Markov Switching Monetary Policy in a two-country DSGE Model

In this paper I show, using both empirical and theoretical analysis, that changes in monetary policy in one country can have important e.ects on other economies. My ew empirical evidence shows that changes in the monetary policy behaviour of the Fed since the start of the Euro, well captured by a Markov-switching Taylor rule, have had significant e.ects on the behaviour of inflation and output in the Eurozone even though ECB’s monetary policy is found to be fairly stable. Using a two-country DSGE model, I examine this case theoretically; monetary policy in one of the countries (labelled foreign) switches regimes according to a Markov-switching process and this has nonnegligible e.ects in the other (home) country. Switching by the foreign central bank renders commitment to a time invariant interest rate rule suboptimal for the home central bank. This is because home agents expectations change as foreign monetary policy changes which a.ects the dynamics of home inflation and output. Optimal policy in the home country instead reacts to the regime of the foreign monetary policy and so implies a time-varying reaction of the home Central Bank. Following this time-varying optimal policy at home eliminates the e.ects in the home country of foreign regime shifts, and also reduces dramatically the e.ects in the foreign country. Therefore, changes in foreignmonetary regimes should not be neglected in considering monetary policy at home. Key words: Markov-switching DSGE ; Optimal monetary policy ; Dynamic programming ;SVAR ; real-time data. JEL Classification: E52 ; F41 ; F42.