| Abstract [eng] |
The main problem is the unstable nature of observed time series and its heteroscedasticity, making it impossible to apply certain time series models. Time series models namely, the generalized autoregressive conditional heteroscedastic (GARCH) models with different lags and distributions were fitted to the data. The stages in the model building namely, identification, estimation and checking has been explored and applied to the data. A best fitting model was selected based on how well the model captures the stochastic variation in the data (goodness of fit). The goodness of fit is assessed through the Akaike information criteria (AIC), the Bayesian information criteria (BIC), the mean absolute error (MAE) and mean square error (MSE). Based on minimum AIC, BIC, MAE and MSE values, the best fit GARCH model tend to be GARCH(1,1). Having satisfied with all the model assumptions, GARCH(1,1) was judged to be the best model for forecasting. Based on the selected model, we forecasted ten (10) days of the conditional volatility out – of – sample period (that is from 02 January 2014 to 15 December2014). From the results, it has been observed that the forecasted series are close to the actual series. |
