How to introduce the Ecu: an analytical framework

S. van Tuyll van Serooskerken (*)

According to the Maastricht Treaty on European Union, the third stage of monetary union will start at the latest on 1 January 1999. From the first day of the third stage, the exchange rates of the partici-pating currencies will be irrevocably fixed and monetary policy will be conducted by the European Central Bank. However, this does not necessar-ily imply that on that day the Ecu will be introduced as the single currency. In this article, some of the technical aspects of the introduction of the Ecu will be discussed.

The Treaty says that at the beginning of stage three the Council shall take the measures necessary for the rapid introduction of the Ecu as the single currency. Until now, however, no decision has been taken as to when and how the Ecu will be intro-duced. The Commis-sion is preparing a Green Paper on this subject, which will be discussed by the European Council in Cannes next June.

In theory, the Ecu could be introduced in several ways. One extreme, the so called 'Immediate big bang', would be to have the Ecu replace all the currencies of participat-ing member states on the first day of stage three. The other extreme would be a gradual introduc-tion of the Ecu, whereby national currencies and the Ecu coexist for several years.

An 'Immediate Big Bang' -whereby all currencies of participating member states are replaced by the Ecu from the first day of stage three- seems to be neither feasible nor desir-able. It is not feasible because prepara-tions for the introduc-tion of the single currency will take more time than the 1½ years up to 1997 or the 3½ years up to 1999. This time is not only needed to print the bank notes and mint the coins, but also to make sure other measures required by the change-over are taken. A large part of these measures will have to be taken by the private sec-tor; they will involve some costs, which can be seen as an investment, that will pay off by efficiency gains from the single currency. These prepara-tion measures are likely to be taken by the private sector only once it is clear which countries will participate in stage three.

Even if the Ecu is not immediately introduced, the beginning of the monetary union will imply some significant changes in the economic environment. The European Central Bank will be in charge of a single monetary policy. Exchange rate risks will have disap-peared, thus eliminating the exchange rate risk premium in interest rates. Any remaining interest rate differential will stem from differences in credit risk and liquidity. Moreover, the deeper capital market should lead to lower interest rates, even in those countries that now already have the lowest interest rates within the ERM. The introduction of the single currency will enhance this effect by further deepening capital markets.

As long as the Ecu has not been introduced, the economies will still be operating in national currencies, be it that the exchange rates have been irrevocably fixed. Two mechan-isms will ensure that these exchange rates remain stable.

One of them follows from article 52 of the ECB Statute, that provides that the Govern-ing Council of the ECB shall take the necessary measures to ensure that banknotes denomi-nated in currencies with irrevocably fixed exchange rates are exchanged by the national central banks at their respective par values. The ECB may even decide that bank balances will be exchanged at par value. This implies that any deviation from par value that may exist -or any deviation in cross rates- will be arbitraged away. Of course no provision exists (1) that compels market participants to trade at exactly the par rates, but market forces will make sure that any deviation from par value only reflects the transaction costs.

The second mechanism to ensure exchange rate stability will be the fact that banks will not have to match their currency balances anymore. Whereas now currency transactions are being matched out of prudential considerations, this will be unnecessary once the currencies have been irrevocably fixed. The purchase of DM by a client of a Belgian bank will be reduced to merely a book transac-tion, that does not need to be matched on the currency market. The bank may cover the administrative costs of such a transaction by charging a fee or by using a small spread(2).

The other extreme way to introduce the Ecu would be a gradual introduction of Ecu notes and coins. The Ecu and national currencies would coexist for several years, with the Ecu gradually replacing the national currencies. Such an option, however, is incon-ven-ient for virtually the entire econ-omy. Any situation in which two currencies coexist as legal tender will require dual pricing, dual cash handling, dual accounting, etc. and will lead to large additional costs. A gradual introduction of the Ecu therefore seems to be inappropri-ate and any parallel circulation of currencies should be avoided.

Having eliminated the immediate replacement of national currencies by the Ecu on the first day of stage three and a gradual introduction of the Ecu, the question is what could be the optimal approach to the introduction of the Ecu. It is important therefore to analyze how financial transactions will be affected between the first day of stage three (called 'T-day') and the full introduction of the Ecu as the single currency (called 'E-day').

Whatever scenario is chosen for the introduction of the Ecu, there will be a need to redenominate all accounts, stocks, and long term contracts such as bonds, rental or labour contracts, etc. at some stage. One-off costs for this operation are inevitable. How-ever, transfers need be converted whenever one of the parties con-cerned is operating in Ecu and the other in national currencies. For this conversion, some conver-sion mechan-ism - interface- has to be built in, a mechanism that by definition becomes obsolete after E-day.

At what level this conversion should take place, can be analyzed by reference to the financial pyramid shown in the diagram. The top of the pyramid, the European Central Bank, will have the Ecu as unit of account. By definition, before E-day the lower tier of the pyramid (the 'end-users of money') will be using national currencies. The main question is where to put the boundary between the Ecu and national cur-rencies in the period from T-day to E-day.

