A Comparison of the Forecasting Performance between GARCH and Implied Volatility Models

• Main Authors: Wei-Peng Chen, 陳煒朋
• Other Authors: Huimin Chung
• Format: Others
• Language: zh-TW
• Published: 1999
• Online Access: http://ndltd.ncl.edu.tw/handle/67696532063274292945

碩士 === 淡江大學 === 財務金融學系 === 87 === Volatility forecasting is very important to derivative pricing, hedging, and risk management. The introduction of option market might provide new information regarding the market''s consensus of future volatility. Previous empirical results on the forecasting performance of implied volatility model find mixed result for the U.S. market. This thesis compare the forecasting performance of a variety of volatility modes for Taiwan and Hongkong covered warrant markets. Using daily data of both Hongkong and Taiwan markets, the empirical results demonstrate that the historical volatility and GARCH-type models outperform implied volatility model. The implied volatility (IV) model has little information content in providing forecast for future volatility. Similar result are found when the intraday data is used to construct daily observation of volatility. The information content of trading volume in term of daily turnover is also examined. When trading volume is added in the IV model the forecasting performance of IV model increases.