Análise temporal dos preços do carvão vegetal, no estado de Minas Gerais

Abstract

Energy is essential to human needs satisfaction. After the consolidation of the machinery industry, man started to depend more and more on the energy .stocked in fóssil fuels, comparatively to the primitive economy. Charcoal is a term-reducer used in the Brazilian pig iron and steel industries. The State of Minas Gerais is the largest producer and consumer of this input. It can be produced from planted forest or from native forest. These two types ofcharcoal possess some technical and economical characteristics thatcan differentiate its prices. The prices of the cubic meter ofcharcoal are formed in an oligopsonic market. Along the time, charcoal prices vary in function of endogenous and exogenous factors, needing, therefore, to be deflated before comparing them. This study analysis the longitudinal series of charcoal prices in Minas Gerais State, their effects on the Brazilian forest politics, starting from the decade of 70s, and their relevant aspects in the metallurgical sector. In particular, it intended: to analyze the longitudinal series of charcoal prices in the State of Minas Gerais, in four consuming regions, in the period 1981-2003; charcoal prices ofnative forest and ofplanted forest, in the period from 1999 to 2003, and; to analyze and to characterize the evolution of the real charcoal prices, to compare these prices with the real petroleum prices, in the period from 1975 to 2002 and the effects of the economical plans and of oil crisis. The average charcoal prices used were those published byABRACAVE. Foranalysis of the series, it was considered the model SARIMA, with the objective offinding the parameters to supply better forecasts for charcoal prices. It was considered the period from January 1975 to December of2002 to adjust the model and the period from January to December of 2003 for the validation of the model forecasts. For the four regions the period from January 1981 to December 2003 was used. Forplanted forest charcoal prices it wasusedtheperiodfromJanuary 1999 to April 2003. For the analysis ofthe series ofprices the model SARIMA was used, with the objective of finding a model that supplies better price forecasts for the four studiedregions. Themostappropriate models were chosen through the graphical analyses of the standardized residues, autocorrelation functions andpartial autocorrelation, stochastic tests andcríteria ofevaluation of the order ofthe model. The analyses ofthe evolution ofcharcoal prices were accomplished in the period from jan/1975 the dez/2002. The prices series of charcoal prices in Brazilian currency were deflated by IGP-DI, base ago/1994=100. The prices were also converted to American dollar and deflated by CPI, base 1982-84=100. Then, they were compared in nominal and real terms. Among the studied models, SARIMA (2,0,1) (0,l,l)i2, for Minas Gerais provided better adjustment and in a parsimonious way. In spite ofhappening a random factor that distorted the real price figures charcoal prices, the projections supplied by the model were inside ofthe established limits ofsignificance. The differences ofcharcoal prices for the four regions seems to be basically related to location; the analyses ofdomain ofthe time and the domain ofthe frequency pointed out that there is difference in the series ofprices for the four studied regions. The áreasofSete Lagoasand Belo Horizonte possess similar prices and generated similar model, probably due to the fact that they are very close; the studied regions presented differentiated models, which supplied good adjustments for the observed series. The best models were SARIMA (l,l,2)x(2,l,0)12, for Belo Horizonte; SARIMA (l,0,l)x(2,l,l)12, for Divinópolis; SARIMA (2,0,0)x(2,l,l)12, for Sete Lagoas and SARIMA (l,0,l)x(2,l,l)i2, for Vertentes. Such models presented, in a parsimonious way, containing a small number ofparameters. Ali ofthe models SARIMA (p,d,q) (P,D,Q)s, for the four studied regions, presented white noise and the price forecast supplied were satisfactory. Among the adjusted models for plantation forest charcoal, SARIMA (2,l,2)(0,l,0)i2 provided better adjustment, while for the native forest charcoal price, SARIMA (0,l,2)(0,l,0)i2 provided better adjustment, demonstrating differences in the tendency among the series. The evolution ofthe real charcoal pricesin domestic currency, or in American dollar, presented a decreasing tendency along time. The inflationary disarray, in the decades of80 until the middle ofthe decade 90, caused a large price variation in the period; from the beginning ofthe XXI century, charcoal prices were more influenced by exchange rate; in the periods of energy crísis (petroleum), charcoal prices underwent several effects that, however, did not sustained along time.