The purpose of the current study is to model Malaysian gold prices, known as Kijang Emas, using a popular class of generalized econometric models called Generalized Autoregressive Conditional Heteroskedasticity (GARCH) and three of its variants. The variant models selected are GARCH in the mean (GARCH-M), Threshold GARCH (TGARCH) and Exponential GARCH (EGARCH). While the standard GARCH and GARCH-M are symmetric models, TGARCH and EGARCH are asymmetric. Using Akaike Information Criterion (AIC) and Schwarz Information Criterion (SIC) as model selection criteria, the best fit model for modelling Malaysian gold is TGARCH.

No abstract

In developing a time series model, parameter estimation is one of the crucial steps. Common methods of estimation include method of moment (MME), ordinary least square estimation (OLS) and maximum likelihood estimation (MLE). The purpose of the current study is to model and forecast the prices of Malaysian gold called kijang emas using Box-Jenkins methodology. To find the best model, parameter estimates using OLS and MLE were computed. Based on the Akaike information criteria (AIC) and mean absolute percentage error (MAPE), the model estimated with OLS was found to perform better.

No abstract

An effective way to improve forecast accuracy is to use a hybrid model. This paper proposes a hybrid model of linear autoregressive moving average (ARIMA) and non-linear GJR-GARCH model also known as TARCH in modeling and forecasting Malaysian gold. The goodness of fit of the model is measured using Akaike information criteria (AIC) while the forecasting performance is assessed using mean absolute percentage error (MAPE), bias proportion, variance proportion and covariance proportion.