TIME SERIES FORECASTING OF LQ45 STOCK INDEX USING ARIMA: INSIGHTS AND IMPLICATIONS

Abstract

This study aims to analyze and forecast the movement of the LQ45 stock index in the Indonesian stock market using the ARIMA (AutoRegressive Integrated Moving Average) model. Through rigorous time series analysis and model selection procedures in SPSS, the ARIMA (4,3,6) model was identified as the most optimal to capture the complex dynamics of the LQ45 index. The model's parameters indicate that the index exhibits a high degree of integration (third differencing) and is influenced by multiple autoregressive and moving average terms. Forecasting results demonstrate that ARIMA (4,3,6) provides accurate short- to medium-term predictions, with a good balance of fit and error metrics. These findings have practical implications for investors seeking data-driven strategies for portfolio management and risk mitigation. Furthermore, the study contributes to the academic field by offering a robust modeling approach for stock index forecasting, encouraging further research to integrate ARIMA with other advanced models and external economic factors for enhanced predictive performance.