Trade Forecasting: Classical and Artificial Neural Network Methods
Abstract
With recent technological advancements, production capacities have increased significantly, contributing positively to the economic development of both countries and firms. Foreign trade plays a crucial role in achieving sustainable economic growth, as it enhances employment, competitiveness, foreign exchange inflows, and value-added production. Therefore, the export and import performance of countries is of critical importance. This study aims to generate forecasts for export and import volumes using Türkiye’s monthly foreign trade data and to comparatively evaluate the accuracy of different forecasting methods. In this context, monthly data covering the period from January 2013 to September 2024 for 13 countries with which Türkiye conducts trade were used to produce 27-month forecasts for the period from October 2024 to December 2026. In addition, data for the period from October 2023 to September 2024 were excluded from the analysis and used to generate new 12-month out-of-sample forecasts. Forecasts were compared with actual values, and the performance of the methods was evaluated using the Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE) metrics. The analysis employed ARIMA, ETS, Auto DeepDENT, and Simple Forecast Combination methods. The results indicate that forecast error values differ across countries for both imports and exports; however, in terms of overall performance, the ARIMA model was found to provide more accurate and reliable forecasts compared to the other methods.
Keywords
Ethical Statement
References
- Alam, T. (2019). Forecasting exports and imports through artificial neural network and autoregressive integrated moving average. Decision Science Letters, 8(3), 249-260.
- Altınay, M. (2024). İzmir dış ticaretinin gelişimi ve yoğunlaşma analizi (2013-2023). Bölgesel Kalkınma Dergisi, 2(3), 235-263, DOI: 10.61138/bolgeselkalkinmadergisi.1576229.
- Bates, J. M., & Granger, C. W. J. (1969). The combination of forecasts. Operational Research Quarterly, 20(4), 451-468.
- Bayır, F. (2006). Yapay sinir ağları ve tahmin modellemesi üzerine bir uygulama (Yayınlanmamış yüksek lisans tezi). İstanbul Üniversitesi.
- Box, G. E. P., & Jenkins, G. M. (1970). Time series analysis: Forecasting and control. San Francisco: Holden-Day Inc.
- Co, H. C., & Boosarawongse, R. (2007). Forecasting Thailand’s rice export: Statistical techniques vs. artificial neural networks. Computers & Industrial Engineering, 53(4), 610-627.
- DTM (Dış Ticaret Müsteşarlığı). (2007). Avrupa Birliği ve Türkiye (6. baskı). Dış Ticaret Müsteşarlığı Avrupa Birliği Genel Müdürlüğü Yayını.
- Eğrioğlu, E., & Baş, E. (2024). A new deep neural network for forecasting: Deep dendritic artificial neural network. Artificial Intelligence Review, 57(7), 171.
Details
Primary Language
English
Subjects
Deep Learning, Neural Networks, Time-Series Analysis, Soft Computing, Applied Statistics
Journal Section
Research Article
Publication Date
March 12, 2026
Submission Date
January 28, 2026
Acceptance Date
February 10, 2026
Published in Issue
Year 2026 Volume: 10 Number: 1