A Study on Volatility Forecasting Models under Different Scenarios－Case of TAIEX Index Option

• Main Authors: Shao-Ping Chen, 陳少萍
• Other Authors: Dr.Yen-Shin Cheng
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/38951960641521166390

碩士 === 國立高雄應用科技大學 === 金融資訊研究所 === 96 === In order to evaluate the price of TAIEX options (TXO) accurately, this study applied Historical model, GARCH Model, Volatility Index (VIX), Implied Volatility Functions (IVF) and Genetic Algorism (GA) to estimate the volatility of option. Moreover, the estimated prices of volatility were inputted into BS model to calculate the theoretical prices, than we could find out the most appropriate model to estimate the volatility through comparing inaccuracy between theoretical prices and realized option prices. The study shows that no single model can be suited to all deal scenarios of TXO. The investors should choose suitable models according to their needs, and they can acquire the most accurate prices. Classified by the differentials between moneyness and maturity date, and analyzing the inaccuracy of the price, it’s the most accurate to adopt GARCH(1,1) model to estimate the volatility in the portion of the call. It’s most efficient to choose the evaluation of IVF model in the portion of the put. In addition, no matter what evaluated model, the theoretical prices what you evaluate will diverge from realized prices if only the days which is away from maturity date is longer. In other words, the inaccuracy of the price is more. Furthermore, this study also reveals that the GA model can give better performance under the situation of higher sampling frequency, both in call and put. Nevertheless, GA model is suited the scenarios which away from maturity date are shorter.