Credibility of Interest Rate Policies in Eight European Monetary System Countries: An Application of the Markov Regime-Switching of a Bivariate Autoregressive Model

This chapter utilizes the Markov regime-switching modelling framework to study the credibility of monetary policy, with respect to the objective of price stability, in some member countries of the EMS throughout its life (i.e., 1979–98). Eight countries are examined for this purpose: Austria, Belgium, Finland, France, Italy, the Netherlands, Portugal and Spain. In addition, Germany is used as a benchmark case.1 A number of other studies have investigated the issue of credibility during the EMS period (see for example, Dahlquist and Gray, 2000; Arestis and Mouratidis, 2004a). One of the main conclusions of this body of literature is that monetary policy may go through different stages of credibility through time, in such a way that it is perceived to be credible in some circumstances and may lack credibility on other occasions. It is, thus, appropriate to use the Markov regime-switching modelling framework to study this phenomenon. Moreover, some of these studies allow the probability of switching between regimes to be a function of macroeconomic variables (see Engels and Hakkio, 1996; Gray, 1996; Dahlquist and Gray, 2000; Sarantis and Piard, 2000). In particular, most of these studies attempt to explain the currency crises of 1992 and 1993 using as information variables in the transition probabilities a number of alternatives: the exchange rate within a band (i.e., a target zone model framework), real exchange rates, budget deficits, interest rate differentials, and other variables. None of these studies includes in the transition probability the two main variables that determine the loss function of monetary authorities, namely domestic output-gap variability and inflation variability (the exception is Arestis and Mouratidis, 2004a).2