EMU and banking supervision

Lecture by Tommaso Padoa-Schioppa Member of the Executive Board of the European Central Bank at the London School of Economics, Financial Markets Group on 24 February 1999

1. I am speaking here, at the London School of Economics, only a few weeks after one of the most remarkable events in the history of monetary systems: the establishment of a single currency and a single central banking competence for a group of countries which retain their sovereignty in many of the key fields where the State exerts its power. To mint or print the currency, to manage it and to provide the ultimate foundation of the public's confidence in it has been, from the earliest times, a key prerogative of the sovereign. "Sovereign" is indeed the name that was given in the past to one currency. And a British Prime Minister not so long ago explained her opposition to the idea of the single currency with the desire to preserve the image of the Queen on the banknotes.

2. For centuries money has had two anchors: a commodity, usually gold; and the sovereign, i.e. the political power. Less than 30 years after the last bond to gold was severed (August 1971), the second anchor has also now been abandoned. Although I personally think that political union in Europe is desirable, I am aware that the present situation, in which the area of the single currency is not a politically united one, is likely to persist for a number of years. This means that we have given rise to an entirely new type of monetary order. For the people, the success of this move will ultimately depend on the ability of governments and political forces to build a political union. For the central banker and for the users of the new currency, the success will be measured by the quality of the currency itself, and such quality will be measured in the first place in terms of price stability. This is not only a requirement explicitly set by the Treaty of Maastricht, it is also, in the opinion of most, the "new anchor" that purely fiduciary currencies need after the gold anchor is abandoned.

3. My remarks, however, will focus on another, less fundamental but still important novelty of the monetary constitution that has just come into existence. It is the novelty of the abandonment of the coincidence between the area of jurisdiction of monetary policy and the area of jurisdiction of banking supervision. The former embraces the 11 countries that have adopted the euro, while the latter remains national. Just as we have no precedent of any comparable size of money disconnected from states, we have no precedent for a lack of coincidence between the two public functions of managing the currency and controlling the banks.

In the run-up to the euro this feature of the system was explored, and some expressed doubts about its effectiveness. I will tonight examine the problems of banking supervision in the euro area. The plan of my remarks is the following. I will first review the existing institutional framework for the prudential control of banks in EMU. I will then examine the likely scenario for the European banking industry in the coming years. Against this institutional and industry background, I shall then discuss the functioning of, and the challenges for, banking supervision and central banking in the euro area, both in normal circumstances and when a crisis occurs.

4. The origin and developments of modern central banks are closely linked to key changes undergone by monetary systems over the past two centuries. Such changes could, very sketchily, be summarised as follows. First, paper currency established itself as a more convenient means of payment than commodity currencies. Second, commercial bank money (bank deposits) spread as a convenient substitute for banknotes and coins. Third, the quantity of money was disconnected from the quantity of gold. Thus, a double revolution in the technology of the payment system, the advent of banknotes and that of cheques or giros, has shaped the functions that most central banks performed over this century: monetary policy and prudential supervision. Man-made money made monetary policy possible. The fact that a large, now a predominant, component of the money stock was in the form of commercial bank money made banking supervision necessary.

Ensuring confidence in the paper currency and, later, in the stability of the relationship, one could say the exchange rate, between central bank and commercial bank money, were twin public functions, and, in general, they were entrusted to the same institution. Just as money has three well-known economic functions - means of payment, unit of account and store of value - so there are three public functions related to each of them. Operating and supervising the payment system refers to money as a means of payment; ensuring price stability relates to money as a unit of account and a store of value; and pursuing the stability of banks relates to money as a means of payment and a store of value. In each of the three functions commercial banks have played, and still largely play, a crucial role.

In an increasing number of countries the original triadic task entrusted to the central bank has now been abandoned in favour of a "separation approach", according to which banking supervision has been assigned to a separate institution. Following the recent adoption by the United Kingdom and Luxembourg of the separation approach, only two of the 12 countries represented in the Basle Committee on Banking Supervision (Italy and the Netherlands) have the central bank as the only authority responsible for banking supervision. In all systems, however, whether or not it has the task of supervising the banks, the central bank is deeply involved with the banking system precisely because the banks are primary creators of money, providers of payment services, managers of the stock of savings and counterparties of central bank operations. No central bank can ignore the need to have a concrete and direct knowledge of "its" banking system, i.e. the banking system that operates in the area of its monetary jurisdiction.

Personally, I have an intellectual attachment to, as well as a professional inclination for, the central bank approach to banking supervision, due partly to the fact that I spent most of my professional life in a central bank which is also to this day the banking supervisor. Yet I can see, I think, the arguments that have led a growing number of industrialised countries to prefer the separation approach. Such arguments basically point to the potential conflict between controlling money creation for the purpose of price stability and for the purpose of bank stability. On the whole, I do not think that one model is right and the other wrong. Both can function, and do function, effectively; if inappropriately managed, both may fail to satisfy the public interest for which banks are supervised.