When the boundary between the Ecu and national currencies is set at a high level -for example between the ECB and the National Central Banks-, only a few, but large, transactions will have to be converted from Ecu to national currencies and vice versa. On the other hand, when conversion takes place at the lowest level of the banking system -for instance when banks are operating in Ecu but the end-users of money are still using national currencies-, it involves many transactions, small as well as large. It will require the conver-sion of many payment systems within the banking sector: transfers, cheques, credit cards, cash transac-tions, etc. Because more transac-tions and more systems are involved, mistakes and fraud are also more likely to occur: any conver-sion mechan-ism will therefore have to be thoroughly tested. One can expect that the costs of building many different conversion mechanisms at a low level of the banking system will be higher than having the conver-sion at the upper level.

Clearly, if the boundary between the Ecu and national currencies is moving, it will require conversion mechanisms at different levels, which seems to be even more complex and expensive.

This argument therefore seems to point in favour of a 'Delayed Big Bang' (3), in which between T-day and E-day the conver-sion takes place at a high level between the ECB and national Central Banks or between national Central Banks and commercial banks when monetary policy is conducted in Ecu. On E-day the rest of the economy could change over in one go: notes, coins, accounts, long-term contracts could be changed over simultaneous-ly. In such a scenario payment systems and the financial accounting systems need hardly be adapted. After E-day the national currency would lose legal tender status and would be treated as a foreign currency, just like the Ecu before E-day. Though the logistics for replacing notes and coins will require a major effort, this scenario seems to be less costly for the banking system and for the economy. Of course, if it turns out that banks can easily build conversion mechanisms at a lower level of the financial pyramid without too much cost, a scenario with conver-sion at a lower level should not be excluded. But this would be a choice for the banks.

In order to assess the feasibility of a Delayed Big Bang, one should have a closer look at the bottlenecks and find out if there are options in which these bottlenecks would be reduced.

One of them is the logistic for introducing notes and coins. It will of course not be possible to replace all notes and coins within one day. Some weeks before E-day, let us say three, Ecu notes and coins will be available at commercial banks. The public may want to take up Ecu cash at their bank or at some ATM in advance, just like one is used to do before going on holiday. Similarly, it will be possible to change national currency notes and coins into Ecu for a long period after E-day. The main logistic problem concerns the banks, who will have to distribute the notes and coins and convert their ATM's. It is, however, unlikely that a logistically easier option can be found. At first sight, a gradual introduction of notes and coins may be easier from this point of view. It is likely, however, that in a gradual introduction, the national currencies remain the dominant means of payment until the very last moment. As its timing becomes more uncertain, the logistics of the changeover would therefore be even more complicated.

Another aspect is psychological. All economic agents will have to adapt to new prices. Will it be easier to do so when there is a period of parallel circulation? A parallel circulation of currencies will not necessarily facilitate the adaptation process. It is likely that in the months before and after E-day many firms will quote two prices (4) for convenience of their customers. This may be useful, especially if firms adopt a clear typo-graphic distinction, such as Ecu in small typeface before E-day and vice versa afterwards. The adapta-tion to new prices will, however, not necessarily be facilitated by actually using two currencies as a means of payment. It may, on the contrary even be misleading.

It is clear that vending machines cannot be adapted overnight. This will require several months. In the weeks up to E-day some vending machines can be adapted already, in order to have some of them operating in Ecu on E-day. The remainder of them will have to change over after E-day. For this purpose national currency notes and coins may remain in circulation -and be distributed by kiosks, shops, etc.-, but without being legal tender. The coins will have the same status as, for instance, telephone tokens have in some countries.

Though the changeover in a Delayed Big Bang will not be an easy one, many of the problems involved have to be dealt with anyway: they cannot be avoided in other options.

A parallel may be drawn with the way Sweden switched in the 1970s from driving on the left side of the road to driving on the right. For obvious reasons this was done overnight. Technically speaking, it would have been possible for instance to drive right on the highways and left on the other roads, provided some kind of interface had been built at every exit. It is doubtful, to put it mildly, however, if it would have made the change easier and less costly.

In order to achieve a smooth and well accepted introduction of the Ecu, it is essential that the change over is well prepared, taking into account all kind of contingencies. The timing has to be adapted to those econ-omic agents that -within reasonable limits- need most time to prepare, which may take up to several years. Of course, if economic agents see some comparative advantage in using the Ecu before E-day, they will be free to do so.

For some, who strongly favour the single currency, any delay before a significant use of the Ecu may seem disappointing. This is what I would call the Ecu introduction paradox: a well prepared changeover in one go after a certain delay may well lead to an earlier, a more smoothly and better accepted introduction of the Ecu than when several sectors move at different speeds.

It is essential, however, that soon a clear view is given on how the Ecu will be introduced. When such a decision is taken, it will be important for the adaptation costs and the incon-venience for the public to be minimised, so as to yield the best chances of realising a smooth and quickly accepted introduction of the Ecu.

A clear timetable for the introduction of the Ecu, fixed when the decision to enter stage III is being taken, will be the best way to ensure that the momentum to achieve the single currency will be maintained.


(*) The author is Economic Advisor at DG II of the European Commission. The views expressed in this article are made in a personal capacity and do not necessarily reflect the opinion of the European Commission. Rue de la Loi, 200 - 1049 Bruxelles.

(1) If such a provision existed, it could easily be evaded by trading via a third currency: DM/$ and $/BFr.

(2) this spread will of course be a fraction of today's spread, which mainly reflects the exchange rate risk

(3) in French: 'un basculement complet préannoncé'

(4) this should be done (compuslorarily) at the fixed parity, in order to avoid deception.