Forecasting the export value of indonesian tuna, skipjack, and mackerel tuna using Autoregressive Integrated Moving Average (ARIMA)

Abstract

This study aims to build a statistically valid forecasting model for the export value of Tuna, Skipjack, and Mackerel Tuna in Indonesia, considering that the ability to project future export performance is crucial for the strategic planning of both government and industry. Using annual time-series data from the period 2005 to 2023 sourced from the Central Bureau of Statistics (BPS) and the Ministry of Marine Affairs and Fisheries (KKP), this study applies the Autoregressive Integrated Moving Average (ARIMA) Box-Jenkins methodology. The entire analysis process, from stationary testing to forecasting, was conducted using Minitab 22 software. After a series of rigorous identification, estimation, and diagnostic checking stages, the ARIMA (0,1,2) model was identified as the best fit. This selection was based on its superiority across various criteria: having the lowest Akaike's Information Criterion (AIC) and Bayesian Information Criterion (BIC) values, the best out-of-sample forecasting accuracy (MAPE = 13.15%), a significant MA(2) parameter (p-value = 0.020), and valid residuals based on the Ljung-Box test (p-value = 0.663). The forecast for the 2024-2026 period shows a sustained upward trend in export value, with a projection of reaching USD 991.65 million by 2026. These findings provide a quantitative basis for the government in setting export targets and for the industry in planning investments, particularly in the development of value-added products to meet global market demand.