벡터오차수정모형(VECM)과 다변량 GARCH모형을 이용한 KOSPI200 현물과 선물간의 선도지연관계에 관한 연구

The purpose of this study was to identify the lead-lag relationships of volatility through spillover effects between log(return) variables of KOSPI200 spot and those of futures markets using VECM and multivariate GARCH model. The findings can be summarized as follows. First, In Johansen cointegration test on log variables of instable series, each market had 1 cointegration vector showing cointegrating relations. Second, this study adopted VECM, a conditional mean equation to test short-run dynamic adjustments and long-run equilibrium relationships between spot and futures, and also selected Off-diagonal BEKK model, a conditional volatility equation. 1 minute data had effects between spot and futures, indicating the existence of bidirectional mean spillover effects, but there was no volatility spillover effects at all. And most of 5 and 15 minute data had effects from futures to spot, indicating the existence of unidirectional mean spillover effects. Both data had effects between spot and futures, suggesting bidirectional volatility spillover effects. Thus, futures market led spot market through mean spillover effects, and spot market led futures market or vice versa through volatility spillover effects between two markets. And 1 minute data showed significant short-run dynamic adjustment factors and the relationships of cointegration between two markets, so both spot and futures adjusted imbalanced conditions. In 5 and 15 minute data, partial and short-run disequilibrium between futures and spot was dynamically adjusted, so the relationships of cointegration could be restored in the section of futures market. Finally, the results are consistent with those of previous studies as there are lead-lag relationships of log(return) between spot and futures market through conditional mean spillover effects and those of volatility between two markets through conditional volatility spillover effects as shown in VECM and Off-diagonal BEKK model.