This paper seeks to examine the nature of spillovers between output and stock prices using both a long annual time series spanning 200 years and a shorter but quarterly observed data set.
The authors’ particular interest is to examine both the time-varying nature of the spillovers and spillovers across the frequency using wavelet analysis.
The results reveal an interesting detail that is missed when considering spillovers for the raw data. Using annual long run data, spillovers in the raw data are in the order of approximately 10 per cent for stocks to output and 25 per cent for output to stocks. But this increases up to 50 per cent and above (in both directions) when considering different frequencies. Similar results are reported with the quarterly data, although the differences between the raw data and the wavelets are smaller in nature. Finally, output explains more of the variation in stocks than stocks explains in output.
The nature of these results is important for policy-makers, market participants and academics alike, while the use of wavelets provides information across different frequencies.
