The objective of this study is to measure the stock market volatility in Korea Security Market using GARCH model and to examine the utility of GARCH model in KOSPI returns.
Daily and Monthly return data are used for this study and it is found that monthly data from daily returns are more adequate for GARCH model than daily return data. The results reveals that daily stock returns can be characterized to the GARCH model. A succinct measure of the persistence of variance as measure by GARCH is the sum, α1 + β1 and it approximately approached unity in the daily return data. This represents the fact that the persistence of shocks to volatility exists in Korea Security Market. To see the effectiveness of GARCH model in forecasting the future volatility with the other forecasting models, ARCH and ARMA model were introduced and we generate forecasts for an daily and monthly returns using VAR model and technical statistics. It is hard to conclude that GARCH model had great explanatory power in the variance equation compared to ARCH and ARMA models.
While this Conclusion is strictly valid only for our sample of KOSPI returns, it is plausible to surmise that similar results would be found for other asset return series that can be explained by GARCH model.