Engle의 DCC모형을 이용한 금융시장내 변동성의 상관관계에 대한 연구

The purpose of this study is to examine interday dynamic correlations of volatility through spillover effects on own and cross financial markets using VAR, VEC, full BEKK-MGAECH and DCC model. The findings are summarized as follows. First, all historical correlations were statistical significant among three variables. Second, coefficient of variables had stable and mean-reverting, so there wasn't a long memory in individual ARFIMA, but there were cointegrating relations in each market on log variables of instable series in Johansen cointegration test. Third, in VEC and Full BEKK-MGARCH model, bond yields made a partial adjustment when it drifted away from long-run equilibrium with stock returns or foreign exchanges. And foreign exchanges parrtially adjusted disequilibrium conditions with stock returns or bond yields. In total, before & after crisis, there was a negatively(positively) unidirectional conditional mean spillover effects from bond yields to stock returns(foreign exchanges), which bond yields led stock returns(foreign exchanges). I found a negatively strong unidirectional conditional volatility spillover effects from residual(volatility) impacts of bond yields to stock returns(volatility), indicating that bond yields led stock returns in all periods. Except crisis period, foreign exchanges(stock returns) increased volatility of stock returns(foreign exchanges) in total and after crisis, so foreign exchanges led stock returns or vice verse through volatility spillover effects. Therefore, this paper showed various correlations of volatility in financial markets in sample period and there were timevarying dynamic conditional correlations of volatility among stock returns, bond yields and foreign exchanges through conditional volatility spillover effects as shown in VAR, VEC, Full BEKK-MGARCH.