Tracking the Economy in the Largest Euro Area Countries: a Large Datasets Approach

Summary

The paper proposes a set of monthly business (growth-) cycle indicators for Germany, France, Italy and the euro area useful for ex post characterization of the cycle, and, most importantly, to assess the current economic outlook. These indicators are projections of quarterly aggregates on the space spanned by a set of regressors extracted from a large panel of monthly series. Being based on static linear combinations of monthly series, they do not suffer from the end-of-sample problem associated with traditional bilateral filters (HP filter). The indicators are used to: (1) study the degree of co-movement and synchronization across economies; (2) derive a dating of the cycle; (3) obtain the ‘stylized’ cyclical facts; (4) assess the predictive content of the panel for GDP growth. The monthly indicators are good forecasters of GDP performing often better than other simple methods. As expected, since the growth cycle indicator is a ‘smoothed’ estimate of the GDP growth, the best forecasts are obtained in terms of year-on-year (rather than quarter-on-quarter) GDP growth.

The authors wish to thank Olivier de Bandt, Sandra Eickmeier, Heinz Herrmann, the discussants and the participants of the conference held in Paris for their helpful comments. Many ideas presented here were first developed in collaboration with Filippo Altissimo, Antonio Bassanetti, Mario Forni, Marco Lippi and Lucrezia Reichlin in a joint research that led to the construction of the euro area business cycle indicator Eurocoin, currently published each month by the CEPR. The usual disclaimer applies.