The dynamics of crude oil price differentials

Abstract

Crude oil price differentials are modelled as a two-regime threshold autoregressive (TAR) process using the method proposed by Caner and Hansen [Caner, M., Hansen, B.E. Threshold autoregression with a unit root. Econometrica 2001; 69; 1555–1596.]. While standard unit root tests suggest that the prices of crude oil of different varieties move closely together such that their price differential is stationary, the TAR results indicate strong evidence of threshold effects in the adjustment process to the long-run equilibrium. These findings suggest that crude oil prices are linked and thus at the very general level, the oil market is ‘one great pool’ (Adelman, M.A. International oil agreements. The Energy Journal 1984; 5; 1–9.). However, differences in the dynamics of adjustment suggest that within this one pool, oil markets are not necessarily integrated in every time period and hence the dynamics of crude oil price differentials may not follow a stationary process at all times. Although the development of a liquid futures market around the crude oil benchmarks has helped make some distant markets more unified, arbitrage is not costless or risk-free and temporary breakdowns in the benchmarks can lead to decoupling of crude oil prices.