5. Against this background, let me now describe the institutional framework currently adopted by the Treaty. As my description will refer to the area in which both the single market and the single currency are established, it will not specially focus on the problems of the so-called "pre-in" countries, including the United Kingdom.

The current institutional framework of EMU (i.e. the single market plus the single currency) is a construct composed of two building blocks: national competence and co-operation. Let me first briefly review the main aspects of these two building blocks and then see how the Eurosystem relates to them.

First, national competence. In a market based on the minimum harmonisation and the mutual recognition of national regulatory standards and practices, the principle of "home country control" applies. According to this principle every bank has the right to do business in the whole area using a single licence, under the supervision, and following the rules, of the authority that has issued the licence. The full supervisory responsibility thus belongs to the "home country". This allows, inter alia, the certain identification of the supervisor responsible for each institution acting as a counterparty to the monetary policy operations of the Eurosystem. The only exception to this principle - the "host country" competence for the supervision of liquidity of foreign branches - is no longer justified now that the euro is in place; hence it should soon be removed.

Second, co-operation. In a highly regulated industry such as banking, a single market that retains a plurality of "local" (national) supervisors requires close co-operation among supervisors to safeguard the public good: namely, openness, competition, safety and soundness of the banking industry. EU directives (the 1st and 2nd Banking Directives and the so-called BCCI Directive) lay the foundations for such co-operation, but they do not contain specific provisions or institutional arrangements to this end. They limit themselves to stating the principle of co-operation among national authorities and to removing obstacles to the exchange of information among them.

6. How does the Eurosystem relate to this construction? Essentially in two ways. First, the Treaty assigns to the Eurosystem the task to "contribute to the smooth conduct of policies pursued by competent authorities relating to the prudential supervision of credit institutions and the stability of the financial system" (Article 105 (5)). Given the separation between monetary and supervisory jurisdictions, this provision is clearly intended to ensure a smooth interplay between the two. Second, the Treaty gives the Eurosystem a twofold (consultative and advisory) role in the rule-making process. According to Article 105 (4), the ECB must be consulted on any draft Community and national legislation in the fields of banking supervision and financial stability; and, according to Article 25 (1) of its Statute, the ECB can provide, on its own initiative, advice on the scope and implementation of the Community legislation in these fields. It should be borne in mind that central banks are normally involved in the process of drawing up legislation relating to, for example, regulatory standards, safety net arrangements and supervision since this legislation contributes crucially to the attainment of financial stability.

7. Two observations should be made about the institutional framework just described. First, such an arrangement establishes a double separation between central banking and banking supervision: not only a geographical, but also a functional one. This is the case because for the euro area as a whole banking supervision is now entrusted to institutions that have no independent monetary policy functions. The separation approach that was chosen for EMU has effectively been applied not only to the euro area as a whole, but to its components as well. Indeed, even in countries where the competent authority for banking supervision is the central bank, by definition this authority is, functionally speaking, no longer a central bank, as it lacks the key central banking task of autonomously controlling money creation.

The second observation is that the Treaty itself establishes (in Article 105 (6)) a simplified procedure that makes it possible, without amending the Treaty, to entrust specific supervisory tasks to the ECB. If such a provision were to be activated, both the geographical and the functional separation would be abandoned at once. The fact that the Maastricht Treaty allows the present institutional framework to be reconsidered without recourse to the very heavy amendment procedure (remember that such procedure requires an intergovernmental conference, ratification by national parliaments, sometimes even a national referendum) is a highly significant indication that the drafters of the Treaty clearly understood the anomaly of the double separation and saw the potential difficulties arising from it. The simplified procedure they established could be interpreted as a "last resort clause", which might become necessary if the interaction between the Eurosystem and national supervisory authorities turned out not to work effectively.

8. When evaluating the functioning of, and the challenges to, banking supervision in the current institutional framework, two aspects should be borne in mind. First, the advent of the euro increases the likelihood of the propagation of financial stability problems across national borders. For this reason a co-ordinated supervisory response is important at an early stage. Second, the sources of banks' risks and stability problems depend on ongoing trends that are not necessarily caused by the euro, but may be significantly accelerated by it. On the whole, we are interested not so much in the effects of EMU or the euro per se, as in the foreseeable developments due to all factors influencing banking in the years to come.

9. It should be noted at the outset that most banking activity, particularly in retail banking, remains confined to national markets. In many Member States the number, and the market share, of banks that operate in a truly nationwide fashion is rather small. Although banks' international operations have increased, credit risks are still predominantly related to domestic clients, and the repercussions of bank failures would be predominantly felt by domestic borrowers and depositors.

