Comparing monetary policy rules in CEE economies: A Bayesian approach

a r t i c l e i n f o JEL classification: C11 E32 E52 Keywords: New Keynesian models Small open economy Monetary policy Taylor rules Bayesian methods Using the Bayesian approach, a small open economy DSGE model was estimated using a sample of quarterly data for three Central and Eastern Europe economies, Czech Republic, Hungary and Poland. The hypothesis that central banks react to exchange rate movements was tested using posterior odds ratio. For these economies , evidence was found that central banks reacted to exchange rate changes. Evidence of similar monetary policy characterized by moderate or low gradualism as well as an active and conservative monetary policy was also found, for the selected countries. When a richer DSGE model featuring habit formation and imperfect pass-through is estimated, the results are generally similar. The inclusion of exchange rate in Taylor rule can also drastically change the dynamics of inflation and output following certain shocks. Whether central banks react or not to exchange rates debate may be traced back to the 2001 contribution of Taylor. He showed that a successful monetary policy should be based on a mix of flexible exchange rate, inflation target and monetary policy rule. The debate followed two main directions. The first direction focused on the benefits brought by a monetary policy responsive to exchange rate, while the second direction focused on how central banks actually behave. This paper follows the second approach, testing whether central banks in several Central and Eastern European (CEE, hereafter) economies responded to exchange rate movements. Univariate approaches were initially used to address the question. Clarida et al. (1998) found, for some industrialized countries, that monetary policy responded to exchange rate fluctuations. Additional evidence was found by Calvo and Reinhart (2002) for the case of emerging economies. According to them, central banks reacted to exchange rate fluctuations through monetary policy in order to smooth the exchange rate variation. More recently, the same topic was studied from a structural perspective within the dynamic stochastic general equilibrium approach. In one of the first studies, Lubik and Schorfheide (2007) estimated a small open economy New Keynesian (NK) model using a Bayesian approach for a selection of developed small open economies (Australia, Canada, New Zealand and UK). They tested based on posterior odds ratios if central banks reacted to exchange rate movements. According to their results, evidence was found that central banks reacted to …