Prediksi Harga Nikel Menggunakan Model Autoregressive Integrated Moving Average with Exogenous (ARIMAX) Dengan Aplikasi R Studio

Abstract

Nickel is one of the strategic commodities whose price movements are influenced by various global economic factors. Accurate nickel price forecasting is important for investors, industries, and policymakers in making decisions related to investment and risk management. This study aims to model and forecast nickel prices using the Autoregressive Integrated Moving Average with Exogenous Variables (ARIMAX) method by incorporating the Consumer Price Index (CPI) and the Federal Reserve interest rate as exogenous variables. The data used in this study consist of monthly nickel prices, CPI, and Federal Reserve interest rates during the 2020–2023 period obtained from Investing.com. The analysis was conducted using R Studio software. The modeling process included stationarity testing, parameter identification, parameter estimation, diagnostic testing, and forecasting evaluation. The results indicate that the ARIMAX model is capable of capturing nickel price movements influenced by external macroeconomic factors. Furthermore, the model provides reasonably accurate forecasting performance based on forecasting error measurements. Therefore, the ARIMAX method can be considered an effective approach for forecasting nickel prices using macroeconomic variables as external factors. The study found that the best ARIMAX model for forecasting nickel prices from January 2024 to December 2025 was ARIMAX(0,1,0), with the lowest AIC value of 881.13 after first-order differencing. The forecast indicates a slight decline in nickel prices from USD 16,701.34/ton in January 2024 to USD 16,647.16/ton in December 2025, influenced by the Consumer Price Index (CPI) and the Federal Reserve Interest Rate Decision. The model achieved a MAPE of 6.68%, indicating high forecasting accuracy and demonstrating that ARIMAX(0,1,0) is reliable for future nickel price projections and decision-making.