Pineapple Trade Projections: Unveiling Future Trends in Philippine Exports with Advanced Time Series Analysis

This study examines data from the Philippine Statistics Authority on pineapple export trade in the Philippines from 2000 to 2022, aiming to project exports from 2023 to 2032. Statistical measures, including Kendall’s tau_b and Spearman’s rho, validate trend strength and robustness. Pre-processing involved outlier detection using the Z-score method and calculating stationarity through first-order differencing. The Ljung-Box test ensured model robustness by checking for autocorrelation in residuals. Various ARIMA models (ARIMA(0,1,1), ARIMA(1,1,0), ARIMA(1,1,1)), Exponential Smoothing (Brown, Damped, Holt) and Moving Average were evaluated to verify on which model best fit for forecasting. Based from the three models, the results provide evidence that across multiple metrics, the 2-year Moving Average method is the best fit for forecasting. It achieved the highest Stationary R² (0.887) and R² (0.887), and the lowest RMSE (71.589) and MAPE (6%). Notably, the moving average forecast for 2023 was 992.22 metric tons, which closely approximates the actual export figure of 939.98 metric tons. In contrast, the forecasts generated by the ARIMA model and Exponential Smoothing were significantly lower, at 470.66 metric tons and 390.88 metric tons, respectively. This indicates the superior accuracy and reliability of the Moving Average method in capturing the trends in the data. These findings aid in understanding export trade and inform policy development for food security and sustainability, and improve the pineapple export strategy for the Philippines moving forward.