Modelling and Forecasting the Volatility and Price of Malaysian Stock Market

Abstract

Modelling and forecasting volatility of a financial time series has been a significant area of research in recent years, owing to the fact that volatility is regarded as an essential notion in many economic and financial applications. Because volatility is not directly observable, financial analysts are especially eager to obtain an accurate estimation of this conditional variance process. As a result, a number of models have been developed that are specifically suited to estimate the conditional volatility of financial instruments, with the most well-known and widely used model being the conditional heteroscedastic models. The objectives of this research are; to model and forecast the volatility and price of the Malaysian stock market; to assess the performance of competing models and to simulate the volatility and price of Malaysian stock market. This research estimates and examines the performance of ARIMA and GARCH family type models, standardized GARCH and GARCH-M for symmetric model and EGARCH and GJR-GARCH for asymmetric model using daily return price data. For GARCH models, two distributions were used which were the normal distribution and the t-distribution. The Malaysian stock market which is Kuala Lumpur Composite Index (KLCI) was studied using daily data over an 11-years period beginning from 1st January 2010 and ending on 31st December 2020. The results showed that the ARIMA method is not suitable in forecasting long term data since the ARCH effect is present. While, the performance of asymmetric GARCH models (GJR-GARCH), especially when the fat-tailed densities are taken into account in the conditional volatility, are better than symmetric GARCH. In addition, the student-t distribution performs better than the normal distribution. Moreover, it was found that the AR (1) GJR-GARCH model provides the best forecast for the Malaysian stock market, KLCI. Thus, it was concluded that the asymmetric AR (1) GJR-GARCH model coupled with the student-t distribution, performed well in modeling KLCI dataset. The forecasting process resulted in three different outcomes where the first one foresees a stationary trend of RM1,825.00 all over the years. While the second forecast indicates a fluctuation of around RM1,825.00 to RM1,625.00 and the last outcome had forecast the best result where the trend shows a positive upward trend of over RM1,850.00 all over the year of 2021.