Commodity price forecasting has become increasingly important because commodity markets play a central role in economic activity, including trade, industrial planning, and investment decisions. Despite this importance, forecasting commodity prices remains challenging because prices are highly volatile and are influenced by the complex interaction of many factors, such as demand and supply conditions, macroeconomic shifts, geopolitical issues, and environmental factors. The literature offers a wide range of approaches to commodity price forecasting, from classical time series models to machine learning, deep learning, and hybrid methods. This study conducts a systematic review of the literature to identify the leading commodity price forecasting models and to map the field’s main trends. It covers key commodity categories such as crude oil, natural gas, various metals, agricultural commodities, and lumber. As a contribution, this review also proposes a taxonomy that organizes the main approaches used in the field. This review provides a clear picture of current directions in the field and offers guidance for future work in commodity price forecasting.

Keywords

Commodity price forecasting, systematic literature review, machine learning, deep learning, hybrid models.