Detecting speculative bubbles in metal prices: Evidence from GSADF test and machine learning approaches


Resources Policy, vol.74, 2021 (SSCI) identifier

  • Publication Type: Article / Article
  • Volume: 74
  • Publication Date: 2021
  • Doi Number: 10.1016/j.resourpol.2021.102306
  • Journal Name: Resources Policy
  • Journal Indexes: Social Sciences Citation Index (SSCI), Scopus, International Bibliography of Social Sciences, Aerospace Database, Aquatic Science & Fisheries Abstracts (ASFA), Business Source Elite, Business Source Premier, Communication Abstracts, EconLit, Index Islamicus, INSPEC, Metadex, PAIS International, Pollution Abstracts, Public Affairs Index, Civil Engineering Abstracts
  • Keywords: Macroeconomic factors, Metal prices, Multiple bubbles, Random forest algorithm
  • Çanakkale Onsekiz Mart University Affiliated: Yes


© 2021 Elsevier LtdThe importance of metal prices to real economic activity and financial markets has increased the focus on detecting price bubbles in precious and industrial metals. Several studies looked at the influence of macroeconomic factors in the formation of a single metal bubble and tried to identify bubble dates. Our study extends the literature and analyzes monthly gold, platinum, palladium, rhodium, silver, and aluminum, copper, lead, nickel, steel, tin prices over 1980M1-2019M12, and contributes to the literature in two ways: First, the analysis incorporates the Generalized Supremum Augmented Dickey-Fuller (GSADF) test to detect potential bubbles. Second, the study evaluates the impact of potential financial, real, and speculative factors in the likelihood of price bubbles using the random forest method. Our findings indicate that financial factors are more critical in predicting precious metal price bubbles. The monetary policy rate and the production index are important to predict bubbles in industrial metal prices. However, our findings suggest that speculative activity may not adequately predict metal price bubbles.