Abstract

This article aims to get model data stock Unilever Indonesia Tbk. use the model ARCH and GARCH as well as comparing forecasting accuracy of the result of the next five days ahead model ARCH and GARCH on the stock Unilever Indonesia Tbk. use MATLAB. The methods used are design application forecasting uses GUI MATLAB, next model ARIMA Box-Jenkins, identification ARCH effect, forecasting use the model ARCH and GARCH, and compares the results second forecasting model that is based on the value of RMSE. On residual ARIMA best namely ARIMA(1,1,1) detected the effects ARCH so that data can modeled ARCH and GARCH. Model ARCH and GARCH best respectively namely ARCH(3) and GARCH(1,1). Based on value RMSE be seen that model best for forecasting the next five days ahead of data Unilever Indonesia Tbk. produced bymodels GARCH(1,1) because it has value RMSE smallest with equation conditional mean and conditional variance