Continuous wavelet transform and rolling correlation of European stock markets

The purpose of this paper is to assess the level of co-movements, contagion and rolling correlation between the stock markets of the PIIGS and those of the UK and Germany. We thus resort to a novel time–frequency approach, namely the continuous wavelet transform, and we analyze the co-movements of the stock index returns at different frequency-scales. We also test the influence of different macroeconomic factors on stock markets co-movements at different time–frequencies. The wavelet analysis results show that, in the short-run, the correlation level is high only during financial distress episodes, while in the long-run, the co-movements are present for the entire analyzed horizon. In addition, at low-frequency levels, the PIIGS stock markets are more correlated with Germany than with the UK. An opposite result is obtained at high-frequency decomposition levels. We also discover that the stock markets' correlation does not increase after all crisis events and depends on the wavelet decomposition levels.