Gasoline price predictability in a border metropolitan economy

This study examines the predictability of local retail gasoline prices in the El Paso metropolitan economy. Given its location on the border with Mexico, the potential influence of cross-border economic variables on gasoline prices in El Paso is taken into account. The study uses monthly frequency time series data from 2001 to 2013. Because historical consumption data are not available, the error correction econometric model employs a reduced form equation in which gasoline prices are functionally dependent on several explanatory variables. Out-of-sample price simulations are compared against random walk and random walk with drift benchmarks. Results obtained indicate that the econometric approach performs fairly well relative to both benchmarks.