Extracting long-run information from energy prices: The role of exogeneity

Abstract

This article considers cointegration analysis to detect key features of long-run structure in the gasoline market. The main purpose of this study is to investigate possible long-run price leadership in the US gasoline market and the characteristics relevant to a competitive market using the vector autoregressive model. After examining the stationarity and cointegration of the weekly gasoline prices in eight different regions of the US we research long-run price leadership and parallel pricing in the framework of the cointegrated vector auto-regression (VAR). The problem is considered on extended data on 901 weekly gasoline prices for eight regions of the US. The discovery of a single common trend has been observed for a smaller number of regions, but when the system is estimated across the US it is found that the single common trend cannot be sustained. In addition to this failure the tests of exogeneity suggest that the extent to which regional gasoline prices in the long-run respond to each other is limited.