Forecasting Malaysian stock market volatility

This study evaluates a battery of forecasting volatility models using daily data of the FTSE Bursa Malaysia CI Index. The forecasting models include random walk model, historical mean model, and moving average models. The mean error statistic, mean absolute error statistic, root mean squared error statistic, mean absolute percentage error statistic, under-predictions mean mixed error, and over-prediction mean mixed error were used to evaluate precision of forecasts. The results suggested that the 3 years moving average model offered advanced forecasts volatility information. Nevertheless, the different models rankings indicated that they are sensitive toward the error statistic employed in evaluating the precision of the forecasts.

PDF 12-2.pdf Restricted to Repository staff only Download (211kB)

Actions (login required)

View Item