|
| |
N° 1997 - 06 |
  |
| April |
| The Exchange Rate Policy of the Euro: a
Matter of Size? |
| Philippe Martin |
|
There are many things we do not know about
European monetary integration as there is no historical precedent to it. However,
we know for sure that the size of the zone created by the EMU will be bigger (far
bigger in the case of EMU15) than European countries taken separately. In this
paper, we start from this size effect of EMU and exploit it to analyse from a
theoretical and empirical point of view the question of exchange rate policy with
the EMU. We argue that country size matters as a determinant of monetary policy
in an open economy policy and volatility because it changes the incentive to use
monetary policy to influence the exchange rate.
One way to analyze this size effect in economic terms is to interpret the creation of the Euro as implying a change in the relative size of players in international monetary relations. From game theory, we know indeed that the size of players matters in determining equilibrium strategies and economic outcomes. We construct a simple model which gives a central role to country size as a determinant of monetary policies and exchange rate variability. We then test its empirical implications and exploit the results to suggest some consequences for the euro exchange rate.
In the theoretical section, a two country model is developed in which unanticipated changes in the exchange rate can help countries stabilize their economy when shocks occur. This is because, during a recession for example, a country can gain employment and output if it succeeds in lowering its real wage below the level of other countries which amounts to a real depreciation. A large country will have less incentive to use strategically its monetary policy to stabilize its economy than a small country because its output depends less on the exchange rate than a small country. Hence, large countries should have more stable exchange rates than small ones. However, when the country becomes " very small ", exchange rate variability becomes more important as a source of output variability than domestic shocks. This implies that " very small " countries will use their monetary policy to stabilize their exchange rate. Hence, we find that exchange rate variability should be a hump-shaped function of country size. The model also predicts that exchange rate variability should be an increasing function of the variability of shocks in both countries (because the exchange rate reacts to those shocks) and a decreasing function of the correlation of shocks (because a high correlation of shocks induces similar monetary policies and a stable exchange rate).
As EMU entails the creation of very large common currency zone, the hump shape relation suggests that the euro exchange rate should be more stable than the European currencies. From a theoretical point of view however, other channels may lead to higher exchange rate volatility. Other authors argue on theoretical grounds that EMU should lead to higher exchange rate instability because being a more closed economy, it will be less concerned with trade imbalances and the inflation consequences of exchange rate changes. These papers belong to the tradition of models which predict that the fixing of the exchange rate (in the case of EMU its elimination) transfers the volatility to other macroeconomic variables (in the case of EMU to the interest rate and exchange rate with other currencies).
As the theory can predict a positive or a negative effect of EMU on exchange rate instability, the next natural step is to attempt to analyze the question at the empirical level.
We test the implications of the theoretical model on a cross section of 210 bilateral exchange rates of the OECD countries for the period 1980-1995. We find that the model works well and that indeed the volatility of the exchange rate seems to have a hump shaped relation with country size even after controlling for several other determinants of exchange rate volatility such as the EMS, the use of the currency as a reserve currency, correlation of shocks, bilateral trade and volatility of shocks. The relation also works reasonably well on subsamples of bilateral exchange rates which exclude certain exchange rates which we believe not to be independent.
We also find that the narrow band EMS currencies have had more stable exchange rates with respect to currencies not in the EMS than non EMS currencies. This is so even after controlling for currencies in the wide band EMS. This is an important result as EMS is the closest experiment of EMU that we have. It suggests that a quasi-fixed exchange rate zone not only stabilizes exchange rates in the zone but also exchange rates of the zone with outside currencies.
Finally, we use the empirical model to make a "back of the envelope" prediction exercise of what the dollar/euro variability might be compared to the past variability of the dollar/DM exchange rate. We find that it should decrease and consistent with the results of the model this decrease should be more important the larger the size of the EMU. The decrease becomes quantitatively important in the case of a large EMU not restricted to the core countries.
Our results can be interpreted as describing how the incentives to use monetary policy to influence the exchange rate in order to stabilise output may change with EMU. They can therefore shed some light on the likely political pressures that the independent ECB will have to resist. Some economists and policy makers fear that if the responsibility of the exchange rate policy is in the hands of the governments rather than the ECB, this may lead to conflicts on exchange rate policy and therefore monetary policy. Our results suggest that the problem of possible conflicts between the ECB and the European governments on the management of the exchange rate policy with respect to the dollar may be exaggerated.
The configurations explain why (but it is not the only reason) big monetary areas cannot agree upon a co-ordination procedure concerning exchange rates in the Williamson sense. |
Abstract |
| |
|
| |
Keywords |
| |
JEL classification |
| To visualise the full text document, use Acrobat
Reader |
Full text (pdf) |
|
|
| |
|
|
|
|
|
