# 13-009/III (2013-01-08; 2013-02-01)

David E. Allen, Edith Cowan University; Michael McAleer, Econometric Institute, Erasmus University Rotterdam, Complutense University of Madrid, and Kyoto University; R.J. Powell, Edith Cowan University; A.K. Singh, Edith Cowan University
Volatility spillovers, Markov-switching GARCH, Cholesky-GARCH, Time-varying correlations
JEL codes:
C22, C32, G11, G15

This paper features an analysis of volatility spillover effects from the US market, represented by the S&P500 index to the Australian capital market as represented by the Australian S&P200 for a period running from 12th September 2002 to 9th September 2012. This captures the impact of the Global Financial Crisis (GFC). The GARCH analysis features an exploration of whether there are any spillover effects in the mean equations as well as in the variance equations. We adopt a bi-mean equation to model the conditional mean in the Australian markets plus an ARMA model to capture volatility spillovers from the US. We also apply a Markov Switching GARCH model to explore the existence of regime changes during this period and we also explore the non-constancy of correlations between the markets and apply a moving window of 120 days of daily observations to explore time-varying conditional and fitted correlations. There appears to be strong evidence of regime switching behaviour in the Australian market and changes in correlations between the two markets particularly in the period of the GFC. We also apply a tri-variate Cholesky-GARCH model to include potential effects from the Chinese market, as represented by the Hang Seng Index.