In an attempt to picture what will happen at the start of the third stage, we could try to predict the future. In particular, we could try to imagine what will happen on Monday 4th January 1999, the first trading day of Monetary Union.
The morning will see the European Central Bank launch its first market operation, probably a fixed-term open market operation (a repurchase agreement). Early in the morning the terms of this operation, especially the volume and the rate, will be publicised, inter alia on the Reuters network. The financial institutions wishing to participate in this operation will send their bids to their respective national central banks within the pre-arranged deadline. The national central banks will collect the bids, check the quality of the underlying securities of the various counterparties, and pass on all the valid bids received from the institutions resident in their country to the European Central Bank.
The European Central Bank will collect the bids passed on by the NCBs, and will allocate the funds at the rates it has set to the counterparties whose bids have been accepted. The accounts of these counterparties held at the NCBs will be credited, in euros, with the amounts allocated.
Within the space of a few hours, this is very likely to be how the ESCB's first open market operation will be carried out, in the same way as it happens today at any central bank.
This first operation marks the moment separating two quite distinct periods: the first, which ends with this particular operation, represents the preparatory stage of the single monetary policy in which the EMI is currently involved, and in which the ECB will be involved from 1998 onwards; the second stage commences with this operation and represents the start of a new monetary regime in Europe.
Let us first look at the preparatory stage. We can do this in reverse, so to speak, starting with Monday 4th January, the first working day of 1999. During the previous weekend, on 2nd and 3rd January, the Governing Council of the ECB will have met to take a series of decisions. Two days beforehand, the conversion rates between the currencies making up the single currency area and the euro will have been irrevocably fixed.
First, it will need to decide on the terms under which Monday morning's repo operation will be carried out; such as its maturity and interest rate. It will also need to define the conditions under which the other regular monetary policy instruments will be used, particularly standing facilities. For instance, it will decide on the rate for the marginal lending facility (lombard facility), which enables counterparties to obtain credit from the ESCB at the end of the day; and the rate for the deposit facility, which allows ESCB counterparties to place deposits with the System. Together, these rates define, respectively, the ceiling and the floor of the corridor for the repo rate, within which will fluctuate the overnight money market rate, which will be the operational objective of the ESCB's monetary policy. These decisions will be taken on the basis of a range of information available to the Governing Council of the ECB, in particular inflationary pressures, the development of the monetary aggregates, banks' liquidity requirements, etc.
The first monetary policy decision which the ECB will take can be likened to that taken by a motorist who, before setting off, has to decide on the cruising speed at which he is going to drive the car. However, before taking that decision, a car will have to be built, with all the features necessary to reach the desired speed; the destination of the journey will also need to have been decided.
Let us talk briefly about the construction of the car - that is to say, the operational framework for the conduct of the single monetary policy - by taking a look at its main characteristics. As regards construction lead times, the car will need to be ready for the test-bed by mid-1998. Approximately six months are being allowed for final testing and the assembly of the last features. The aim is thus to present the Governing Council of the ECB at the beginning of 1998 with a basic model and several options, the adoption of which will be for the Governing Council to decide. By the end of 1996 the preparatory stage of the prototype model will have been completed, at least as regards the main characteristics. The Treaty in fact stipulates that "at the latest by 31st December 1996, the EMI shall specify the regulatory, organisational and logistical framework necessary for the ESCB to perform its tasks in the third stage". The whole of 1997 and part of 1998 will be devoted to finalising the project and to implementing it.
The lead times have been estimated on the basis of the need to adapt the existing infrastructures at the national central banks and the financial institutions concerned, and that to create new information technology structures to interconnect the various elements of the ESCB, for we shall not be starting from scratch. An infrastructure already exists at each NCB that will need to be made compatible for participation in the European System of Central Banks. Thus it will be a question of harmonising, regulating, expanding and interconnecting the existing infrastructures in order to produce a single coherent vehicle.
The basic model can easily be deduced from the scenario for the first day of Monetary Union.
Regular open market operations will probably be the main tool whereby the ESCB will inject or withdraw liquidity from the money market. This instrument is currently used by all the central banks. It will need to be harmonised and integrated within a single framework to enable the decentralised execution of these operations. It will also be possible to carry out this kind of operation more frequently during the day.
