Analisis Peramalan Harga Saham Bank Rakyat Indonesia (BRI) menggunakan Autoregressive Integrated Moving Average (ARIMA)

Abstract

The uncertainty in stock price movements makes it difficult for investors to make the right investment decisions. One of the leading stocks in Indonesia is Bank Rakyat Indonesia (BRI), whose price movements are very dynamic and influenced by various factors. This research aims to build a BRI stock price forecasting model using the ARIMA (Autoregressive Integrated Moving Average) time series statistical approach, as well as evaluate how accurate the model is in providing predictions. The data used is the daily closing share price of BRI during the period from January 2 to May 20, 2025. The analysis method includes stationarity testing using the Augmented Dickey-Fuller (ADF) test, model identification based on ACF and PACF patterns, parameter estimation through Maximum Likelihood Estimation, and model accuracy evaluation using MAE, RMSE, and MAPE. The results of the analysis showed that the ARIMA model (0,1,2) was the best model with a MAPE value of 2.16%, which was included in the category of being very accurate. The forecast was made for the next 30 days, and the results showed that BRI's share price tends to be stable with a gradual upward trend, without extreme fluctuations. These findings suggest that the ARIMA model is effective in capturing short-term patterns in stocks with relatively consistent movements. Nonetheless, this model has its limitations because it relies only on historical price data without considering fundamental or macroeconomic factors. Therefore, for future research, it is recommended to consider external variables so that the forecasting results become more comprehensive. Practically, the results of this research can be used by investors in developing short-term investment strategies based on historical data and statistical analysis.