Abstract
Employing the MS-ARJI-GJR-GARCH-X model, in which the parameters for the jump process, the asymmetric GARCH effect and the impacts of oil price shocks are regime-dependent, this paper analyzes the impact of crude oil price shock on stock return dynamics. Empirical results reveal three interesting findings. First, incorporating the asymmetric GARCH effect and the oil price shock can substantially improve fitting ability. Second, the GARCH and jump components show very different behaviors during turbulent and stable periods. Third, the effects of current and past oil price shocks differ. The conditional mean, mean of jump size and variance of jump size immediately respond to a current oil price shock. A one-period lagged oil price shock, no matter whether positive or negative, can affect the transition probability that the stock market will remain conditional in the next period. Moreover, the effects of lagged positive and negative shocks on transition probabilities are very different. © 2013.