The two other basic elements are the standing facilities. There will be a marginal lending facility, which currently exists at ten of the fourteen NCBs. This facility, like all credit-granting activity on the part of the ESCB, will be fully collateralised. It is necessary, therefore, to define the procedure, access and type of securities which will be eligible within the framework of this instrument. The other standing facility will be a facility for placing deposits with the NCBs. This facility is currently only available in a few countries, but its establishment does not give rise to any particular problems. These two instruments determine the margin for fluctuation of the overnight rate, which will be the operational objective of monetary policy. If the overnight rate exceeds the marginal lending rate, operators would prefer to have recourse to central bank credit, thereby setting a ceiling on the market rate; however, if the market rate falls below the deposit facility rate, operators would gain by placing deposits with the central bank; this would thus set a floor on the market rate.
Let us turn now to the options which will need to be available. There is in fact only one major option: compulsory reserves. I do not wish to enter the debate for or against compulsory reserves. Compulsory reserves are basically used to achieve two goals: i) monetary control; and ii) the stabilisation of money market rates. For each of these goals, the effectiveness of this instrument depends on a range of conditions such as the controllability of the money supply, the possibilities of substitution between monetary assets, and the nature of the monetary shocks that will affect the single currency area. At present, ten out of fourteen countries have compulsory reserves, but with different ratios and rates of remuneration. It is difficult to foresee at present whether the economic and financial conditions of the single currency area at the beginning of 1999 will warrant the use of compulsory reserves and, if so, on what terms.
The instrument of compulsory reserves will definitely need to be available to the ESCB. Given the length of time necessary to construct this instrument in the single currency area, it will need to be prepared in advance to ensure that the ECB will indeed have a choice. This preparation is scheduled to be carried out over the next two years. One objective is to minimise the costs of preparation for financial intermediaries, particularly by using simplified statistical reporting rules.
In any case it will be for the ECB to decide whether or not this instrument is to be used and to select its characteristics, particularly concerning the reserve ratio, basis, level of remuneration, and penalties to be imposed in the event of non-compliance, etc. This decision will be taken towards mid-1998, to allow a gradual implementation and testing of the system during the second half of 1998.
Foreign exchange market interventions represent a further option which could occasionally be used by the ECB. As we shall see shortly, it is unlikely that this instrument will be a key component in the conduct of monetary policy. However, it should not be excluded from the range of instruments available. The ESCB will be able to perform either centralised or decentralised foreign exchange market interventions via the NCBs. The EMI is in the process of drawing up plans to implement the infrastructure necessary to carry out interventions in accordance with these two models.
The other major decision which will need to be taken concerns the destination of the car, in other words the strategy of the single monetary policy. The Governing Council of the ECB will probably make its decision towards the end of 1998. The success of monetary policy also depends on the markets and the general public having a sound understanding of the monetary strategy of the ESCB.
Once again, it is difficult at present to give an answer to this question. It will depend on the macroeconomic conditions facing the ESCB in three years' time. While continuity and coherence are important principles for the central banks, so too are pragmatism and realism. The ESCB must take account of the fact that the move to Monetary Union and the changeover to the single currency may entail considerable changes in the fundamental economic and financial behaviour of operators.
What are the options? There are essentially two, given the current practice in European countries.
The first option is represented by a quantitative monetary target. We know that the effectiveness of this option depends in particular on the stability of money demand. This is an empirical issue which is currently being studied. Some results of this empirical research are encouraging and in fact show that a European monetary aggregate would be quite stable, even relatively more stable than national aggregates. Further research in this field is necessary.
The other alternative is a direct inflation target. This consists in deriving a forecast of inflation from a structural macroeconomic model of the economy. If the forecast moves away from the target, the central bank acts accordingly on the instruments of monetary control. Here again, a sound macroeconomic model is needed for forecasting European inflation which specifies the relationship between monetary policy instruments and the effective and nominal variables of the European economy.
One option which does not appear to be feasible is that of an exchange rate target. Admittedly, one cannot exclude the possibility that the single currency area will have exchange rate agreements with other countries, particularly countries with a derogation, but such agreements will have to safeguard the ability of the ESCB to pursue its primary objective in terms of price stability.
It is difficult to say which of the first two options will be chosen. In effect, the ECB will have to rely, in terms of analytical capacity, on either a money demand model or a structural system of the European economy which would make it possible to predict inflationary trends. The choice of which target to use will depend upon the forecasting ability of the two models. Which brings us back to the old debates on the forecasting ability of structural models as opposed to the reduced forms which were so popular with economists and econometricians in the 1960s.
