Thank you for giving me the opportunity to present the Eurosystem's view on the current state of affairs in European cross-border retail payments.
As you are aware, today's date is an important one in European history. Exactly eleven years ago, East German citizens tore down the Berlin Wall and paved the way for German unification. If I can draw, from this big historical event, an analogy for our small world of payment arrangements, I would like to express my sincere wish that today's meeting will help tearing down the walls impeding full integration of retail payments in the euro area. The Eurosystem is strongly committed to achieving this objective very soon. Indeed, 2002 should be the landmark in this process.
In my remarks today, I will first recall the central bank role in retail payments. Next, I will explain what our common objective for 2002 should be and evaluate what has been done so far. Before concluding, I shall also comment on the instruments that the Eurosystem has at its disposal to perform its payment system functions and on its oversight role.
Central banks have been created to operate payment systems. At their very origin, they were entrusted with the monopoly for the issuance of banknotes, as paper currency previously issued by commercial banks or non-bank enterprises proved unsatisfactory or even unreliable. Central banks were then called "banks of issue". Later, they became "central banks" because of their role as "bank of the banks". Their functions in the fields of banking supervision and monetary policy followed the payment systems one.
In recent years, the attention of central banks has focused on systemic risk and on the oversight of large value payment systems, with the adoption of the Lamfalussy report and the development and use of RTGS. As a result, the view sometimes prevails, even among central banks, that their role in payment systems is now played more (or even exclusively) in the field of large value systems, rather than in retail payments.
I do not share this view. Retail payments are in fact the heart of central banking. Central banks would betray their origins and their social function if they neglected that field. In Europe, this view is clearly supported by the Treaty. Indeed, Articles 3 and 22 of the Statute of the ESCB neither limit the tasks of the Eurosystem to risk considerations nor do they limit its possible intervention to large-value payment systems. The Statute explicitly allows for an operational involvement of the Eurosystem that is based on efficiency grounds, and makes no distinction between different categories of payments.
One can also refer to the US approach. A key provision of the 1913 Federal Reserve Act (one without which US Congress would perhaps have never created a central bank) states that all checks should be cleared nation-wide under the same conditions. More recently, in 1998, a committee headed by Alice Rivlin the Vice Chair of the Board of Governors of the Federal Reserve System, analysed the role of the Federal Reserve System in retail payment services. The committee liased with market participants and evaluated several options for the Fed's involvement in retail payments, ranging from complete withdrawal to taking a leading role. The report confirmed the operational role of the Federal Reserve in retail payment systems and strongly recommended that it plays a leading role towards enhancing the efficiency, effectiveness and convenience of retail payment services, in co-operation with market participants.
Why have central banks been chosen for such leading role? First, because of the experience they have accumulated in the field, which comes from their history of providing payments instruments (and not only banknotes) to the private sector. Second, because central banks are "banks of the banks", one should even say "banks for the banks". For those reasons, they are the public authorities that have the operational knowledge and the practical experience required to exercise the public functions involved in the payment system.
At the beginning of the 20th almost the totality of retail payments were made in central bank money. Over time, this monopoly came to be shared with commercial banks, when deposits and their transfer via checks and giros became widely accepted. Banknotes and commercial bank money became fully interchangeable payment media that customers could use according to their needs. While transaction costs in commercial bank money were shrinking, cashless payment instruments became increasingly used, at the expense of banknotes.
When looking at the 2002 horizon, however, we have to recognise a dramatic discrepancy. On their side, central banks are producing an enormous effort to prepare the cash changeover (14 billion banknotes and 60 billion coins will have to be replaced, which represents 170 square kilometres of paper to be printed and 270,000 tons of metal to be minted). That effort will make sure that, within the first few weeks of 2002, the currency borders for banknotes in the euro area will disappear. The banknotes will be identical and freely circulating in any part of the area without costs.
For central bank money, the distinction between domestic and cross-border payments will disappear in early 2002. Indeed, the very notion of "cross-border" payment will then lose any meaning within the euro area. All payments in Euroland will be domestic.
Now, and this is the key point I want to make today, it is absolutely crucial that borders will cease to exist also for commercial bank money. This means that, after 2002, indifference between commercial bank money and central bank money will have to apply to the whole of Euroland. Only by achieving this goal would the banking system continue to fulfil the responsibility it has taken up more than a century ago when it began replacing central bank money with commercial bank money of its own creation.
It is true that, when a retail payment involves entities based in two different countries, the cost of processing it is today higher than when the entities are in the same Country. This is why payments between Lyon and Torino remain less efficient than payments between Torino and Palermo. In early 2002, however, the distinction between domestic and cross border payments will have to disappear.
The Eurosystem is very concerned that this might not be the case for commercial bank money in practice. This concern is based on two reasons. First, because the Eurosystem has the mandate to promote the smooth functioning of payment systems. Second, because, if there remains a distinction between domestic and cross-border payments in the euro area after 2002, this will disappoint the expectations of European citizens. The Eurosystem cannot be indifferent to such development. Nor should the banks. As you know, non-bank institutions have already started to lobby EU institutions in view of obtaining an "EU passport", using the deficiencies of the banks in the retail field as an argument.
