Price Interactions and Predictability in the Steelmaking Industry: A Time Series Approach

The Thesis proposes a statistical in-depth analysis based on econometric models of the steelmaking industry, which is not only retained as one of the most important industrial sectors in many countries but also represents the backbone of their economic development. In particular, the main focus of the analysis concerns the long-term cointegration relationship between crude steel and steel scrap prices in light of the ever-increasing adoption of recycled materials in the manufacturing process of finished steel goods. For this purpose, a Vector Error Correction Model has been adopted to evaluate the long-run adjustment mechanism in which the variables are potentially involved. After having determined the presence and directions of the causal relationships underlying the factors of interest, the focus of the Thesis shifts toward the validation of the adopted model and the analysis of information spillovers through the calculus of the relative Impulse Response Functions and the Forecast Error Variance Decompositions. Since the scrap prices are characterized by a strong exogeneity in their relationship with primary steel prices, an ARIMAX model has been implemented to explain the short-run fluctuations of the dependent variable in relation to external factors’ influences. Some ground is also provided for a forecast analysis of future crude steel prices on the basis of the ARIMAX predictive power.