本文以ARJI與GARCH-Constant Jump兩種跳躍-擴散模型,探討立委選舉事件對股匯市所產生跳躍狀態的頻率與跳躍狀態的變異。實證結果發現:(l)ARJI模型的配適度優於GARCH-Constant Jump模型。(2)股匯市均存在著異常資訊所造成瞬時的跳躍行為。(3)立委選舉事件使股匯市發生跳躍的機率增加,且較平時為大。
This study employs jump-diffusion model, including ARJI model and GARCH-Constant Jump model, to examine jump frequency and risk of Taiwan stock and foreign exchange markets during legislative election period. The empirical results indicate that, firstly, ARM model fits data better than GARCH-Constant Jump model. Second, Taiwan stock and foreign exchange markets exist jump behavior. Thirdly, Legislative election event made a great jump probability impact on Taiwan stock and Foreign Exchange Markets.