DP6654 Should the Euro Area be Run as a Closed Economy?
|Author(s):||Carlo A. Favero, Francesco Giavazzi|
|Publication Date:||January 2008|
|Keyword(s):||DSGE models, ECB, monetary policy, yield curve|
|JEL(s):||E43, E52, E58|
|Programme Areas:||International Macroeconomics|
|Link to this Page:||cepr.org/active/publications/discussion_papers/dp.php?dpno=6654|
The European Economic and Monetary Union (EMU) has created a new economic area, larger and closer with respect to the rest of the world. Area-specific shocks are thus more important in EMU than country-specific shocks used to be in the previous states, e.g. in Germany. It is thus not surprising that the models built by the staff of the European Central Bank (ECB) to study optimal monetary policy in the Euro area (for instance Smets and Wouters, 2004a, 2004b) typically assume that this works essentially as a closed economy, hit by domestic shocks - the same assumption made in standard models of U.S. monetary policy (see e.g. Christiano et al., 1999 ), where all shocks are domestic with the only possible exception of energy price shocks. Two-country models exist at the ECB (e.g. de Walque, Smets, Wouters, 2005) but they overlook asset price fluctuations and their international comovements. This paper studies monetary policy in the Euro area looking at the variable most directly related to current and expected monetary policy, the yield on long term government bonds. We explore how the behaviour of European long-term rates has been affected by EMU and whether the response of long-term rates to monetary policy has got any closer to that consistent with a closed economy. We find that the level of long-term rates in Europe is almost entirely explained by U.S. shocks and by the systematic response of U.S. and European variables (inflation, short term rates and the output gap) to these shocks. Our results suggest in particular that U.S. variables are more important than local variables in the policy rule followed by European monetary authorities: this was true for the Bundesbank before EMU and has remained true for the ECB, at least so far. Using closed economy models to analyze monetary policy in the Euro is thus inconsistent with the empirical evidence on the determinants of Euro area long-term rates. It is also inconsistent with the way the Governing Council of the ECB appears to make actual policy decisions. We also find that Euro area long rates respond more to financial shocks, in particular shocks to term premia, than they do to monetary policy "shocks" - i.e. instances when the ECB deviates from its rule. This finding point to the importance of incorporating into the analysis of Euro area monetary policy of the effects of fluctuations in international asset prices.