Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | 王士顯 | en_US |
dc.contributor.author | 鍾惠民 | en_US |
dc.contributor.author | 周幼珍 | en_US |
dc.date.accessioned | 2014-12-12T01:18:29Z | - |
dc.date.available | 2014-12-12T01:18:29Z | - |
dc.date.issued | 2007 | en_US |
dc.identifier.uri | http://140.113.39.130/cdrfb3/record/nctu/#GT009539527 | en_US |
dc.identifier.uri | http://hdl.handle.net/11536/39373 | - |
dc.description.abstract | 預測波動度對於衍生性商品定價、資產配置、風險管理以及套利都是非常重要的議題。ARCH以及GARCH衍生出的各種預測波動度模型也廣泛的被使用。最近的研究顯示,對於預測波動度來說高頻率(5min)的資料比日資料能提供更多資訊。Corsi(2004)提出了heterogeneous autoregressive (HAR)預測模型、Ghysels, Santa-Clara, Valkanov (2006)則提出mixed data sampling (MIDAS) 預測模型。之前的研究顯示HAR與MIDAS比ARCH跟GARCH有更高的解釋能力。近年來,使用無母術方法將波動度分離成連續以及跳躍的不同部分,更加強了模型的解釋能力。我們將使用Andersen,Bollerslev和Diebold(2005)的RV與Christensen和Podolskij(2007)的RRV兩種分離連續與跳躍的方法。並加入HAR與MIDAS模型中一起比較HAR-RV-CJ、HAR-RRV-CJ、MIDAS-RV-CJ與MIDAS-RRV-CJ模型的預測能力。 關鍵字:HAR | zh_TW |
dc.language.iso | zh_TW | en_US |
dc.subject | HAR | zh_TW |
dc.subject | MIDAS | zh_TW |
dc.subject | RV | zh_TW |
dc.subject | RRV | zh_TW |
dc.subject | 連續 | zh_TW |
dc.subject | 跳躍 | zh_TW |
dc.subject | HAR | en_US |
dc.subject | MIDAS | en_US |
dc.subject | RV | en_US |
dc.subject | RRV | en_US |
dc.subject | continue component | en_US |
dc.subject | jump component | en_US |
dc.title | HAR-CJ與MIDAS-CJ模型預測波動度之研究 | zh_TW |
dc.title | Forecasting Volatility by HAR-CJ Models and MIDAS-CJ Models | en_US |
dc.type | Thesis | en_US |
dc.contributor.department | 財務金融研究所 | zh_TW |
Appears in Collections: | Thesis |