Common Stationary and Non-Stationary Factors in the Euro Area Analyzed in a Large-Scale Factor Model

In this paper, we rely on techniques recently developed by Bai and Ng (2004a) to estimate common euro-area stationary and non-stationary factors using a large-scale dynamic factor model. We find that euro-area economies share four non-stationary factors or trends and one stationary factor. By means of rotation techniques, we propose estimates of a euro-area business cycle and a euro-area stochastic trend. It turns out that our business cycle estimate is a very good match to EuroCOIN, the euro-area coincident business cycle indicator published by the CEPR. Fluctuations of common euro-area factors mainly reflect variations of German and/or French real economic activity as well as of prices, financial prices (long-term interest rates and/or real effective exchange rates) and confidence indicators in various countries. As concerns the transmission channels, macroeconomic shocks seem to proliferate in the euro-area more strongly through trade, exchange rates and long-term interest rates than through stock prices. Among the external driving forces, shocks to US economic activity seem to be more strongly linked to shocks to the euro-area factors than oil price shocks. We finally find evidence of mild overall convergence; results for individual countries are mixed