GARCH-jump models of metal price returns, while allowing for sudden movements (jumps), apply the same specification of the jump component in both 'bear'and 'bull' markets. As a result, the more frequent but relatively small jumps that occur in both bear and bull markets dominate the characterization of the jump process. Given that large jumps, although less frequent, are still quite common in copper (and other metal) markets, this is a potential shortcoming of current models. More flexibility can be added to the modeling process by allowing for regime-switching. In this paper we specify a model that allows for switching across two separate regimes, with the possibility of different jump sizes and frequencies under each regime, along with a regime-specific GARCH process for the conditional variance. This model is applied to daily copper futures prices over the period of January 2 1980 through the end of July 2007. The model is estimated both with and without factors such as interest and exchange rate movements entering into the specification of the state-dependent mean of the conditional jump size. In some respects, a Regime Switching GARCH-Jump Model performs well when applied to the copper returns data. The results are mixed in terms of whether or not variations of the model that allow jump sizes to be a function of interest or exchange rates offer much of an advantage over a pure time series approach to the modeling of copper returns over the past three decades.
Download Info
To download:
If you experience problems downloading a file, check if you have the
proper application to
view it first. Information about this may be contained
in the File-Format links below. In case of further problems read
the IDEAS help
page. Note that these files are not on the IDEAS
site. Please be patient as the files may be large.
Publisher Info
Paper provided by University of Alberta, Department of Economics in its series Working Papers with number
2009-13.