Recent events in Eastern Europe have provided a windfall of new opportunities for the analysis of economic phenomena. In particular, the unification of Germany in July 1990 permits the study of exchange rate behaviour in a regime in which market participants knew with certainty that impending monetary union would fix the exchange rate for ever. In Discussion Paper No. 485, Research Fellow Michael Burda and Stefan Gerlach consider the prediction of standard exchange rate theory that the exchange rate in such a regime should be a weighted average of `fundamentals' relative money supplies, price levels, economic activities and other factors relating to money demand and also expectations of the `pegging' or the terminal exchange rate, with weights related to the sensitivity of money demand to its opportunity cost.

In an efficient market on the day before an anticipated fixing, where the terminal rate is unknown, the realization of the fundamentals will have some impact on the currently observed exchange rate, but the market exchange rate will be dominated by expectations of the terminal rate. In contrast, when the anticipated currency unification is far in the future, the weight on the expected terminal rate will be small, and the market rate will be largely determined by the fundamentals. Burda and Gerlach estimate and test such a model using daily data for the Ostmark/Deutschmark cash exchange rate in West Berlin during August 1989-June 1990. Standard regression procedures are inappropriate, since the expected final pegging rate and the `fundamentals' are largely unobservable. The authors therefore employ state-space (Kalman filtering) techniques to estimate key parameters, assuming that the market believed over the entire sample that 1 July 1990 would be the day of monetary union at an unknown exchange rate, but with a time-dependent variance.

Burda and Gerlach find that the behaviour of the exchange rate changed over the sample period. A random-walk model fits the data well for the first half of the period, during which considerable uncertainty remained whether monetary union would in fact take place. For the second half of the sample, during which the likelihood of monetary union increased, the exchange rate behaved as a weighted average of fundamentals and an expected terminal exchange rate. The sharp appreciation of the Ostmark after March 1990 supports these conclusions, but the data nevertheless cannot reject the possibility that the market anticipated German monetary union over the entire period, i.e. as early as September 1990.

Exchange Rate Dynamics and Currency Unification: The Ostmark-DM Rate
Michael Burda and Stefan Gerlach

Discussion Paper No. 485, December 1990 (IM)