Pipeline capacity and the dynamics of Alberta crude oil price spreads

Abstract

AbstractWe investigate the effect of binding pipeline capacity on price spreads between crude oil in Alberta, Canada, and competing crude oils in export markets at the other end of the pipeline. Our goal is to determine whether these price spreads are consistent with capacity‐constrained spatial arbitrage models by estimating Markov switching models of the price spreads for both heavy and light crude oil. Our results for heavy crude oil agree with the theory: we find one regime with both a low mean and a low variance, consistent with sufficient pipeline capacity, and another regime with both a higher mean and a higher variance, consistent with constrained pipeline capacity. The difference in mean spreads between regimes then provides an estimate of the mean shadow price of capacity. Price spreads for light crude oil do not display significant differences in mean spreads across regimes, suggesting that pipeline capacity constraints are less important for this commodity. Comparing our results with pipeline capacity data, we find that tight capacity is not always associated with large price spreads.