-3-
domestic variables with no feedback effects possible.1 Finally, in contrast to Pesaran et
al. (2004), we choose the weighting in the construction of the common factors by using
principal components analysis.
Another series of papers has empirically checked the existence and significance of
common patterns in the international dynamics of macro variables. The focus has been on
the changes over time of business cycle synchronization across the most important
economies originating in common global disturbances. Assessing international business
cycle co-movement is mainly an empirical task and the main drivers of the development
may shift over time. Usually, this literature has taken into account only a limited set of
real quantities such as output, consumption and investment, which is, however,
sometimes researched for a large number of countries (Stock and Watson 2005).
When a broader range of variables is included in the analysis, the focus of the
literature tends to switch from the common driving forces of fluctuations to the spillover
of shocks. Many authors have investigated these issues.2 Most of them detected a
tendency for national business cycles to converge over the period of the second
globalization. Artis and Okubo (2008) provide a long-run historical perspective which, by
revisiting the era of the first globalization before the First World War, demonstrates a
tendency for globalization to produce a high degree of synchronization in national
business cycles. Stock and Watson (2005) conclude from their analysis that co-movement
has fallen during the 1984-2002 period relative to 1960-83 due to the absence of
common shocks. In this paper, we adopt a wider perspective and study co-movements
among the G7-countries plus the euro area, using a larger data set than previously
employed in the literature, including both real and nominal variables.
In the seminal contribution of Kose, Otrok and Prasad (2008), more than 100
countries are analyzed over the 1960-2005 period. Fluctuations in economic activity are
decomposed into global, country group and country specific factors. During the second
period of globalization (1985-2005), business cycles have converged among the group of
1 Not too different from our approach but in contrast to Bagliano and Morana (2009) who model all
variables as endogenous from the outset, Dees, di Mauro, Pesaran and Smith (2007) model each country
separately, with foreign variables treated as weakly exogenous.
2 See for example Artis and Zhang (1997), Kose, Otrok and Whiteman (2003), Artis, Krolzig and Toro
(2004), and Canova, Ciccarelli and Ortega (2006).