It is true that the matter is not an easy one. Payment systems depend on infrastructures which have been set up at domestic level and which are difficult to interconnect. Changing them requires high costs and long lead times. Unfortunately, banking systems have done little in this field until the start of 1999, when the Eurosystem decided to tackle this issue.
The Eurosystem approached the problem in a pragmatic way. We first listened to the banks. They gave us a number of arguments explaining why retail cross border payments are more slow and costly: there is no business case; the infrastructures are inadequate; no common standards are used; double charging cannot be avoided; and reporting requirements for balance of payments are expensive. Let me briefly review these various arguments one by one.
We were not convinced by the business case argument. Is there a business case for a payment between a village in Aquitaine and one in Alsace having the same cost and speed as one between Paris and Lyon? Are prices high because traffic is low, or is traffic low because prices are high? Even if there were no "business case", in a narrow sense, there is a very strong one in a broader sense. In fact, banks understood this, but in 1999 only.
In the field of the infrastructures, the banking sector has been very active. In particular, they have decided to set up a new system called STEP1, which will have an extensive coverage within the banking sector. The ECB has welcomed this initiative and has assessed the system positively.
On the standards front, when the Eurosystem started to be heavily involved in 1999, most of the standards needed were already available. The two key standards have been the International Bank Account Number (IBAN) and the International Payment Instruction (IPI). Only one element was missing for complete straight through processing: a special SWIFT message for the interbank part of the transactions. The ECB launched the discussion and participated in the inter-bank work which have led to the definition of the missing link (so called "chaînon manquant"): in the jargon of technicians this is the MT 103+ message. One well known problem remains: the implementation date for all these standards. Banks tend to delay the implementation date of the standards because of the costs involved and because there is no incentive to be among the first to implement standards, since these are only beneficial when they are widely and commonly used. So we did what all central banks would do in such circumstances: we urged banks to speed up the process and to publicly commit themselves on an early implementation date.
Double charging is obviously unacceptable and today even illegal in some cases (according to the 1997 Directive). In the postal service, double charging disappeared about two centuries ago. One solution to avoid double charging could have been to decide that only the sending bank charges a payment fee, and that the money collected on the sending side would also cover expenses linked to incoming payments. This scheme has often been adopted at the domestic level. However, within the euro area there are large imbalances between the North (which is net sender) and the South (which is net receiver). The absence of a mechanism to remunerate the receiving bank would clearly be detrimental for banks which are active in Southern Europe. Of course, banks could have found a solution by themselves before the 1997 directive was implemented. But they did not. In 1999, the Eurosystem acted, as other central banks would have done. We have urged the banks to define a short-term solution and they have proposed to set up a Multilateral Interbank exchange Fee (MIF). We have evaluated this arrangement against other alternatives and discussed the matter with the Commission. Our conclusion is that the MIF represents a solution, at least in the short run, provided the fee is low.
Reporting requirements for balance of payments purposes have been the argument traditionally used by banks to return the problem of high charges to the public authorities. The dilemma for public authorities, including central banks, is that they want both good statistics and good payments. The ECB has promoted bilateral discussions, first, between statisticians and Eurosystem payment systems experts and, later, between statisticians and banks. Eventually, the statistical authorities have decided to exempt all payments below EUR 12,500 from any reporting requirements. Many banks would have preferred a higher threshold. One should, however, be realistic.
This brief review suggests that all the conditions are now set for banks to be able to fulfil the criterion of indifference, i.e. to ensure equal cost and speed of commercial bank money transfers throughout Euroland. Of course, the positive achievements mentioned before are recent and their impact will take time to materialise. The central bank can understand this argument, but is convinced that 2002 should be set as the key year just as it is the key year for banknotes and coins.
Let us now turn to the instruments available to the Eurosystem to carry out its payment system functions. When central banks promote reforms of payment systems, their preferred strategy is to reach common agreements through co-operation. In all countries, the central bank and the local banking community have found a way to organise discussions and find a consensus on most payment systems issues. ZKA (Zentraler Kreditausschuss) in Germany, APACS (Association for Payment Clearing Services) in the UK and CFONB (Comité Français d'Organisation et de Normalisation Bancaire) are three examples. These bodies are not totally comparable in status and activities, but they are all fora where banks and central banks can co-operate.
The Eurosystem has still to firmly establish such a forum, although it is increasingly discussing payment systems issues with the banking community of the euro area. Regular meetings are being organised with banks on various issues, like retail payments, TARGET, or securities settlement.
Central banks may also have recourse to two other instruments: operational involvement, and regulation, either directly, when the central bank has regulatory power, or indirectly, by making proposals for legislation.
