Forecast of global steel prices

Abstract

304Economic forecasts rarely come true with a high degree of accuracy, however, this does not lead to the eradication of an interest in such forecasts. We can see an analogy with weather forecasts, which are frequently incorrect, but they do not express such a strong devia tion from the real weather as other forecast methods. Economic forecasts restrict the future ambiguity and thus promote more confident long term development strategies and investments. If one is to choose between a forecast and its absence, the former is explicitly prefer ential because it allows for the reduction of losses from probable wrong decisions, taking advantage of opening opportunities or escaping potential threats.Long term price forecasts have a special signifi cance. Long term investment decision making in any branch, including ferrous metallurgy, is based on the understanding of price trends during the implementa tion and pay off phases of a project. In the ferrous met allurgy industry, the time slot can exceed 10 years. It is vital that the trends of final product prices be compara ble with the expected cost dynamics. It allows for more reliable forecasts of producers' margins and, conse quently. of investment payback periods.The object of the analysis in this article are long term price tendencies in the global rolled iron market. The purpose was to propose a forecast procedure for world steel prices that would include compatible trends of steel prices and costs.As a result, we have developed a medium term fore cast algorithm for the average world steel price, which is based on a regression between metal roll prices, global steel capacities and the dynamics of steel production costs. The derived regression model provides a good description of the historical dynamics of world steel prices since 1980, including the drop in prices during the crisis of 2009. The simplicity of the model allows for use in the calculation of price targets for strategic busi ness plans in metallurgical companies.Approaches to price forecasting in the steel industry. The lack of popular intuitive prediction methods among practitioners in the metallurgical industry [1] results in a situation when experts frequently substitute forecasts by wishful thinking, i.e., instead of forecasting what is going to be, they express their anticipations. In addi tion, a considerable amount of time is spent on the coordination of different trends (the growth rate of steel consumption, steel prices, raw material prices, etc.). Formal methods allow for the elimination of the above mentioned subjective judgement and additional labor costs through the identification of regularities in the object to be forecast through the construction of eco nomic models.In the process of price forecasting (in virtually every industry), we can distinguish the following approaches to model construction:(1) decomposition of price time series: detection of the trend, autocorrelation, seasonal and random com ponents;(2) factor analysis based on a regression model: allo cation of essential price f...

Econometric modeling of recycled copper supply

“…Given that values of MI are normalized to a scale of 100, +1% change in MI would lead to about +1.1% change in refined copper scrap supply. Previous work has shown that steel is price inelastic and the demand elasticity is in the range of −0.2 to −0.3 (Malanichev and Vorobyev, 2011). Therefore, based on the estimated price elasticity, we argue that refined copper scrap supply is price inelastic, and its main drivers are world GDP and level of mining activity.…”

Development of a Group Method of Data Handling Technique to Forecast Iron Ore Price

“…The design of another effective parameter (neuron number) on the GMDH model needs another parametric study. To do this, considering the previous studies (e.g., [21]), values of 2,4,6,8,10,12,14,16,18, and 20 were selected to be used as number of neurons in the parametric study and results of GMDH models based on R 2 are shown in Table 4. Among the obtained results, as it can be seen in Table 4, GMDH model number 8 with 16 neurons shows the best prediction performance and due to that neuron number of 16 was selected in the rest of modelling process.…”

“…A major share of distributed iron ore is produced by the three companies of Vale (a semi-private company) (Rio de Janeiro, Brazil ), Rio Tinto(London, UK), and B.H.P Billiton(Melbourne, Australia). Because of the fact that each ton of crude steel is made by using 1.5 ton of iron ore, production of crude steel can be introduced as a demand agent and because provision of iron ore is almost exclusive, producers provide the market with an output totally in association with instant demand [10].…”

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

“…The global steel forecast model was tried by Malanichev and Vorobyev (2011) with multivariate regression using raw material costs and capacity utilization. In the research by Kapl and Müller (2010), the authors compared the efficiency of the autoregressive integrated moving averages (ARIMA) model with covariates versus multi-channel singular spectrum analysis (M-SSA).…”