MODELLING AND FORECASTING STOCHASTIC VOLATILITY: EVIDENCE FROM THE STOCK EXCHANGE OF THAILAND

In financial markets, volatility particularly refers to the spread of all likely returns of an asset and it is very important in financial modeling, such as in return predictability modeling. This paper use stochastic volatility approach to modelling and forecasting stock return volatility based on the various data frequencies. Empirical results from the Thai stock market showed that the stochastic volatility model has better (unobserved) conditional volatility forecasting power than GARCH (1,1) in the low frequencies date, i.e. monthly, quarterly and yearly. The results suggest the application stochastic volatility in the volatility modeling and forecasting for the portfolio allocation and risk management in the medium- to long- horizons, which is the crucial parts of the portfolio management and derivative pricing.