재무연구 제 권 호 (2011년 9월)
Asian Review of Financial Research, Vol., No..
pp.404~451
pp.404~451
Bivariate GARCH-Jump Model
Chuang-Chang Chang Department of Finance, National Central University Jung-Li 320, Taiwan
Hsiao-Wei Ho Department of Finance, National Central University Jung-Li 320, Taiwan
Tzu-Hsiang Liao Department of Finance, Ming Chuan University, Taipei 111, Taiwan.
Yaw-Huei Wang Department of Finance,National Taiwan University, Taipei 106, Taiwan.
In this paper we consider the pricing of quanto derivatives with the bivariate GARCH-Jump model, in which jumps take place in the price kernel, and consequently in foreign asset returns and in exchange rates. In the empirical investigation, we use Dow Jones, NASDAQ and NIKKEI 225 indexes, exchange rates and corresponding index warrants data to examine the effects of jump on derivative pricing. The empirical results suggest that the unrestricted bivariate NGARCH-Jump model outperforms the other four models considered in this study. The evidence also reveals that the average pricing error is the smallest for the unrestricted bivariate NGARCH-Jump model. Hence, the nonlinear asymmetric model with jumps captures the dynamics of index return and exchange rate well.
Chuang-Chang Chang
Hsiao-Wei Ho
Tzu-Hsiang Liao
Yaw-Huei Wang
In this paper we consider the pricing of quanto derivatives with the bivariate GARCH-Jump model, in which jumps take place in the price kernel, and consequently in foreign asset returns and in exchange rates. In the empirical investigation, we use Dow Jones, NASDAQ and NIKKEI 225 indexes, exchange rates and corresponding index warrants data to examine the effects of jump on derivative pricing. The empirical results suggest that the unrestricted bivariate NGARCH-Jump model outperforms the other four models considered in this study. The evidence also reveals that the average pricing error is the smallest for the unrestricted bivariate NGARCH-Jump model. Hence, the nonlinear asymmetric model with jumps captures the dynamics of index return and exchange rate well.
