ARIMA-GARCH Model Price Forecasting in PT. Unilever Indonesia Tbk

Abstract

The purpose of this research is to predict the closing stock price of PT. Unilever Indonesia Tbk using the ARIMA-GARCH method. The data used in this study covers the period from February 20, 2020, to February 17, 2023, consisting of 734 daily observations. The data processing is performed using E-Views software. The closing stock price data of PT. Unilever Indonesia Tbk is non-stationary, thus requiring natural logarithm transformation and differencing. This is followed by model identification, parameter estimation, and diagnostic checking. The best-selected ARIMA model is ARIMA ([3,9],1,0), which accounts for the presence of heteroscedasticity. Subsequently, the GARCH method is applied, including model identification, parameter estimation, and diagnostic checking. The best GARCH model is GARCH (1,1), with the mean equation = 0,000047 + 0,230821 + 0,681968σ , which is free from heteroscedasticity effect. The forecast using the ARIMA ([3,9],1,0) GARCH (1,1) model yields a Mean Absolute Percentage Error (MAPE) of 2.423%, indicating a close approximation to the actual data. From the results of this research, the best model for forecasting PT. Unilever Indonesia Tbk for the next period was obtained. Therefore, the findings can assist PT. Unilever Indonesia Tbk and prospective investors in making decisions regarding the sale and purchase of shares in PT. Unilever Indonesia Tbk.

Keywords

• stock price
• arima model
• garch model