10. Assessing the internationalisation of euro area banks is a complex task because internationalisation can take a number of forms. One is via cross-border branches and subsidiaries. Although large-scale entry into foreign banking markets in Europe is still scarce, reflecting persisting legal, cultural and conduct-of-business barriers (less than 10% on average in terms of banking assets in the euro area; Table 1), there are significant exceptions. The assets of the foreign branches and subsidiaries of German and French banks account for roughly a third of the assets of their respective domestic banking systems (Table 2). The Dutch banking system is also strongly diversified internationally.

Another way to spread banking activity beyond national borders is consolidation. Cross-border mergers or acquisitions still seem to be the exception, although things have started to change. The recent wave of "offensive" and "defensive" banking consolidation has mainly developed within national industries, thus significantly increasing concentration, particularly in the smaller countries (Table 3); it may be related not so much to the direct impact of EMU as to globally intensified competition and the need to increase efficiency.

In the coming years internationalisation is likely to increase, because, with the euro, foreign entrants can now fund lending from their domestic retail deposit base or from euro-denominated money and capital markets. The relatively large number of foreign branches and subsidiaries already established could be a sufficient base for an expansion of international banking activity (Table 4) since a single branch, or a small number of branches, may be sufficient to attract customers, especially when they are served through direct banking techniques, such as telephone and Internet banking. Also, the cross-border supply of services on a remote basis is likely to spread as direct banking techniques develop. As to cross-border mergers and acquisitions aimed either at achieving a "critical mass" for wholesale financial markets, or at rapidly acquiring local expertise and customers in the retail sector, they may remain scarce because the cost savings from eliminating overlaps in the retail network are likely to be limited and the managerial costs of integrating different structures and corporate cultures are substantial.

11. However, banks' internationalisation does not provide the full picture of the interconnections of banking systems. As "multi-product" firms, banks operate simultaneously in many markets which have different dimensions: local, national, continental (or European) and global. The advent of the euro is likely to enlarge the market for many banking products and services to the continental dimension; this will "internationalise" even those banks that remain "national" in their branch networks and organisation.

The formation of the single money market in the euro area has largely taken place already. The dispersion in the euro overnight rate across countries, as reported by 57 so-called EONIA banks, fell in January from around 15 to 5 basis points. The variation between banks has been significantly greater than between countries. The TARGET system has rapidly reached the dimension of Fedwire, with a daily average value of payments of €1,000 billion, of which between €300 and €400 are cross-border. The ever stronger interbank and payment system links clearly increase the possibility of financial instabilityspreadingfrom one country to another. Through these links the failure of a major bank could affect the standing of its counterparties in the entire euro area. On the other hand, the deeper money market could absorb any specific problem more easily than before.

As regards the capital markets, the effects of the euro will take more time to manifest themselves, but are likely to be substantial. The single currency offers substantial opportunities for both debt and equity issuers and investors. The increase in the number of market participants operating in the same currency increases the liquidity of the capital markets and reduces the cost of capital. The low level of inflation and nominal interest rates and diminishing public sector deficits are additional supporting factors of capital market activity, especially private bond market activity which has so far been relatively limited (Table 5). Banks will thus operate in increasingly integrated capital markets and will be exposed to shocks originating beyond their national borders.

As to corporations, they may concentrate their operations (treasury, capital market and payment management) in a single or few "euro banks", while the disappearance of national currencies may break links between firms and their home country "house bank". This dissociation would make the domestic economy indirectly sensitive to foreign banks' soundness, thus creating another propagation channel of banking problems across countries.

12. When considering the industry scenario for the coming years, the viewpoint has to be broadened beyond the impact of the euro. Rather than the exclusive, or even primary, force for change, the euro is expected to be a catalyst for pre-existing trends driven by other forces. The recent ECB report prepared by the Banking Supervision Committee on "Possible effects of EMU on the EU banking systems in the medium to long term" gives a comprehensive analysis of such trends, which can be summarised as follows. First, regulation: the industry has yet to feel the full impact of such fundamental, but relatively recent, regulatory changes as those related to the single market legislation. Second, disintermediation: other financial intermediaries and institutional investors will grow relative to banks, pushed by demographic and social changes, as well as by the increasing depth and liquidity of the emerging euro area-wide capital market. Disintermediation is expected to take the form of increasing recourse to capital market instruments relative to bank loans by firms, and diminishing investment in deposits by households relative to mutual funds and related products. Third, information technology: bank products, operations and processes are changing rapidly, while technology offers increasing possibilities for dissociating the supply of a large number of services from branches and face-to-face contact with customers. The current tendency in the EU banking systems to reduce over-branching and over-staffing will grow stronger.

