短期利率之动态条件变异与预测绩效之探讨.pdfVIP

Academy Papers 17 94 17-32 The Stochastic Volatility and Predictability Efficiency of Short-Term Interest Rate Bor-Yi Huang Jer-Shiou Chiou Shih Chien University, Shih Chien University, Department of Finance and Banking Department of Finance and Banking Cho-Min Lin Chien-Hung Chen Ling Tung College, Department of Finance Tamkang University, Graduate Institute of Money, Banking and Finance GARCH (Generalized Autoregressive Conditional Heteroskedasticity) (Markov Switching Model) (Constant Intensity Jump Model) MAD RMSE MAPEMAPE GARCH GARCH 104 70 TEL 886-28927 E-mail:jschiou@mail.usc.edu.tw . 18 94 6 Review of Financial Risk Management Abstract Time series econometrics models of GARCH, Markov Switching and Constant Intensity Jump Model are adopted in this study for seeking a more reasonable and much effective model in predicting Taiwans short-term interest rate. The result indicates that Constant Intensity Jump Model would be a better forecasting model by the criterion of MAD, RMSE and MAPE. Consequently, base upon basic GARCH Model, and considering the factor of structure changes and jumps, Constant Intensity Jump Model can improve the forecasting ability of Taiwans short-term interest rate. Key Words: GARCH Model, Markov Switching Model, Constant Intensity Jump Model 1.