A Comparative Study of the Taiwan and Japan Equity and Foreign Exchange Markets: Modeling, Estimation and Application of the Component Garch-in-Mean Model

The main purpose of this paper is to verify the effectiveness of the bivariate Component GARCH-in-mean (GARCH-M) model and analyze the interactions and risk premium of equity markets by exploring the short- and long-run volatility components on both the Taiwanese and Japanese equity markets. We show that unexpected shocks of volatility will in general influence the fluctuations of both equity and foreign exchange markets. Persistence on the long-run volatility components of both markets is also found. The results also reveal that the positive risk-return relation on equity markets can be further verified when the impacts of short and long-run volatility components are decomposed by the Component GARCH-M model. The decomposition can also facilitate reflecting the transitory and permanent volatility impacts of foreign exchange exposure on the returns of equity markets.