Determinants of Import Prices: An Econometric and Machine Learning-Based regularisation Approach
Year: 2026
Type: article
Abstract: This study examines the determinants of import unit prices by highlighting the importance of the price dimension in Türkiye’s nominal imports over the period 2005-2024. In this regard, the analysis focuses on economic and global factors that affect import unit prices. The objective is to contribute to the development of a more realistic and forward-looking perspective in the evaluation of foreign trade policies through the price channel. Using a broad dataset, various indicators affecting import unit prices are investigated using the commonly used OLS method and multicollinearity robust Ridge regression approach. Ridge reduces the problem of multicollinearity, preserves the direction and relative importance of the coefficients in a manner consistent with OLS, and confirms the robustness of the results. The findings imply that both methods yield consistent results in terms of variables affecting import prices. Accordingly, commodity prices and crude oil prices are the variables with the strongest and most positive impact on import prices. Furthermore, while real effective exchange rate volatility, geopolitical risk, world industrial production, inflation rate, and the baltic dry index maintain positive coefficients, trade policy uncertainty stands out as the only variable showing a negative impact on import prices. This study points to import prices as a significant determinant in the process of developing foreign trade policies. Accordingly, developing policy frameworks on reliable forecasts of the factors influencing import prices will contribute to the design of more realistic, consistent, and effective foreign trade policies.
DOI: 10.26650/jepr1953672
Authors: Zaim Reha Yaşar, Ahmet Bağcı
Institutions: Kırıkkale University, Ministry of Agriculture and Forestry
Language: English
Subfield: Economics and Econometrics
Domain: Social Sciences
Open Access status: diamond
OpenAlex ID: W7203875246
