Using VIX Index and Range-based Volatility to enhance the Volatility Forecasting Performance of GARCH Models

碩士 === 德明財經科技大學 === 服務業經營管理研究所 === 97 === This research data was based on the S&P500 index from the period January 2nd 2001 till December 31st 2007, and the prediction period was from December 21st 2003 to December 31st 2007, totally 508 observations. By using the VIX index and range-based volat...

Full description

Bibliographic Details
Main Authors: Chen Hsin-Hua, 陳信華
Other Authors: Ni Ren-Xi
Format: Others
Language:zh-TW
Published: 2009
Online Access:http://ndltd.ncl.edu.tw/handle/00551982882809532660
Description
Summary:碩士 === 德明財經科技大學 === 服務業經營管理研究所 === 97 === This research data was based on the S&P500 index from the period January 2nd 2001 till December 31st 2007, and the prediction period was from December 21st 2003 to December 31st 2007, totally 508 observations. By using the VIX index and range-based volatility as the exogenous variable, the objective of the research was to identify the model, other than the GARCH model, capable of improving the one-step-ahead volatility prediction performance under the circumstances with or without consideration of the error distributions respectively. The rolling window method was used in the one-step-ahead volatility prediction to compare the advantages and disadvantages of the models. The results, which were volatile projections, were obtained first by using the MSE, MAE, MME, and VaRE, and the SPA test to reveal the statistical significance among the competing models.