These factors will increase competition, exert pressure on profitability and oblige banks to reconsider their strategies. Such effects are already visible throughout the EU. They produce changes in organisation, new products and services, mergers, strategic alliances, co-operation agreements, etc. They also involve strategic risks, because the pressure for profitability and some losses of revenue due to the euro, for example from foreign exchange, may push some banks to seek more revenue from unfamiliar business or highly risky geographical areas. Inadequate implementation of new technologies or failure to reduce excess capacity may also affect banks' long-term viability. In the short term, the structural adaptation process could be made more difficult by the combination of factors like the protracted financial difficulties of Asia and Russia, or the preparations for the year 2000.

13. Against the background of the institutional framework and the industry scenario I have outlined, let me now turn to the functioning of banking supervision in the euro area. Two preliminary observations. First, the objective of financial stability pursued by banking supervisors is only one in a range of public interests, which also includes competition policy and depositor and investor protection policy. Second, current supervision and crisis management involve different situations and procedures and will therefore be examined in sequence.

14. Starting with current supervision, let me consider banking regulation first. As observed earlier, the regulatory platform for the euro area banking industry combines harmonised rules with country-specific (non-harmonised, but mutually recognised and hence potentially competing) rules.

The harmonised part of the platform includes most of the key prudential provisions that have been developed in national systems over the years. More than 20 years ago (1977), the 1st Banking Co-ordination Directive adopted a definition of a credit institution and prescribed objective criteria for the granting of a banking licence. In 1983 the first Directive on carrying out supervision on a consolidated basis was approved, and in 1986 the rules relating to the preparation of the annual accounts and the consolidated accounts of banks were harmonised. In 1989 the 2nd Banking Co-ordination Directive (which became effective on 1 January 1993) marked the transition from piecemeal to comprehensive legislation, introducing, inter alia, the principle of "home country control". A number of other specific directives have subsequently addressed the main aspects of the regulatory framework - notably, own funds, solvency ratios and large exposures. A Directive imposing deposit guarantee schemes supplemented the legislation in support of financial stability. All in all, the European Union, including the euro area, now has a rather comprehensive "banking law" consistent with the Basle Committee's rules and with the 1997 Core Principles of Banking Supervision.

The country-specific, non-harmonised, part of the platform is also quite relevant and very diversified. It includes, among other things, the different organisational arrangements for the conduct of banking supervision (central bank, separate agency or a mixed arrangement); the tools used by banking supervisors (e.g. supervisory reporting, on-site inspections); provisions for the liquidation and restructuring of banks; and the definition and legal protection of financial instruments and contracts. Even the key notion of a regulated market is harmonised only to a very limited extent.

15. Such "neutrality" and "incompleteness" on the part of the EU legislator with respect to key aspects that are normally incorporated in the regulatory framework is a unique feature of EU banking regulations and is likely to trigger a deregulatory process, pushed by competition among the national systems and the different financial centres in the euro area, and beyond that in the EU. Against the background of the increasing competition and other changes in the banking industry, one can expect that the regulatory platform will evolve in the years to come. Additional EU legislation may prove necessary to complete and strengthen the harmonised part. One important part of common legislation, namely the draft Directive on liquidation and re-organisation measures for credit institutions, has not yet been adopted and, indeed, has been stalled for years. This Directive is needed to bring legal certainty to the framework for banking crisis management. In this regard, it would be useful for the Eurosystem, if necessary, to be able to exclude counterparties from the single monetary policy on prudential grounds. Also, the non-harmonised part of the platform will come under pressure to converge, as I have just mentioned, through the process of "regulatory competition". Like any other rapidly changing industry, the banking sector will require careful attention by regulators. As indicated earlier, the ECB will have the possibility of contributing to the rule-making process through its advisory tasks under Article 105 (4) of the Treaty and Article 25.1 of the Statute of the ESCB.

16. On the whole, and taking a euro area perspective, the legislative-cum-regulatory platform of the banking industry, although rather unusual and very diversified in comparison with those of most currency jurisdictions, does not seem to present loopholes or inconsistencies that may hamper the pursuit of systemic stability. Seen from the point of view of the regulatory burden, it is a light system. It will become even more so if competition among national banking systems and financial centres encourages national regulators to free their banks from regulatory burdens that are not required by the EU Directives. Conversely, seen from the point of view of its flexibility, i.e. how quickly it can adapt to new situations, it is, on the contrary, a heavy system. This is the case both because the EU legislative process is slow (three years or even longer may be needed to pass Directives) and, perhaps more importantly, because many provisions are embodied in the Community primary legislation (i.e. Directives) rather than in Community secondary legislation (amendable through simpler comitology procedures).

The establishment of EMU does not seem to determine a need for revising the pillars of the current legal framework. What seems to be necessary, however, is a more flexible legislative procedure which allows for a faster and more effective revision of Community legislation, whenever needed in relation to market developments.

17. Let me now turn to the execution of banking supervision. It should immediately be recalled that supervision, contrary to regulation, is a national task, exercised by what the jargon of the Directives calls the "competent authority". Since the euro area has adopted a separation approach between supervisory and central banking functions, it is natural to examine first the functioning of the "euro area supervisor" (i.e. the co-operative system of national supervisors) and then turn to the tasks and needs of the "euro area central banker" (i.e. the Eurosystem).

18. The euro area supervisor can be regarded as a rather peculiar entity composed of national agencies working in three modes: stand-alone, bilateral and multilateral. Let us briefly examine each of them.

The stand-alone mode is the one in which the supervisor exclusively operates in the national (or even local) context. Today it is by far the most predominant mode. In most cases, this approach is sufficient to achieve the objectives of banking supervision because most banks in Europe are operating in a context that does not even reach the nationwide market of the country of origin. Such a decentralised model is even more effective because it allows the efficient use of information that may not be available far from the market in which the bank operates. That is why it is actually applied even within countries. In Italy, for example, over 600 of the 900 licensed credit institutions at end-1998 were entirely supervised by the Banca d'Italia branch of the town in which the bank is licensed.

The bilateral mode involves co-operation between two supervisory agencies. It is used for cross-border supervision of the same type of financial institutions, such as credit institutions, or the supervision of different types of financial institutions operating in the same market, such as credit institutions and securities firms. The instrument that has been devised to organise bilateral co-operation between banking supervisors is the Memorandum of Understanding (MoU). With the implementation of the 2nd Banking Co-ordination Directive, the Member States began to negotiate extensively MoUs in order to establish the necessary co-operation between "home" and "host country" authorities to supervise efficiently institutions that have cross-border activities or foreign country establishments.

By the end of 1997, 78 bilateral MoUs had been signed between the EEA banking supervisory authorities. The key aims of MoUs are to establish a regular exchange of information between national supervisory authorities. While the "gateways" for the exchange of information have been laid down in Community legislation, MoUs provide a practical framework for communication to be carried out between supervisors. Moreover, MoUs define procedures and reciprocal commitments between pairs of EU supervisors related to the various parts of the supervisory process, such as establishment procedures and on-site examinations.

Finally, the multilateral mode is the one in which a group of supervisors works collectively as, say, a single consolidated supervisor. Such a mode is required when the problems involved are area-wide. They may be area-wide for a number of reasons with regard to the institutions, or groups, involved: their dimension; their linkages with a number of different markets in various countries; the role they play in the payment system or in other "systemic" components of the market, etc. Multilateral co-operation can also enhance the quality of supervision by examining common macroeconomic influences on the banking system and common trends in the financial system that may not be revealed from the national perspective only.

Today, the Banking Supervision Committee is the key forum for multilateral co-operation. It is composed of representatives of the banking supervisory authorities of the EU countries, either forming part of the respective NCB or separate bodies. The Banking Supervision Committee's main functions are the promotion of a smooth exchange of information between the Eurosystem and national supervisory authorities and co-operation among EU supervisory authorities. Another forum for dealing with the requirements of the multilateral mode is the Groupe de Contact, a group of EU banking supervisory authorities which, for many years, has discussed individual banking cases in a multilateral way, but at a lower organisational level than the high-level Banking Supervision Committee.

19. So far, the need to develop the multilateral mode has been relatively limited, as the emergence of a single banking market in the European Union has been slow and the euro was not yet in place. Thus, the fact that the multilateral mode has not gone, for the moment, beyond periodic discussions among supervisors and occasional industry-wide analyses should not be a cause for concern.

I am convinced, however, that in the future the needs will change and the multilateral mode will have to deepen substantially. Over time such a mode will have to be structured to the point of providing the banking industry with a true and effective collective euro area supervisor. It will have to be enhanced to the full extent required for banking supervision in the euro area to be as prompt and effective as it is within a single nation.

There are no legal impediments to that. The existing legislation, whether Community or national, permits all the necessary steps to be made. Information can be pooled; reporting requirements and examination practices can be developed and standardised; common databases can be created; joint teams can be formed; and analyses of developments across the whole banking system can be conducted. The Community legislation providing for the unconstrained exchange of confidential information between supervisors does not distinguish between bilateral and multilateral co-operation, but the common interpretation is that it covers both modes. It will be the task of the Banking Supervision Committee, for its part, to develop the multilateral mode among EU banking supervisors.

20. If the above concerns primarily the euro area supervisor, what about the euro area central banker, i.e. the Eurosystem? The euro area central banker has neither direct responsibility for supervising banks nor for bank stability. It is, however, no stranger in this land. It has a vital interest in a stable and efficient banking industry; it is, therefore, keen to see its action complemented with an effective conduct of the supervisory functions by the competent authorities; it needs a clear and precise knowledge of the state of the euro area's banking industry as a whole and of its major individual players; and it may have a role to play, as we shall see, in the management of crises.

For the Eurosystem, natural reference models are provided by the central banks of countries that apply the separation approach, for example: Germany before the euro; the United Kingdom after the creation of the Financial Services Authority; or Japan. In all these cases the central bank has a well-developed expertise in the micro and macro-prudential field; each distinctively plays a role in the macro-prudential field by addressing threats to the stability of the banking system and analysing the soundness of the structural features of the system. For their own purposes, these central banks also have precise and comprehensive information about the banks in their respective country. This is obtained either from performing practical supervisory duties, as in the case of the Bank of Japan or the Bundesbank; or from the national supervisory authority; or through direct contacts with the banking industry, as in the case of the Bank of England.

The Banking Supervision Committee is in a good position to co-operate with the Eurosystem in the collection of information. Indeed, the so-called BCCI Directive has removed the legal obstacles to the transmission of confidential information from competent supervisory authorities to "central banks and other bodies with a similar function in their capacity as monetary authorities". This includes national central banks and the ECB. Of course, the provision of supervisory information is voluntary and its development will have to be based on an agreed view of the central banking requirements the Eurosystem will have in this field.

21. In normal circumstances central banking and prudential supervision have an arm's length distance between them. In crisis situations, however, they need to act closely together, often in co-operation with other authorities as well. Charles Goodhart and Dirk Schoenmaker have made here at the London School of Economics a valuable contribution to analysing the handling of major banking problems in the history of industrial countries. One of their conclusions is that, in most instances, central banks have indeed been involved. Banking problems are so close to monetary stability, payment system integrity and liquidity management that this finding hardly comes as a surprise. The advent of the euro will not, by itself, change this state of affairs.

22. When discussing crisis management, it should not be forgotten that, while central banks have a direct and unique role to play when the creation of central bank money is involved, this represents just one category of emergency action. Another category refers to the injection - by politically liable Finance Ministries - of taxpayers' money into ailing or insolvent credit institutions. There is also a third, market-based, category, consisting of the injection of private money by banks or other market participants. These three typologies of emergency action all require the involvement of policy-makers, but they must not be mixed up when evaluating the existing arrangements. Therefore, before discussing the much debated question of the lender-of-last-resort, let me briefly comment on the two, probably less controversial cases where central bankers are not the providers of extra funds.

23. First, the "private money solution". This market-based approach is clearly the preferable option, not just to save public funds and avoid imbalances in public finances, but also to reduce the moral hazard problem generated by public assistance to ailing institutions. Indeed, policy-makers are increasingly aware that the expectations of a helping hand can increase financial institutions' risk appetite in the first place. However, even when a market-based solution is possible, on the grounds of private interest, private parties may not be able to reach a solution for lack of information or co-ordination. Public authorities have therefore an active role to play for the market solution to materialise. The recent rescue package co-ordinated by the Federal Reserve Bank of New York to prevent the LTCM hedge fund from collapsing is a good example of public intervention being used to achieve a private solution.

Acting as a "midwife" in brokering a private sector deal is not the only example of managing crises without injecting public funds. Banking supervisors have at their disposal a number of tools to intervene at the national level to limit losses and prevent insolvency when a bank faces difficulties. These tools include special audits, business restrictions and various reorganisation measures.

In the euro area, national supervisors and central banks will continue to be the key actors in the pursuit of market-based solutions to crises. The Eurosystem, or the Banking Supervision Committee, would become naturally involved whenever the relevance of the crisis required it.

24. Second, the "taxpayers' money solution". Taxpayers have been forced to shoulder banks' losses in the past, when public authorities felt that otherwise the failure of a large portion of a country's banking system or of a single significant institution would have disrupted financial stability and caused negative macroeconomic consequences. In such instances banks have been taken over by the state, or their bad assets have been transferred to a separate public entity to attract new private investment in the sound part of the otherwise failed banks. The US savings & loans crisis of the 1980s, the banking crises in Scandinavia in the early 1990s and the current banking crises in Japan and some East-Asian countries are examples of system-wide insolvency problems that have triggered taxpayers' support. Crédit Lyonnais and Banco di Napoli are recent examples of public support to individual insolvency problems.

The introduction of the euro leaves crisis management actions involving taxpayers' money practically unaffected. The option of injecting equity or other funds remains available for the Member States, since these operations are not forbidden by the Treaty. Nevertheless, the European Commission will be directly involved in scrutinising and authorising such actions, since any state aid must be compatible with the Community's competition legislation. This happened, for example, in the cases of Banco di Napoli and Crédit Lyonnais.

The handling of solvency crises is not within the competence of the national central banks nor that of the ECB, although national central banks are likely to be consulted, as they have been in the past.

25. Third, the "central bank money solution". This is the lender-of-last-resort issue that has brought the Eurosystem under vigorous criticism by distinguished academics and the IMF's Capital Markets Division of the Research Department. The criticism has been that the alleged absence of a clear and transparent mechanism to act in an emergency raises doubts in the markets about the ability of the Eurosystem to handle crisis situations. It is said that the uncertainty generated by the present arrangements would entail new risks, including the possibility of investors requiring an additional risk premium at times of financial market volatility and, ultimately, of the credibility of EMU being damaged. Two examples of these concerns deserve an explicit mention. The IMF "Report on Capital Markets", September 1998, stated that "it is unclear how a bank crisis would be handled under the current institutional framework - which is not likely to be sustainable".Similarly, the first report of the CEPR (Centre for Economic Policy Research) on monitoring the ECB entitled "The ECB: Safe at Any Speed?" expressly suggested that the Eurosystem lacks crisis management capacity and is too rigid to pass the A-Class test to keep the vehicle on the road at the first steep turn in financial market conditions in Europe.

26. My response to this criticism is threefold. To my mind, the criticism reflects a notion of lender-of-last-resort operations that is largely outdated; it underestimates the Eurosystem's capacity to act; and, finally, it represents too mechanistic a view of how a crisis is, and should be, managed in practice.

27. The notion of a central bank's lender-of-last-resort function dates back more than 120 years, to the time of Bagehot. This notion refers to emergency lending to institutions that, although solvent, suffer a rapid liquidity outflow due to a sudden collapse in depositors' confidence, i.e. a classic bank run. A bank could be exposed to depositors' panic even if solvent because of the limited amount of bank liquidity and an information asymmetry between the depositors and the bank concerning the quality of bank's assets that do not have a secondary market value.

Nowadays and in our industrial economies, runs may occur mainly in textbooks. They have little relevance in reality because, since Bagehot, many antidotes have been adopted: deposit insurance, the regulation of capital adequacy and large exposures, improved licensing and supervisory standards all contribute to the preservation of depositors' confidence and minimise the threat of a contagion from insolvent to solvent institutions.

A less unlikely case is a rapid outflow of uninsured interbank liabilities. However, since interbank counterparties are much better informed than depositors, this event would typically require the market to have a strong suspicion that the bank is actually insolvent. If such a suspicion were to be unfounded and not generalised, the width and depth of today's interbank market is such that other institutions would probably replace (possibly with the encouragement of the public authorities as described above) those which withdraw their funds. It should be noted, in this respect, that the emergence of the single euro money market lowers banks' liquidity risk, because the number of possible sources of funds is now considerably larger than in the past.

Given all of these contingencies, the probability that a modern bank is solvent, but illiquid, and at the same time lacks sufficient collateral to obtain regular central bank funding, is, in my view, quite small. The textbook case for emergency liquidity assistance to individual solvent institutions has, as a matter of fact, been a most rare event in industrial countries over the past decades.

28. What if this rare event were nevertheless to occur and cause a systemic threat? The clear answer is that the euro area authorities would have the necessary capacity to act. This is not only my judgement, but also that of the Eurosystem, whose decision-making bodies have, as you can imagine, carefully discussed the matter. I am not saying that we are, or shall be, infallible; no one can claim such a divine quality. I am saying that there are neither legal-cum-institutional, nor organisational, nor intellectual impediments to acting when needed. In stating this, I am aware that central banks may be the only source of immediate and adequate funds when a crisis requires swift action, while solvency remains an issue and failure to act could threaten the stability of the financial system.

In these circumstances the various national arrangements would continue to apply, including those concerning the access of central banks to supervisors' confidential information. As is well known, such arrangements differ somewhat from country to country.

29. The criticism I have referred to also underestimates the Eurosystem's capacity to act. To the extent that there would be an overall liquidity effect that is relevant for monetary policy or a financial stability implication for the euro area, the Eurosystem itself would be actively involved.

The Eurosystem is, of course, well equipped for its two collective decision-making bodies (the Board and the Council) to take decisions quickly whenever needed, whether for financial stability or for other reasons. This readiness is needed for a variety of typical central bank decisions, such as the execution of concerted interventions or the handling of payment system problems. Indeed, it has already been put to work during the changeover weekend and in the first few weeks of this year.

A clear reassurance about the capacity to act when really needed should be sufficient for the markets. Indeed, it may even be advisable not to spell out beforehand the procedural and practical details of emergency actions. As Gerry Corrigan once put it, maintaining "constructive ambiguity" in these matters may help to reduce the moral hazard associated with a safety net. I know of no central bank law within which the lender-of-last-resort function is explicitly defined.

The question of who acts within the Eurosystem should also be irrelevant for the markets, given that any supervised institution has an unambiguously identified supervisor and national central bank. As to the access to supervisory information, the lack of direct access by the Eurosystem should not be regarded as a specific flaw of the euro area's institutional framework, as has been frequently argued, since this situation also exists at the national level wherever a central bank does not carry out day-to-day supervision.

30. Finally, the criticism reflects an overly mechanistic view of how a crisis is, and should be, managed in practice. Arguing in favour of fully disclosed, rule-based policies in order to manage crises successfully and, hence, maintain market confidence, is almost self-contradictory. Emergency situations always contain unforeseen events and novel features, and emergency, by its very nature, is something that allows and even requires a departure from the rules and procedures adopted for normal times or even in the previous crisis. Who cares so much about the red light when there is two metres of snow on the road? As for transparency and accountability, these two sacrosanct requirements should not be pushed to the point of being detrimental to the very objective for which a policy instrument is created. Full explanations of the actions taken and procedures followed may be appropriate ex post, but unnecessary and undesirable ex ante.

31. So far, I have focused on the provision of emergency liquidity to a bank. This is not the only case, however, in which central bank money may have to be created to avoid a systemic crisis. A general liquidity "dry-up" may reflect, for example, a gridlock in the payment system or a sudden drop in stock market prices. The actions of the Federal Reserve in response to the stock market crash of 1987 is an often cited example of a successful central bank operation used to prevent a dangerous market-wide liquidity shortfall. This kind of action is close to the monetary policy function and has been called the "market operations approach" to lending of last resort. In such cases, liquidity shortfalls could be covered through collateralised intraday or overnight credit, or auctioning extra liquidity to the market. The Eurosystem is prepared to handle this kind of market disturbance.

32. In my remarks this evening, I have looked at the euro area as one that has a central bank which does not carry out banking supervision. This would be normal, because in many countries banking supervision is not a task of the central bank. What is unique is that the areas of jurisdiction of monetary policy and of banking supervision do not coincide. This situation requires, first of all, the establishment of smooth co-operation between the Eurosystem and the national banking supervisors, as is the case at the national level where the two functions are separated. The most prominent reason for this is, of course, the scenario where the provision of liquidity from the central bank has to be made in a situation that is generated by problems of interest to the supervisor. But beyond that, I do not know any country in which the central bank is not very closely interested in the state of health of the banking system, irrespective of its supervisory responsibilities.

33. In my view, we should move as rapidly as possible to a model in which the present division of the geographical and functional jurisdiction between monetary policy and banking supervision plays no significant role. I do not mean necessarily a single authority or a single set of prudential rules. Rather I mean that the system of national supervisors needs to operate as effectively as a single authority when needed. While the causes of banking problems are often local or national, the propagation of problems may be area-wide. The banking industry is much more of a system than other financial institutions.

34. I am clearly aware that we are far from having a common supervisory system. But since the euro has just been launched and will last, we have to look in prospective terms at what needs to be set in place. There is no expectation, at least to my mind, that the division of responsibility in the euro area between the central bank and the banking supervisory functions should be abandoned. Although the Treaty has a provision that permits the assignment of supervisory tasks to the ECB, I personally do not rely on the assumption that this clause will be activated. What I perceive as absolutely necessary, however, is that co-operation among banking supervisors, which is largely voluntary but which finds no obstacles in the existing Directives or in the Treaty, will allow a sort of euro area collective supervisor to emerge that can act as effectively as if there were a single supervisor. This is desirable in the first instance to render the supervisory action more effective against the background of current and future challenges and, second, to assist the Eurosystem in the performance of its basic tasks.

TABLES

Table 1. Market share of branches and subsidiaries of foreign credit institutions as % of total domestic assets, 1997

Source: ECB report "Possible effects of EMU on the EU banking systems in the medium to long term" (February 1999).

Table 2. Assets of branches and subsidiaries of domestic credit institutions in foreign countries as % of total domestic assets, 1997

Source: ECB report "Possible effects of EMU on the EU banking systems in the medium to long term" (February 1999).

Table 3. Concentration: Assets of the five biggest credit institutions as % of total assets

Source: ECB report "Possible effects of EMU on the EU banking systems in the medium to long term" (February 1999).

Table 4. Number of branches and subsidiaries of foreign credit institutions, 1997

Source: ECB report "Possible effects of EMU on the EU banking systems in the medium to long term" (February 1999).

Table 5. Private non-financial enterprises' bonds, credit institutions' bonds and government bonds outstanding as % of GDP, 1997

Source: ECB report "Possible effects of EMU on the EU banking systems in the medium to long term" (February 1999).