Diesel oil consumption has rapidly grown in Europe during the last decades up to 202 million tons in 2008. This increase is mainly attributed to the growing transport needs of trucks, and to the switch between gasoline and diesel cars. Nowadays most of the passenger car sales are diesel cars. This increasing share of diesel engines for passenger cars could be associated with the evolution of the household income, the spread between gasoline and gas oil prices, and with the performance of the diesel engine with regard to fuel consumption. In this context, we have built an econometric cointegration model in order to establish a long-term equilibrium between the diesel car share and the GDP, the price spread and the engine fuel consumption in France. We use quarterly data from 1985-Q1 to 2008-Q3. The classical Chow test and cusum tests point out several structural break dates over the sample. When we carry out the unit root test with structural break, some of these dates are significant. Consequently, we apply the Bai and Perron (1998, 2003) procedures to test for multiple break points and the recent Kerjrival and Perron (2008) developments for cointegration models. We compare the different test procedures and establish a long-term equilibrium for the diesel car share with a decreasing impact of the price spread and an increasing effect engine performance. We analyze the consequences of these structural breaks on the long term energy demand forecast in the case of diesel oil.
INDJEHAGOPIAN, J.P., LANTZ, F. et SENTENAC-CHEMIN, E. (2011). Multiple Structural Break Change and Energy Forecasts: Application to Diesel Cars Equipment in France. Dans: 31st International Symposium on Forecasting.