A Study of Applying Asymmetric GARCH Model to Taiwan Stock Market

• Main Authors: Ping-Lun Tu, 杜秉倫
• Other Authors: Chu-Hsiung Lin
• Format: Others
• Language: zh-TW
• Published: 2002
• Online Access: http://ndltd.ncl.edu.tw/handle/30014516259703615579

碩士 === 國立高雄第一科技大學 === 財務管理所 === 90 === Black(1976) confirmed that stock market volatility is asymmetric in its response to news。The GARCH model doesn’t allow good news and bad news to have a different impact on volatility, so we adopt the asymmetric GARCH model instead of GARCH model. The article researches two subjects, one named Day-of-the-Week Effects and Behavior of Volatility in Taiwan Stock Market：Before and After the Periods on SIMEX Index Future Trading and the other named a study on the asymmetric mean reverting of short-horizon stock returns in Taiwan. In the first subject, we use GJR GARCH model. This subject analyses the impact of SIMEX Taiwan index future on daily return seasonality and volatility on Taiwan from January 5, 1993 to December 30, 2000. This subject indicates that low returns are observable in the Taiwan stock market on Tuesday before SIMEX index future listing. After SIMEX index future listing, the empirical result that Monday and Tuesday have negative return in the cash market is different from other countries. We also find Monday have negative return in the future market. The aim of this subject investigates explanations for the existence of anomalies. We find the day-of-the-week phenomenon in the Taiwan stock market, which would be regarded as weekend effect of the SIMEX index future. Next, SIMEX index future trading has significant effect on the spot market volatility. But it doesn’t improve the quality and speed of information to spot market and make market more efficient. Besides, this subject thinks that market dynamics doesn’t explain the reason of the asymmetric response of volatility to news in Taiwan market completely. In the second subject, we use Asymmetric Nonlinear Smooth-Transition GARCH model、Smooth-Transition GARCH model and Threshold-Switching GARCH model to design mean equation to analyze asymmetric mean reverting of short-horizon stock returns in Taiwan and to utilize variance equation to capture the characteristics of the Taiwan stock return’s volatility. Using capital market weekly data from January 4, 1986 to October 26, 2001, we find strong evidence of asymmetric mean reverting in Taiwan stock market. It implies that anomalous return following unexpected price drop tend to gain more quickly than it does following unexpected price rise. Moreover, we find that the relationship between future volatility and risk premium is negative when a negative return shock is realized. The findings support the market overreaction hypotheses. The asymmetry is due to the mispricing behavior on the part of investors who overreact to certain market news. Besides, the empirical results are shown that Asymmetric Nonlinear Smooth-Transition GARCH Model is similar to Smooth-Transition GARCH Model and Threshold-Switching GARCH Model in capturing the behavior of volatility.