<![CDATA[Stock price forecasting is important because investors need empirical information to reduce uncertainty in short-term investment decisions. This study aims to analyze the daily stock price movement of PT Bank Nationalnobu Tbk and to determine the feasibility of the ARIMA model for forecasting the stock price. The study used a quantitative time series approach based on daily closing prices from May 2, 2023, to April 30, 2026, with 712 observations. The analysis was conducted using RStudio through descriptive statistics, stationarity testing, first differencing, ACF and PACF identification, ARIMA model comparison, diagnostic testing, and forecast accuracy evaluation. The results show that the data were non-stationary at the original level because the Augmented Dickey-Fuller p-value was 0.1847. After first differencing, the p-value decreased to 0.0100, indicating stationarity. The ARIMA(0,1,2) model was selected because it had the lowest AIC value. The Ljung-Box test produced a p-value of 0.1675, indicating no residual autocorrelation. The model yielded an MAE of 29.2604, an RMSE of 38.9040, and a MAPE of 5.2949%. These results indicate that the ARIMA(0,1,2) model is sufficiently accurate for short-term forecasting of NOBU stock prices, although residual normality remains a limitation of the model.]]>
Copyrights © 2026
