, have a very major impact in the host country. For these institutions we need to develop arrangements where they do not currently exist – for example in respect of EU branches - which recognise the need for the host regulator to have influence and even rights in the decisions affecting those institutions.

  14. Let me give examples of what this would mean in practice for bank subsidiaries. The present draft of the Capital Requirements Directive in Article 129 proposes that, in the event of the range of home and host supervisors being unable to agree, the home supervisor should be given supremacy in the decisions as to Basel I model recognition for a group, including the subsidiaries within the group operating in other member states. The purpose of this is to simplify the model recognition process which a bank group operating across the EU has to undergo – in itself a desirable objective. But I hope that, in practice, we can agree to distinguish between banks whose operations in a host country are of small import, for whom model recognition decisions could safely be left to the home regulator, and those whose activities in the host country were of major or systemic import. In the latter case it is important that home supervisors, whatever their formal powers, should take full account of any legitimate concerns of host regulators in making their decisions. This seems to me to be one of a number of areas – including day to day supervision of high impact firms - in which home and host supervisors should collaborate more effectively than in the past. Note that this would not necessarily mean that the systemically important financial institution would have to undergo multiple examinations or provide information in multiple ways. It should not be beyond the scope and the ingenuity of regulators in different member states to devise ways of working together to avoid this. I say this with considerable confidence because the FSA has been in the forefront of developing such models of cooperation, both internationally and in the regulation of UK firms. But it would mean that the host regulator would retain the rights that it currently has in respect of subsidiaries, which could be used in the event that there were difficulties with the home regulator. And, if this approach were to be adopted, it would help mitigate the problems of democratic accountability, of competence and of policy divergence which I earlier identified.

  15. A similar approach could be adopted for the treatment of clearing and settlement services provided through branches of EU financial services companies. This arrangement is not at present the norm. Rather clearing and settlement services are at present provided through subsidiaries of EU firms, and the regulation of the subsidiary in any host country is done by the host regulator, in close collaboration with the home regulator (and other hosts). It is quite probable however that in future such services will be provided through branches. The terms of the new directive on clearing and settlement have yet to be worked out. Were that directive to follow narrowly the approach set out in existing EU directives, however, the host regulator would have no day to day prudential oversight or powers in respect of the activities of the branch. It is, however, manifest that the branch of a EU wide clearing and settlement firm would be highly likely to be of systemic importance to the host country, and because of this desirable that the host regulator should have some locus, rather than simply accepting the decisions of the home regulator. We would therefore wish to press for the home regulator to have some more rights, de facto, than at present in respect of branches supplying systemically important clearing and settlement services. The host regulator might, for example, have the right to undertake some level of oversight of the branch, and to seek information from the firm which was relevant to its operations, and hence the risks it poses, in the host country. It would also give the host regulator the right, if it could demonstrate that the activities of the branch were creating risks in the host market, first to raise issues with the home regulator and – in the exceptional circumstances of there being no meeting of minds between home and host regulators – then to take actions in respect of the branch even if these actions were not agreed with the home regulator. I would emphasise one very important point here. In pressing for these various forms of enhanced cooperation, we are not looking for any change in the legal framework within which we operate in Europe. Article 56 of the MiFiD Directive makes provision – without spelling out the implications – for "proportionate cooperation arrangements" in respect of regulated markets. We believe that the kind of arrangement I have outlined would be an appropriate interpretation of that provision in respect of day to day supervision, though only – I would emphasise – for systemically important branches. Article 22 of the Banking Coordination Directive also provides some precedent for providing host supervisors with some scope to assume powers over branches in certain circumstances. Again, I should make clear that I do not expect these in extremis rights to be used often; in an ideal world they would not be used at all. But I believe that it is necessary to make provision for their use if necessary.

  16. Article 22 of the Banking Coordination Directive is of particular interest in this context as it makes provision in certain specified and extreme circumstances – when the activities of a bank branch are violating legal rules in the host state and the home regulator has failed to respond to the host regulator's representations – for the host regulator to assume formal powers over the branch. Although this power is explicitly granted to the host regulator in the Directive, much remains to be done to translate the legal right into practice. I will comment in a moment on some practical steps which are in hand to advance this. But it is worth considering whether the legal power which has so explicitly been recognised as needed for banks should not also be recognised as being equally required for other financial institutions. I see advantage in developing the principle to cover other financial institutions such as insurance companies and securities houses, as well as clearing and settlement systems.

