Modeling and forecasting cointegrated relationships among heavy oil and product prices
30.11.2005
Alessandro Lanza (Eni,ย Fondazione Eni Enrico Mattei,ย CRENoS), Matteo Manera (Department of Statistics, University of Milan-Bicocca,ย Fondazione Eni Enrico Mattei), Massimo Giovannini (Department of Economics, Boston College)
C22, D40, E32
Heavy oil prices, Product prices, Error correction models, Forecasting
Science Direct
"Energy Economics", Volume 27, Issue 6, November 2005
In this paper we investigate heavy crude oil and product price dynamics. We present a comparison among ten prices series of heavy crude oils and fourteen price series of petroleum products in two distinct areas (Europe and Americas) over the period 1994โ2002. We provide a complete analysis of crude oil and product price dynamics using cointegration and error correction models (ECM). Subsequently we use the ECM specification to predict crude oil prices over the horizon January 2002โJune 2002. Finally we compare the forecasting performance of ECM with a naรฏve model in first differences which does not exploit any cointegrating relation.