FORECASTING EXPORTS IN ALBANIA

Abstract

Time series forecasting plays a crucial role in various fields, including economics, finance, and weather prediction. This study explores the application of time series forecasting techniques to predict future values based on historical data patterns. Different models, such as ARIMA, SARIMA, and exponential smoothing, are commonly employed to capture the underlying trends, seasonality, and irregularities in time series data. The accuracy of the forecasts depends on selecting the appropriate model and fine-tuning its parameters. Regular validation and refinement of the forecasting models are essential to ensure their reliability and adaptability to changing conditions. Overall, time series forecasting serves as a valuable tool for decision-making, resource allocation, and planning in numerous domains. The Autoregressive Integrated Moving Average (ARIMA) model was used in this research to predict Albanian exports. The ARIMA models provided reliable forecasts for the near future and placed more importance on recent observations rather than distant past data. However, a limitation of this study is the lack of using multiple models on new data to improve future forecasts. Considering the significance of exports in Albania, it is crucial to periodically validate and refine the model-building exercises. Based on the findings, the study recommends using the ARIMA (3,2,3) model for forecasting. Furthermore, the forecast errors were statistically tested and validated, confirming the model's strong predictive ability. It is important to note that a forecasting technique that performs well in one situation may not be suitable for another. Therefore, the validation of a specific model should be regularly assessed as time progresses. To forecast the annual exports of Albania, researchers can employ these models.