  17. If we develop the principle which is helpfully explicit in Article 22 of the Banking Coordination Directive we should establish a consistency of approach which would be helpful. The alternative, of different principles governing activities which are often carried on by the same firm, would be awkward and lead to all sorts of regulatory arbitrage of an obviously unattractive nature. Let me emphasise again that we are not looking for significant changes in the legal frameworks for home/host collaboration within Europe. I do believe, however, that within these formal frameworks we need to construct a series of protocols to govern the behaviour of both home and host regulators. In fact, I would go further and say that the need for such a set of protocols goes some way beyond Europe and should govern our day to day approach to global regulatory relationships. These protocols should cover information exchange: what information can the home regulator require the host regulator to provide – and what information should the home regulator be expected to provide to the host regulator? How should differences in defining the information required be dealt with? How are the information rights and duties going to be graded according to the importance of the firm's and exchange's activities in a particular country – for the smaller that importance, the less relevant to the home regulator is information from the particular host country, just as the need for the host regulator to have information about a low impact firm is reduced? What is the best means of exchanging information?

  18. Regulators have already made considerable progress in devising practical answers to these and other questions, often on a basis which goes beyond the boundaries of the EU. For the two large Swiss banks, for example, twice a year there are meetings of the Swiss Banking Commission (who are the lead regulator), the Federal Reserve Bank of New York and the FSA - the three most relevant regulators in the three capital market centres of most importance to Credit Suisse and to UBS – to compare views, exchange information and coordinate any necessary actions. For HSBC we at the FSA have recently begun to chair meetings of what we call the college of HSBC regulators, dealing with HSBC's activities in the most important markets in which they operate. By bringing together regulators from the US, Canada, Switzerland, Hong Kong, France and the UK we are able to cover some 80 per cent of HSBC's assets. We need to develop this sort of approach to provide further guidance and help to those countries where the firm's activities are very significant to the country, even if not particularly important to the whole of the HSBC group. In another context, arrangements have been developed between the Belgian, Dutch, French, Portuguese and British securities regulators to supervise the activities of Euronext.

  19. Alongside these protocols defining information exchange, there is clearly scope for other work to develop supervisory practices. There are many aspects of this: larger and longer established regulatory organisations can share their experience – both successes and mistakes – with smaller or newer counterparts (something which the FSA has done annually in its international regulators' seminar, the fifth of which is taking place at the moment with over 100 participants). There is scope for regulators working with the Commission to assess the success or otherwise of how EU directives have been implemented (something which CESR has started to do through its work on the Market Abuse and Prospectus Directives); through the Lamfalussy process in the EU there is a greater prospect of identifying instances of inadequate or inappropriate implementation, first by applying peer pressure and – if this proves unsuccessful – acting as an early warning system so that the Commission can initiate infraction proceedings. The opportunities for the three Lamfalussy Committees to develop along these lines are clear. I think it important, however, that there should be a degree of realism in the expectations placed upon them. None of the committees is more than three years old; CEBS, the banking committee, and CEIOPS, the insurance committee, have just been established. CESR, the oldest committee, has been preoccupied with the enormous task of translating the stream of level one directives into advice at level two for the Commission – a task which has understandably consumed almost all its resources. One benefit of the fact that the legislative phase involved in the Financial Services Action Plan is coming to an end is that – provided it is not succeeded by a further wave of legislation – CESR will be able to develop a series of initiatives dealing with the important questions of implementation. Recent consultation papers from CESR have outlined the territory and I, and many of my colleagues on CESR, look forward to the opportunity to take these initiatives forward.

  20. Let me sum up. In this paper, I have attempted to identify the practical problems which have to be overcome if, as is highly desirable, we are to turn the concept of lead regulator or home regulator into a practical reality. We are certainly not looking to overturn the legal basis for regulatory cooperation in Europe. I have however suggested an approach based on recognising that, if a bank, exchange or other financial institution is of systemic or major importance in a host country, the host regulator should have certain rights alongside those of the home regulator. It is hard to imagine that, whatever the legal framework, hosts should not be able at least to have some direct knowledge of the risks posed by systemically important entities in their markets, to have a voice in making representations to home regulators and to expect significant risks to be dealt with – provided of course that a rigorous case can be made. And I have indicated some legal and supervisory precedents on which we can build. As you would expect from my position at the FSA, I have approached this as an issue which is important for not only banking, but also for insurance and securities trading. It is a pleasure to be able to build on a suggestion from Tomasso Padoa-Schioppa.

More Speeches: