Payments and the Eurosystem

Speech by Tommaso Padoa-Schioppa, Member of the Executive Board of the European Central Bank, Sibos, Munich, 13 September 1999


  1. Introduction

  2. A glance at two centuries

  3. The role of the Eurosystem

  4. The first fifteen months

  5. The main policy issues today

  6. Conclusion

I. Introduction

1. Payment systems and securities settlement systems (PSSS) represent one of the basic infrastructures needed for the proper functioning of market economies, as vital as the legal or the transport system. They are an essential component of the transmission mechanism for monetary policy. The soundness and the smooth operation of PSSS are therefore indispensable for the stability of the currency, the financial system and the economy in general.

In the very same years in which the Treaty of Maastricht was being negociated, a technological revolution was taking place that highlighted the relevant public interest at stake in PSSS. This led the Treaty to give explicit and considerable responsibilities to the Eurosystem in the field. And since the central bank of the euro only came into existence fifteen months ago, both its reflections and its actions about the appropriate fulfilment of such responsibilities must be carefully weighted, knowing the importance of the age of foundation for the long future awaiting the euro.

The relevance of PSSS issues is only matched by the persistent lack of familiarity with it of most central bank observers, whether in academia, in the media, or in the market. An effort will therefore be made in this paper to explain the issues in a way that requires little prior knowledge from the reader.

The next Section will recall why central banks have always been concerned with the efficiency and safety of payment arrangements and how, in practice, their responsibilities have evolved over time. Section III will analyse the statutory tasks of the ECB and the Eurosystem in the field of PSSS. Section IV summarises the activities of the Eurosystem in the first fifteen months of its existence. Finally, and more importantly, Section V tries to review the main policy issues in which the Eurosystem is likely to be involved over time.

II. A glance at two centuries

2. The origin of central banks is closely linked to a revolutionary change experienced by monetary systems in the course of the 19th century: the replacement of commodity currencies with paper currency as means of payment.

While a commodity currency was supported by its intrinsic value, the value of paper money depended on the confidence in the issuer's ability to redeem paper money in commodity money at par. The "act of faith" on which paper-based systems rely is still reflected in the vocabulary: banknotes are called monnaie fiduciaire in French, moneta fiduciaria in Italian and fiat money in the English literature. Freely issued banknotes entailed uncertainties and risks. They were difficult to recognise because there were so many of them. Their value could diminish or even disappear whenever doubts arose about the creditworthiness of the issuer. To ensure the efficiency and the safety of paper currency systems, public authorities gradually entrusted one single commercial bank with the task of issuing banknotes. In Europe, this bank became known as bank of issue.

3. A second revolutionary change occurred during the second half of the 19th century and the beginning of the 20th century with the development of bank money, made material by means of entries in the books of the banks: monnaie scripturale in French, moneta scritturale in Italian. Here again, the value of this form of money depends on the creditworthiness of the issuer. Yet, unlike the issuance of banknotes, the issuance of bank money has remained decentralised within the banking sector.

Today, the public generally consider banknotes and commercial bank money as perfectly interchangeable and equally safe. Banks, however, have always made the difference and have insisted on interbank claims being settled in fiat money:originally banknotes, and, progressively, the dematerialised form of book entries at the banks of issue. Banks of issue became to be known as central banks and their sight liabilities (in paper or book-entry form) became for banks and other market participants the ultimate way to settle obligations. Central bank money also became legal tender, i.e. the settlement means which market participants must accept. In the following, I shall use the terms "final money" or "central bank money" interchangeably.

4. This historical development shows how much the raison d'être of central banks lies in the promotion of the efficiency and the safety of the payment arrangements. Indeed, the two other key central banking functions, supervision of commercial banks and monetary policy, stem from the fact that central banks are the issuers of final money. Banking supervision developed because central banks needed to evaluate the soundness of the claims on banks which they accepted as a counterpart for the issuance of final money. Monetary policy, whichdeveloped only subsequently in this century, was made possible by the gradual loosening (and eventually the severance) of the link between central bank money and gold. Nowadays, the largest proportion of the stock of money are liabilities of commercial (rather than central) banks, and confidence in the currency ultimately rests on two pillars: first, the possibility of converting commercial bank money into central bank money at par at any moment and, second, the ability of the central bank to maintain price stability.

5. After these revolutionary changes, payment technology settled down. Many decades without changes made central banks' functions in the payment field gradually less prominent than banking supervision and monetary policy. The efficiency of the payment system came to be considered a merely technical matter, what Gerry Corrigan called the "plumbing" of the monetary system, an area generally dealt with by a sub-unit of the IT department in most central banks. Although payment systems progressively moved from paper-based into electronic systems, their organisational model remained the 19th century-style clearing house, with multilateral netting of payments and settlement of net balances in central bank money at the end of each day. Exceptions could be found only in countries, such as Germany or the United States, where a decentralised organisation of the central bank could hardly be reconciled with the central clearing house model. In these countries, interbank obligations were settled partly on a gross basis to facilitate the transfer of funds from one central bank entity to another.

6. In the middle of the 1970s a new revolution started. It was due to two developments. First, the exponential increase in the value of the payments produced by the phenomenal expansion of financial transactions. Second, the advent of new money transfer procedures combining electronic data processing and telecommunication technology.

Central banks quickly understood that settlement risks in the traditional clearing house model had become unmanageable. Payment system issues moved from the back offices to the board rooms again, in both central banks and market participants Central banks reacted to the new developments in two ways: first, they required substantial improvements in the safety features of netting systems and, second, they promoted real-time gross settlement (RTGS) systems. Both reactions were largely organised through international co-operation.

The first was developed in 1990 at the G10 level, through the "Report on Interbank Netting Schemes",also known as the Lamfalussy ReportThe report set out minimum standards to ensure that netting systems would be able to complete the settlement phase even in the event of the failure of the participant with the highest debit position. Following the Lamfalussy Report, most netting systems in the world have revised and improved their operational rules and procedures.

The second reaction, the promotion of RTGS systems, was started by the pioneering precedent of the Federal Reserve System in the United States and developed at the EU level within the framework of the European Monetary Institute. RTGS systems allow the intraday balances between banks to be eliminated and therefore radically reduce systemic risks. Today, technological progress has allowed commercial banks to use central bank money as a means of payment on a continuous basis throughout the day. The velocity of circulation of central bank money has risen so much that efficiency no longer requires netting. The evolution was particularly remarkable in the EU. In 1993 the "market share" of RTGS systems was about 5% in value terms. In 1998 it reached 50%. In "Euroland", after the successful introduction of TARGET, the share is now close to 70%.

7. Payment systems oversight isthe expression used to describe the function of central banks aiming at ensuring the efficiency and the soundness of payment systems. It shares with prudential supervision the objective of financial stability. Both functions aim to reduce the risk of instability. However, while prudential supervision looks at institutions, payment systems oversight looks at systems. Another significant difference is that, while central banks have progressively abandoned commercial lending and deposit collection to avoid conflicts of interest, in the field of payment systems they are not only overseers, but also service providers. This is unavoidable because central bank money has no substitute as it is the sole means that provides finality to payments.

8. A more recent evolution is the extension of payment systems oversight to securities settlement systems. In the settlement of securities transactions central banks used to be involved only as agents for the management of government debt. Nowadays, they are concerned about the smooth functioning of overall systems, because of their link to payment circuits and monetary policy. Since this evolution is not always understood, and sometimes even challenged, it is useful to devote a few words to it.

For safety reasons, securities transactions have come to be more and more settled, over recent years, on a delivery-versus-payment basis. As a result, banks include the payment flows stemming from securities settlement in their intraday liquidity management. If these funds are not delivered, or not delivered on time, payment settlement systems could become gridlocked. Moreover, since the liquidity in RTGS systems depends on intraday credit that central banks only grant against collateral, RTGS systems would be blocked if securities were not delivered to the central bank on time. This would in turn seriously hamper the implementation of monetary policy.

III. The role of the Eurosystem

9. This glance at two centuries helps to understand that ensuring not only a sound, but also an efficient payment mechanism has been a core function of central banks from their very beginnings and throughout their historical development. It also gives an idea of the richness of the legacy inherited by the central bank of the euro area when it was established last year. I now turn tothe heir, i.e. the Eurosystem.

The name "Eurosystem" has been coined to designate the European Central Bank (ECB) and the National Central Banks (NCBs) of the Member States which have adopted the euro. It is the central bank of the euro and of Euroland, in the same way as the Federal Reserve System is the central bank of the dollar and of the United States. The Member States which do not participate in the euro have retained their separate currencies and continue to conduct independent monetary policies. Their central banks are not part of the central bank of the euro.

Although a federal and decentralised central bank is not a novelty, the Eurosystem is a special case. The economy and the financial system of the euro area have a much deeper (national) segmentation than any other. The NCBs that are part of the Eurosystem have for many generations performed the full range of central banking functions on their own responsibility and in a national context. They have been accountable to, and sometimes dependent on, national institutions. Public opinion has perceived, and still perceives, them as national entities. The notion of the public interest they were referring to was the notion of a national interest. Significant differences existed, and partly remain, in their tasks, organisation, statutes and cultures. In this situation, the Eurosystem has to make an appropriate distinction between being national in the organisational sense and being euro area-wide in the definition of the public interest pursued. This distinction also needs to be drawn in the field of PSSS.

10. From the very beginning of the debate on Monetary Union, it was agreed that the future central bank of Europe would have responsibilities in the field of payment systems. I remember the then President of de Nederlandsche Bank, Wim Duisenberg, proposing to the Delors Committee (on which I sat as co-secretary) that the future European central bank should be "responsible for the maintenance of a properly functioning payment system." This was in the autumn of 1988; the Delors Report was published in April 1989.

The Treaty followed the recommendations of the Delors Report, which in turn took stock of the revolution experienced by central banks in this field. As holds true of all central bank legislation and statutes adopted at the beginning of the 1990s, the Treaty explicitly recognises both the oversight and the operational functions of the Eurosystem. Its Article 105 (2) states that one of the basic tasks of the Eurosystem is to "promote the smooth operation of payment systems". The annexed Statute goes further, granting regulatory powers to the ECB under Article 22: "The ECB and national central banks may provide facilities, and the ECB may make regulations, to ensure efficient and sound clearing and payment systems within the Community and with other countries".

It should be noted that the Treaty makes no distinction between retail and large-value payments. In this respect, it does not follow the doctrine according to which central banks should concentrate only on systemic risk issues, which have limited relevance for retail payments. On the contrary, the doctrine implicit in the Treaty is that fostering the efficiency and safety of retail payment systems is an integral part of the central bank's responsibility for the currency. I think this is the right approach, because the public's confidence in the currency would be undermined by financially unsound providers of retail payment services, as well as by an inefficient functioning of retail payment systems. This explains why the Eurosystem is concerned with retail payment systems and, in particular, with their ability to handle the circulation of money in a reliable way and at a low cost. I shall come back to this point.

Another observation concerns securities settlement systems. The Treaty (like most other central bank laws) does not explicitly mention this field. The reason is that, contrary to the oversight of payment systems, central banks' involvement in securities settlement systems has only developed after the adoption of the Treaty. Yet, as we saw earlier, central banks need to be increasingly concerned about the smooth functioning of securities settlement systems as a consequence of their statutory duties in the field of payment systems and monetary policy. The Eurosystem will not be an exception.

11. Within this framework, the Eurosystem will exert its statutory tasks in accordance with two general principles stipulated by the Treaty itself: the market principle and the decentralisation principle. Let's briefly examine each of them.

The market principle, laid down in Article 105 (1) of the Treaty, states that: "the [Eurosystem] shall act in accordance with the principle of an open market economy with free competition, favouring an efficient allocation of resources (...)".This requires the Eurosystem to take a neutral position in competition between (i) systems, (ii) financial centres and (iii) categories of banks. The important implications of the principle of neutrality can be seen in the light of the profound structural and functional changes undergone by national securities and settlement systems in recent years. These changes were largely driven by market forces, as a result of innovative developments in payment technologies, and by the greater integration of the European economies. As a result, PSSS have progressively moved towards increased, although uneven, concentration and specialisation. This in turn has prompted a radical restructuring and reorganisation of the industry.

Figures speak eloquently. In 1997 the euro area (still split in 11 countries) had 18 large-value payment systems, 23 securities settlement systems and 13 retail payment systems. The United States has 2 large-value systems, 3 securities settlement systems and 3 retail payment systems. This striking difference indicates unmistakably the type of evolution one can expect in Europe for purposes of reaping the full benefits of economies of scale. A vast reorganisation is under way and already well advanced in large-value payment systems. By contrast, securities settlement systems have only just started to move, and the restructuring of retail payment systems has yet to come.

As long as market forces do not lead towards unsound or inefficient solutions, the Eurosystem will remain neutral in the current and prospective restructuring of the financial infrastructure. Indeed, in a market economy public authorities should intervene only when there is a strong public interest, which the market fails to address.

12. The decentralisation principle stems from Article 12.1 of the ECB Statute, which states that "to the extent deemed possible and appropriate, (...) the ECB shall have recourse to the national central banks to carry out the operations which form part of the tasks of the [Eurosystem]". Following this principle, national central banks will continue to be involved in payment system operations in different ways, in line with their national traditions but in accordance with the jointly defined policy. Also the actual implementation of oversight, including the discussions with banks and system operators, will take place at the local level whenever a system is strongly rooted in one individual country.

It should be noted that the implementation of the decentralisation principle does not mean that the responsibility for operational tasks will fall exclusively onto the NCBs and that oversight responsibilities will fall only onto the Eurosystem. A clear distinction will have to be made between what can be handled at the national level and what must be handled at the Eurosystem's level. For example, the ECB is operationally involved in the end-of-day procedures of TARGET and many operational features of TARGET have been decided collectively. As for oversight, the Eurosystem will in many instances define only the general principles that need to be followed throughout the euro area in order to ensure a level playing-field. Electronic money is a good example of a subject to be discussed at the Eurosystem's level. The elaboration of a single policy line is also required for cross-border systems, such as the Euro 1 system of the Euro Banking Association (EBA), the oversight of which is performed by the ECB. Action at the Eurosystem level could also be undertaken in the field of cross-border retail payments in euro.

IV. The first fifteen months

13. When the ECB was established, in June 1998, technical preparations were well advanced. Yet, many important decisions could only be taken after the establishment of the ECB and of the Eurosystem. Let me briefly review them, under the three headings of large value payments, securities settlement and retail payments.

14. In the field of large value payments, the Eurosystem had to achieve two objectives before the start of EMU: the creation of TARGET and the evaluation of the soundness of the other systems which would continue to process large-value payments in parallel with TARGET.

TARGET is an acronym standing for Trans-European Automated Real-time Gross settlement Express Transfer system. It is the RTGS system which the euro needed to achieve between the national central banks of the euro area the same easy transferability of central bank money that existed within countries. Without such a system, supply and demand for central bank funds would have had to be met on the books of each of the NCBs. As a result, money market interest rates would have differed among financial centres, making the implementation of a single monetary policy very difficult, if not impossible. In this respect, the creation of TARGET by the Eurosystem mirrors the creation of Fedwire by the Federal Reserve System in 1918, shortly after its establishment.

The establishment of TARGET was, however, also a consequence of the central banks' willingness, in the course of the 1990s, to develop RTGS systems as a safe way to process payments and to minimise systemic risk. TARGET had indeed two objectives: to serve the needs of the single monetary policy and to achieve high security in the payment system infrastructure in the euro area.

TARGET consists of the fifteen national RTGS systems of the EU countries and the ECB payment mechanism (EPM). These are interlinked so as to provide a uniform platform for the processing of domestic and cross-border payments within the euro area. Although technical and organisational features continue to differ across NCBs, the TARGET concept has been built in such a way that the use of the system in the domestic and in the cross-border mode is very similar for commercial banks. In accordance with the market principle, the use of TARGET is not mandatory, except for the few payments that are directly connected with monetary policy operations or with the settlement of large-value netting systems in euro.

15. In the first weeks of its existence, the ECB Governing Council took a series of key decisions which could not be taken earlier. It adopted a legal guideline addressing the relationship between the ECB and all participating NCBs. It decided on the price structure for cross-border TARGET payments. Finally, it set the conditions under which banks and national central banks of the EU countries, which had not adopted the euro, could connect to TARGET.

This last point requires a short explanation. Although the Treaty does not grant "out" NCBs any special right to be a part of TARGET, in 1995 the EMI Council decided that all EU NCBs would be allowed to connect to TARGET. The decision was based on the consideration that the technical preparations had to start long before the selection of EU Member States participating in EMU. The EMI indicated, however, that the connection of non-participating central banks would be subject to conditions. Against this background, the ECB Governing Council decided in July 1998 on the conditions under which these central banks could also provide intra-day liquidity in euro to their banks. The ensuing arrangement was a very innovative one, because it was the first time that a central bank allowed other central banks to provide settlement facilities in its currency. A "Policy statement" issued by the ECB in November 1998 clarified that central bank money in euro can only be provided by the central banks belonging to the Eurosystem and indicated that the possibility offered to "out" central banks was a specific exception. Moreover, the policy statement invited any central bank located outside the euro area to inform the ECB whenever they became aware of a planned settlement arrangement in euro.

16. TARGET started its operations on 4 January 1999. The statistical data show that the system now settles two-thirds of the value of all payments processed in euro and that both the volumes and values are stable. In July 1999, a daily average of nearly 170,000 payments was recorded, for a value of more than EUR 920 billion. During that month, domestic TARGET payments represented an average value of EUR 550 billion per day, whereas cross-border TARGET payments reached EUR 370 billion. The high proportion of cross-border payments is an indication of the successful integration of the euro area money market. The importance of payment flows processed also shows that TARGET has met a market need. The figures can be compared with those of Fedwire in the United States which on average processed about 390,000 payments daily in 1998, for a value of about USD1,300 billion.

17. The second objective of the Eurosystem in the field of large-value payments concerns the safety features of systems which continue to operate in parallel to TARGET. There are six such systems. Three - in Luxembourg, Finland and Spain - are very small (they process in total less than EUR 10 billion daily); the three remaining ones are the Euro 1 system, Euro Access Frankfurt (EAF) and Paris Net Settlement (PNS). The ECB Governing Council assessed the soundness of these systems prior to authorising them to settle in the books of the Eurosystem.

The payment systems infrastructure of the euro area is very sound since two-thirds of the value of the payments exchanged within the area are made in the RTGS mode, while the rest is settled in systems which meet, as a minimum, the standards set out in the Lamfalussy Report.

18. I now turn to the activities of the Eurosystem in the field of securities settlement systems. The preparatory work conducted in this field aimed at establishing an appropriate infrastructure for the custody and settlement of assets used as collateral in Eurosystem credit operations. Two main actions related to the management of collateral were carried out in this respect: the establishment of the "Correspondent Central Bank Model" (CCBM) and the definition of standards for the use of securities settlement systems in Eurosystem credit operations.

The existence of a single pool of collateral for monetary policy operations, as well as for Eurosystem intraday credit operations, was a pre-condition for the emergence of a unified money market for the euro. This required a common definition of the assets eligible as collateral, safe and efficient domestic procedures, and the existence of an efficient mechanism to ensure the cross-border transfer of eligible assets. In 1996, when it became clear that the market would not be able to create such a mechanism, the EMI decided to create CCBM. CCBM is based on links between central banks for the exchange of information and messages on the securities used as collateral for their credit operations. The concept of CCBM is simple: a market participant can receive credit from its home NCB using collateral located in other Member States; the NCB of the country where the collateral is deposited acts as correspondent for the NCB that grants the credit.

CCBM has made the integration of European securities markets possible, allowing cross-border use of collateral in operations involving central banks. At the end of August central banks held an amount of EUR 74 billion on behalf of each other as a result of cross-border use of securities through CCBM. This figure stands for about 12% of the collateral used in Eurosystem credit operations.

19. The second activity in the securities settlement field, i.e. the definition of standards for the use of settlement systems in Eurosystem credit operations, was also a prerequisite for the smooth start of Monetary Union. The objective that all systems would apply the same high standards of safety and efficiency stemmed both from the necessity to ensure a level playing field among the counterparts of the Eurosystem and from the need to protect the Eurosystem against undue settlement or custody risks.

The Eurosystem standards have been published in January 1998. They reflect a strong preference for clear legal arrangements, the use of central bank money for settlement, strict limitations on custody risk, and intraday finality. It is worth noting that the standards, which were defined as standards required by the Eurosystem as a user, have de facto become supervisory standards. It is so because many systems have committed themselves to implementing the standards not only for operations related to central bank credit, but also for all kinds of operations. This evolution is similar to the one experienced long before in the field of banking supervision, when the counterparty assessment made by central banks to protect themselves against the possible failure of their creditor banks became the starting point of prudential rules.

In December 1998, after an assessment of domestic procedures, the Eurosystem started evaluating the quality of links between securities settlement systems, with the view of ensuring that such links offer the same reliability as the systems themselves. A first series of 26 links was authorised in June 1999.

20. Finally, retail payment systems. The Eurosystem devoted limited attention to this area during its first year of activity, because concentration on large-value payment systems was critical for the smooth transition to the euro. A report on electronic money, however, was adopted in August 1998. In that report, the Governing Council confirmed its position that only credit institutions should issue electronic money, in order to maintain a level playing field in the euro area. That view is not shared by all central banks worldwide. Compared to those of other countries, European banking laws are very liberal in allowing licensed banks to conduct a wide range of activities, but tend to be strict on requiring a license. As a result, reserving the right to issue electronic money to credit institutions is equivalent to inviting issuers of electronic money to adopt a certain corporate structure; it is not a way of creating a privilege for an existing group of institutions. The Eurosystem has also insisted on the need for electronic money to be redeemable in bank money or in central bank money at par. As monetary history shows, public confidence in the currency is linked to this requirement.

V. The main policy issues today

21. I now turn to the policy issues ahead. Those in which the Eurosystem and the ECB are likely to be involved in the coming months and beyond fall again under the three main headings of large-value payments, securities settlement systems, and retail payments.

22.On the large-value payments front, future work will primarily concern TARGET and is likely to concentrate on three issues: (i) technical improvements; (ii) the establishment of closer relations with the users; and (iii) the area of jurisdiction.

Technical improvements are necessary because in its first months of existence TARGET encountered some difficulties, mainly in the software, which interrupted the flow of cross-country payments on several occasions. Although these difficulties did not seem to damage the confidence of banks and became less frequent and less serious after the first two months of operations, their causes need to be addressed. The technical infrastructure of TARGET is composed of fifteen RTGS systems and fifteen Interlinking components, which have been developed under the responsibility of the national central banks. This heterogeneous nature of TARGET raises problems of both efficiency and cost. Efficiency problems are due to the fact that any system with many interfaces is likely to encounter more frequent difficulties than a centralised system or a decentralised one with identical local components. Cost problems are primarily due to the fact that each time a modification has to be made to the TARGET software, it has to be made fifteen times.

When TARGET was designed it was decided that harmonisation should be "minimal" (so called minimum harmonisation approach) essentially because it would have taken too long to seek agreement on the design of a single TARGET component that could have been used by all NCBs. Moreover, for many of them, this would have implied substantial losses, because their RTGS investments were recent (or even still underway) and far from being recovered. Finally the selection of countries participating in EMU was not done. As a matter of fact, when the Working Group I chaired agreed on the minimum approach in 1993 and 1994, the main reasons given were more practical than conceptual. Today, the same arguments relating to costs and efficiency which advised against substantial harmonisation in the preparatory phase of TARGET speak -in the operational phase of the system -in favour of this idea.

The minimalism of harmonisation may not be there for eternity. It will have to be reconsidered in the future and, over time, the various components of TARGET may converge towards a single model or, perhaps, even towards a unified infrastructure, as holds true of Fedwire and ELS, the German RTGS system, today. There has been as yet no special debate within the Eurosystem in this respect, but there is no doubt that the subject needs careful analysis. In any case, decisions about a unification in TARGET would, and should, not run counter to the principle of decentralisation. In the TARGET environment decentralisation consists primarily in the exclusive account relationship that NCBs maintain with commercial banks in their respective countries. That principle is, and will remain, one of the building blocks of the Eurosystem's architecture.

23. The Eurosystem-users relations will also need to improve. During the preparatory phase of TARGET, NCBs were intermediaries between the banks and the working parties in charge of designing the system. Although banks sometimes complained that they were not associated closely enough with the preparatory work for the system, this method proved efficient in allowing the key discussions to take place within the short time available for the design phase. TARGET would probably not have been ready on time if banks had been involved more directly in the discussions. Today, the situation is different. Banks need to re-discuss some of the existing TARGET rules in the light of experience, either to change them or to complement them. The Eurosystem will be able to accommodate such requests over time, while firmly preserving its key objectives in terms of safety and efficiency. Fruitful discussions between banks and central banks can now start at the euro level.

24. Finally, the area of jurisdiction of TARGET is an issue for the following practical considerations. Twelve countries in central and Eastern Europe have applied to join the European Union and, later, to adopt the euro, and others may follow. It is not certain whether TARGET could cope in its present shape with the need to operate perhaps as many as 25 connections. What is clear is that, once applicant countries join the euro, their banks will need to have equal access to TARGET. The enlargement of the European Union will therefore exacerbate the efficiency and cost problems already mentioned. Moreover, most of the applicant countries are likely to process a low volume of payments and, if they have to build their connections to TARGET the way this has been done so far, they would probably never recover their costs. My conclusion is that the geographical extension of TARGET will require innovative solutions for the system as a whole or, at least, for the applicant countries.

25. I would now like to briefly consider three other issues closely related to, but not coincident with, large value payment problems. The first concerns the six other payment systems that continue to process large-value payments in the euro area in parallel with TARGET.

The multiplicity of large value payment systems is exceptional compared with that of other monetary areas where both market forces and policy makers led to a single system (United Kingdom) or two systems (United States). It is possible that the present infrastructure in the euro area, which requires banks to divide their liquidity between several systems, will consolidate further. However, the large-volume systems, which operate in parallel with TARGET, are very sound and, so far, have not caused segmentation of the money market. For these reasons, the Eurosystem has no reservation against the fact that banks find the use of systems other than TARGET more convenient for some categories of payments, commercial payments in particular, for which immediate finality is not a compelling necessity.

The only aspect on which the Eurosystem will continue to express reservations relates to the use of systems other than TARGET for the processing of very high value payments, e.g. more than EUR 500 million or even 1 billion. Processing such payments in systems other than TARGET gives rise to concerns about efficiency (liquidity limits lower than in TARGET reduce the flow of transactions) and safety (payments should be final as soon as possible).

26. The second issue is the inauguration of the CLS Bank. For those who are not familiar with this development, let me give a little background. CLS stands for Continuous Linked Settlement. The CLS Bank will manage a settlement arrangement in which foreign exchange contracts are settled on a gross, "payment-versus-payment", basis in order to eliminate the so called "Herstatt risk", i.e. the risk of a loss arising from the non delivery of one of the two legs of a two-currency transaction. For several years now, central banks had been pressing banks to find a solution to eliminate that risk, which has obvious systemic implications. In response, a group of major international banks has developed CLS. If successful (and I have no reason to believe that it could not be successful), this project will represent a major achievement for the banking sector and an example of good co-operation between banks and central banks. The Eurosystem strongly supports the launch of CLS. It is also happy to see that the euro is expected to be among the first group of eligible currencies to be settled in the CLS Bank.

27. The third issue concerns developments in the field of correspondent banking. In principle, correspondent banking, like CLS, is also a mechanism that settles in commercial bank money because it is based on reciprocal services provided between banks operating in different countries. It has hitherto been limited in scope and based on bilateral credit assessments. Few systemic implications were involved. Now, however, correspondent banking is evolving and becoming a special type of service. Major banks may now have hundreds of customers, and fund transfers can be settled directly on their books, by transfers from one account to the other. In the end, this development could lead to new kinds of fund transfer systems, with settlement taking place in commercial banks' money. It is easily understandable that central banks, including the Eurosystem, will wish to evaluate the safety features of this evolution.

28. Let me now turn to securities settlement systems. In preparing for the start of Monetary Union, the Eurosystem had adopted a minimum approach for reasons of simplicity. This approach was based on the selection of those eligible securities settlement systems, which met the minimum standards mentioned above. NCBs have maintained securities accounts with their local systems (and, a few of them, with Cedel and Euroclear) for the delivery of local securities. For the delivery of foreign collateral (i.e. collateral located outside their jurisdiction), they have developed CCBM. While it was sufficient to launch Monetary Union, such model may not remain the most efficient mechanism in the long run. Alternative solutions such as the use of direct links between securities settlement systems have been envisaged and in some cases are already being used. Moreover, the possibility for NBCs to use accounts with central securities depositories located in other countries of the euro area, so-called remote access, is being considered. Let me briefly discuss these two alternatives.

29. The possibility of using direct links between securities settlement systems, whereby one securities settlement system becomes a participant in another, has already been approved by the Eurosystem, at the condition its minimum standards are met. To date, as mentioned earlier, 26 such links have been approved and can be used for the collateralisation of the central bank's credit operations. About 20 more direct links are likely to be approved in the coming months.

The next step will be to investigate under which conditions the Eurosystem could accept collateral delivered through the so-called indirect links, whereby two securities settlement systems use one or more custodian bank(s) or another securities settlement system as intermediary(ies). At the same time, the Eurosystem is investigating whether securities held in the books of custodian banks could also be used as collateral in its credit operations. Answers to all these questions will no doubt have an impact on the market. It is, therefore, essential that the Eurosystem takes positions that are exclusively based on risk and efficiency considerations.

30. Remote access is also being studied. In the present context, remote access refers to the possibility for NCBs of using accounts with securities settlement systems located in countries other than their own for settlement of their monetary policy or intraday credit operations. If such facility has not been used so far, it is because it was not deemed necessary within the framework of the minimum harmonisation approach, not because of safety or efficiency considerations. These aspects need now to be analysed in greater depth. The results of this analysis are expected to become available by the end of this year.

31. The consolidation of securities settlement systems has yet to come. Interbank cross-border transfers of securities in the euro area still rely extensively on the use of custodian banks. Custodian banks are to securities what correspondent banks are to payments: intermediaries that provide access facilities to a foreign settlement system. But, contrary to what has happened on the payments side with the creation of TARGET, it has been difficult for banks to reduce the costs linked to the use of custodian banks in the euro area because the securities settlement system industry is still very fragmented. Moreover, whereas the introduction of the single currency has eliminated the multiplicity of currencies within the Euroland, thereby reducing the need for traditional correspondent banking services, differences between national issuers still remain thereby increasing the fragmentation of the securities industry. A consequence is that, even today, it is not easier for a euro area bank to settle trades in "bunds" than it is to settle trades in "gilts", although the first are denominated in euro and the second are not. It is likely that the process of integration of the securities industry will affect also the activities of bank custodians. These activities tend to concentrate more and more in a very small number of big operators.

32. The Correspondent Central Banking Model, the use of links between securities settlement systems and remote access will not be sufficient. Market participants want a further consolidation of the securities settlement system industry of the euro area in order to save costs. They have already made it clear that they want to be able to settle all their securities transactions in euro from a single gateway. This is not very surprising. Securities settlement systems clearly represent what economic theory would call a natural monopoly and the euro area will see the same trend towards consolidation that has already been seen in individual countries.

Several projects have been developed so far to further consolidate the securities settlement industry. It is not yet clear which single solution will emerge and how long it will take for it to emerge. The segmentation in different systems has existed for so many years that a great deal of resistance can be expected. For its part, the Eurosystem will not oppose the trend towards the consolidation of central securities depositories in Europe that the market requires. It will remain neutral with respect to the competing formulas for consolidation and will leave it to market forces to decide the lines along which the restructuring should take place. Neutrality of the Eurosystem does not consist in giving "a bit to everyone". It means, instead, defining clearly and openly the standards the central bank of the euro deems necessary. Neutrality of course requires that these standards are not modified to accommodate individual requests of some market participants. Once the market decides which model (or models) prevails, the Eurosystem will be willing to discuss with the operators how best to deal with the risk issues involved.

33. The third category of policy issues in which the Eurosystem is involved concerns the area of retail payments. I have already indicated that, until recently and with the exception of electronic money, the Eurosystem had not paid much attention to this area, notwithstanding the responsibilities defined by the Treaty. This attitude had to change because substantial improvements can no longer be delayed in the field of retail cross-border payments in the euro area. In a recently published report, the ECB Governing Council has set a number of objectives to be pursued by the banking community in order to improve the speed and cost conditions of retail cross-border payments. The report also explains why the current low level of efficiency of retail cross-border payments in the euro area is indeed a source of concern for the ECB, identifies some causes of the problem and defines the Eurosystem's strategy and objectives. Let me elaborate a little bit more on this.

The current situation in the field of retail cross-border payments is not satisfactory. Although exchange rate risks have disappeared within the euro area, charges for such payments have not, as had been expected by consumers, decreased substantially. Whereas the fees charged to customers for domestic credit transfers rarely exceed between EUR 0.10 and EUR 0.20, the fees for most cross-border credit transfers come to between EUR 3.50 and EUR 26, with even higher peaks in some countries. The speed of execution for domestic transfers is sometimes less than a day, whereas transfers on a cross-border basis typically take several days, occasionally even more than a week. The full benefits of Monetary Union will not be reaped by individuals and small enterprises until a payment between Milan and Munich is as efficient, secure and quick as a payment between Munich and Frankfurt. The euro area still has to become a single payment area.

34. Several initiatives, including legislation, were taken with a view to abolishing the barriers to integrated payment systems within the Single Market. In particular, studies conducted by the EU Commission in 1993 and 1994 led to the adoption, in 1997, of the EU Parliament and Council Directive on cross-border credit transfers, which lays down common rules in the area of transparency and performance to the benefit of both end-customers and financial institutions. EU Member States were due to implement this Directive by August 1999 at the latest. Important improvements in banks' practices should result from the implementation of the Directive. However, since it applies to the whole of the European Union and is inspired by considerations related to the Single Market, that Directive could only address aspects that are specific to a multi-currency environment. Customers within the euro area expect that greater benefits from the cross-border payment services become available to them. This expectation is a concern of the central bank of the euro, i.e. the Eurosystem.

The Eurosystem has identified several reasons, which explain why the gap between consumers' expectations and what the market is offering is still so wide in the area of cross-border retail payments. They include the predominant use of correspondent banking and the lack of standardisation. The gap, however, needs to shrink quickly to a more reasonable level. If it does not, confidence of citizens and businesses in Monetary Union will suffer. For its part, the Eurosystem has come to the conclusion that, at least for the moment, it should work as a catalyst for change rather than increasing its operational involvement. This explains the publication of a series of objectives that the banking community is invited to meet by 2002 at the latest, when all the day-to-day transactions of the citizens of the euro area will be denominated in euro.

VI. Conclusion

35. I have looked at the process of restructuring of payment and securities settlements systems in Europe from a central banker's perspective. This process has not only been driven by technology and market forces, it has also been a response to the establishment of the Single Market and Monetary Union. These two events have set in motion a fundamental transformation, that is far from complete, of the monetary and financial sector of Euroland. In the new environment, old national infrastructures could not remain; they need restructuring and consolidation. The Eurosystem welcomes a process that will help reach all benefits that can be expected from financial and monetary integration in Europe.

Looking at the PSSS industry, the greatest advances have been made in the field of large-value payment systems, where many reforms have been achieved. National central banks and, since it was established, the Eurosystem have made their contribution to these reforms, in particular, through the creation of TARGET. The trend towards consolidation in the securities settlement industry is a more recent phenomenon. In this area, market forces still need to find and develop those arrangements that meet the requirements of concentration and specialisation. As long as market forces do not lead to unsound or inefficient solutions, the Eurosystem will not influence the trend towards consolidation of the central securities depositories in Europe.

In the field of retail payments, substantial efforts are required to make cross-border payments as efficient, secure and quick as domestic payments. For the European citizens and businesses, the euro area still has to become a single payment area. The Eurosystem has recently set targets for the banking industry that may help it reach that objective within a reasonable period of time.

The responsibility of the Eurosystem is to ensure that the vast restructuring process of the industry takes place in a way that preserves the safety and the smooth functioning of payment and securities settlement systems and increases their efficiency. The future accession of central and eastern countries to the EU and, later, Monetary Union will make the current process even more complex. Within its field of responsibility, the Eurosystem stands ready to address this challenge.

Media contacts