Operational involvement is an instrument available to most central banks, including the Eurosystem. TARGET has been created to contain systemic risk through the continuous use of central bank money. In the retail field, however, there may be no such need, because the banking sector itself should be able to process retail payments sufficiently. Only if banks fail to deliver efficient and cheap payment services throughout the euro area, would the Eurosystem be compelled to set up a "TARGET for the citizen". This idea deserves careful examination.
If the issue were a lack of interbank infrastructure, TARGET could easily be used to develop a solution for retail payments. Today, however, efficient solutions for processing payments between banks, like the STEP1 system, exist or are under preparation. The problem we are facing now is rather one of relations between banks and their customers. In this field, the operational involvement of the Eurosystem would become more complex because central banks do not have a network of ordinary customers. That said, it remains that, if banks fail to deliver what the society expects from them in the early days or weeks of 2002, a Eurosystem solution cannot be excluded.
Concerning regulation, this instrument has explicitly been granted to the Eurosystem by the Treaty (Article 22 of the Statute). While this tool has not been used so far, it should, by no means, be forgotten. Regulation could be used as an outcome of co-operation, with a view to formalising an agreement, or as a substitute for co-operation, when the latter failed.
In practice, however, central banks are reluctant to substitute regulation to negotiation and moral persuasion. Regulation, indeed, is rarely a good instrument to solve efficiency problems. This is shown, for example, by the limited effect the 1997 Directive has had on prices for retail cross border payments. Regulation may also prevent the market from finding appropriate outcomes. One example of this is the law of cheques in France which dates from the 1960s and which has clearly led to an excessive use of cheques that is now difficult to modify.
The risk of over-regulation also exists at the European level. For instance, it has been argued that the 1997 directive on cross-border transfers should be revised because it imposes a six-day delay for processing payments, while most banks can now process payments in three days. The fact that most banks do better than required should be applauded. When technology and market forces work, a new regulation in this field would serve no purpose. It could even be counterproductive.
Another instance of potential over-regulation is suggested by the idea to prepare a directive on a European direct debit scheme. The creation of a European direct debit is indeed an objective, but one may doubt whether it is appropriate to start with a legal instrument. The direct debit has been introduced in the past twenty years in most EU countries. The process never started by a regulation, but by a study made by the banking sector, in general with the strong involvement of the central bank. These studies generally concluded that some rules had to be changed for the schemes to be effective. Legislation then took place, only when necessary.
Let us now move to the Eurosystem oversight. Oversight is the expression used to describe the function of central banks aiming at the efficiency and soundness of payment systems. The main objectives pursued are the maintenance of systemic stability, the promotion of efficiency and the safeguarding of the transmission channel for monetary policy.
Payment systems oversight shares with prudential supervision the objective of financial stability. However, while prudential supervision looks at institutions, payment systems oversight looks at systems. The word "system" could refer either to a particular funds transfer system, or to the general organisation of payment services in the economy. For my remarks today, this second aspect is more relevant.
Oversight may or may not be based on explicit legal provisions. Central banks developed oversight frameworks in the 1980s, mostly on a non-statutory basis. Later, however, their oversight role gradually became recognised in law. For the Eurosystem, it is embedded in the Treaty.
In June 2000, the ECB published a document entitled "Role of the Eurosystem in the field of payment systems oversight". This document explains the set up of the payment systems oversight function in the decentralised environment of the Eurosystem. It also deals with retail payments.
The decentralisation principle means that, to the extent possible, the tasks of the Eurosystem are carried out through the national central banks. In the area of payment systems, this means in practice that the Governing Council of the ECB formulates the common oversight policy stance in those cases where the functioning of the systems may affect the implementation of monetary policy, systemic stability, the establishment of a level playing field between market participants, or cross-border payments within the EU or with other countries. In other areas, NCBs apply specific policies tailored to their national environment.
Decentralisation is sometimes, but inappropriately, referred to as a case of application of the principle of subsidiarity. The principle of subsidiarity leaves to the national level competencies for which there is no advantage to exercise them at the European level. Decentralisation is a modality to carry on a European, or euro area, competence. In the field of payment systems, a competence has clearly been entrusted to the Eurosystem. In that field, subsidiarity does not apply. Decentralisation does.
It is time to conclude. The improvement in retail cross-border payment services depends on infrastructures and procedures more than on regulation. Banks have done too little until 1999 when they "discovered" the political problem they will face in 2002. At this point they started moving.
In early 2002, the moment of truth will come. Banks in the euro area will face a critical decision. Indeed, they will have to decide whether they might change their pricing policies in the field of cross-border payments or instead face unpopularity. This is a difficult choice. In making it, banks may want to consider the costs of the promotion campaigns they may have to conduct to redress their image. This is a kind of cost that payment systems experts may not be aware of. But the banks' CEOs are.
The conclusion is that the royal way may be for banks to price domestic and cross-border payments at a similar level, and make all efforts possible to allow for the costs also to become the same. The challenge is to prepare the 2002 changeover in a way which will ensure that central bank money and commercial bank money are offered to customers under the same conditions, so that the principle of indifference between the two is respected.
Reproduction is permitted provided that the source is acknowledged.