碩士 / 國立臺灣海洋大學 / 航運管理學系 / 98 / This research take Baltic Dry Index (BDI) as the analysis object, whether the range-based model is optimal or not with the forecasting performance, and compares with returns-based GARCH model and the EGARCH model, the sample period is 1999/11/1~2010/2/28, has including around the term of Financial Crisis. The empirical results show that both the in-sample and the out-sample forecasting range-based model are better than GARCH model, forecasts in the achievements outside the sample. Discusses the BDI in the EGARCH model whether to have the leverage effect, discovered that the leverage effect’s coefficients are not all significant, therefore, the BDI does not exist the leverage effect. Finally, the discussion model joins the exogenous variable, on amplitude, the upward range in the ACARR model matches on suitable and all exogenous variables are insignificant, except the coal variable on downward range ACARR model is significant. The GARCH model exists the fuel oil and the coal are significant for model suitable. Joins the exogenous variables to the models (ACARR-X model and GARCH-X model) in the long-term forecast and has not distinct improvement performance, but in short-term forecast, ACARR-X model is better than GARCH-X model in the forecasting achievements. In addition, the upward range and exogenous variable are insignificant, demonstrated that downward range to have the sensitive influence regarding the external condition, is also the external condition in downward range implies more information than upward range, this conclusion is same with the case which Yang, Chou and Wu(2008) obtains.
Identifer | oai:union.ndltd.org:TW/098NTOU5301053 |
Date | January 2010 |
Creators | Chih-Chung Yi, 易至中 |
Contributors | Heng-Chih Chou, 周恆志 |
Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
Language | zh-TW |
Detected Language | English |
Type | 學位論文 ; thesis |
Format | 76 |
Page generated in 0.0018 seconds