Central banks are more pragmatic. The two approaches can be considered as complementary. On the one hand, a quantitative money target is also based on a forecast of the rate of inflation and potential economic growth. On the other, an inflation target requires the modelling of the reaction of the economy to the monetary variables controlled by the central bank, and thus also requires a money demand function.
In short, the choice will be made on the basis of detailed analyses which will be carried out first by the EMI and then by the ECB. This choice will be put forward and discussed in academic and market circles in the second half of 1998.
Once this decision has been taken, the ESCB's itinerary will have been finalised. For the Governing Council of the ECB, all that will remain to be done at the meeting to be held over the weekend preceding 4th January 1999 will be to turn the key in the ignition and start the car.
It is perhaps possible to outline briefly what we can expect after the first few hours of 4th January 1999, that is, to describe what market operators will be doing. The EMI is invariably cautious when it comes to predicting the behaviour of other institutions, whether public or private.
What we can certainly expect to see, however, is that operators will start to take advantage of the arbitrage opportunities which will emerge on the money market. This is a necessary condition for the implementation of a single monetary policy.
In the interbank segment, transactions will be effected not only within each country, but also between operators located in the various financial markets in the single currency area. There will no longer be a money market for country X or Y, but rather a single money market for the single currency area.
This will be made possible by the payment infrastructure that the EMI is in the process of implementing. The TARGET system (the Trans-European Automated Real-time Gross settlement Express Transfer system) will allow the settlement - in real time and on a gross basis - on accounts held at the ESCB of all transactions between participants in the system. This infrastructure will be more advanced than that currently in place in European countries, most of which are not yet real-time gross settlement systems. The TARGET system will be composed of the national systems and the Interlinking mechanism, which will interconnect the various national systems.
The TARGET system will be able to operate in euro from the start of Stage Three, and settlement will be effected via accounts held at the NCBs, likewise in euro. Opportunities for arbitrage and transactions within the single currency area will arise naturally. The single monetary policy will be conducted in a manner based on the principle of competitive markets offering counterparties from each country the opportunity to participate in the operations of the ESCB as a whole. This is to say that there will not be quotas of central bank money allocated to each country, and liquidity will be distributed throughout the area by market mechanisms. Therefore, operators will need to be prepared to effect monetary transactions with counterparties across the area. The creation of a new integrated market requires a certain degree of standardisation. The denomination of operations in euro will be a first step. But it will also be necessary to standardise certain procedures, such as the calculation of interest rates, maturities, and collateral. In short, a market agreement will be needed for the single currency area. The role of the market operators will be to prepare and develop this agreement, and the authorities will certainly provide all the necessary support.
The process I have just described will extend to all the other segments of the financial markets. Advantage will be taken of the opportunities for arbitrage between securities denominated in various monetary units to the point at which returns are equalised, with the exception of credit risk.
Within a relatively short period of time, there will be a single money market in the single currency area. No geographical market segment can afford to remain isolated from this area, since it would run the risk of being at a disadvantage, especially in terms of relative liquidity conditions.
First of all, let me remind you that the preparation of the single monetary policy is only one of the many challenges that will face the monetary authorities over the next two years in ensuring the start of Stage Three. Other challenges will be just as important, and also represent conditions necessary for the start of Monetary Union. The first is to ensure the convergence of the economies of the countries which will be participating in this area, in terms of the criteria laid down by the Treaty. The second is to ensure that the changeover to the single currency by a group of Community countries does not lead to a rift in the Community, and, therefore, that there will be mechanisms to ensure that the other countries speed up the convergence process in order to be able to participate when the time comes. The third challenge is to ensure that the changeover to the single currency is accompanied by a strengthening, not only of the monetary, but also of the fiscal and financial stability of the area.
As far as the preparation of the single monetary policy is concerned, the EMI's work is in line with the established deadlines. Estimates show that the vehicle will be ready to be made available to the ECB once the latter has been established.
The setting-up of the EMI was an exception in the history of the Community. In fact, it was the first time a temporary institution had been established in the Community. In view of the precedents of temporary international institutions, none of which has ever ceased to exist, one might worry that the EMI will last forever. All members of the EMI, from its President and Council to the staff, are committed to ensuring that the conditions are created which will enable the EMI to accomplish its task and indeed cease to exist when the time comes to pass the baton to the European Central Bank.