Forecasting of the U.S. Steel Prices with LVAR and VEC Models

Özet

Base metal prices, especially steel, play a significant role in industrialeconomics, making them worth knowing about future values. In most cases, we expectsuperior performance from multivariate forecasting models comparing univariatemethods due to the involvement of explanatory variables in the system. Standard vectorauto regressive model can only capture short-run dynamics because of the differencingprocess for non-stationary series that eliminates the possible long-run relationship.Instead, performing non-stationary series on levels through the vector auto-regressiveframework does not suffers such loss. Moreover, the vector error correction model candefine both short-term and long-run dynamics explicitly. These models can yield morerobust forecasts in the mid-term and long-term by investigating short-run and long-runrelationships simultaneously. The current study aims to perform an out-of-sampleforecast for the United States steel prices index 18 months ahead using cointegratedvariables. The results suggest that the non-stationary vector auto-regressive modeloutperforms the vector error correction model regarding mean absolute percentageerror and root mean square error as forecast accuracy measures.