Strategic alignment of pricing oil products from Brent conditionally on market regimes using Markov and cluster-based switching dynamics

Asymmetries and non-linearities in oil prices are of high importance for energy policy and strategy. The effectiveness of energy policy depends heavily on the ‘conceptualization’ of states, regimes and less on the absolute price level. Based on the strategic alignment theory and the matching perspective, we claim that recurring oil market dynamics showcase the existence of various regimes across which pricing relationships differ. Regimes are examined by combining two methodologies: Cluster-based Regression (CBR) and Markov-Switching Dynamic Regression (MSDR). CBR identifies spatial regimes by clustering Brent as a proxy variable, while MSDR identifies temporal regimes by presuming a probabilistic superposition of states which transition as a discrete time Markov chain. Results have both policy and methodology contributions. A reversion and a cyclicality phenomenon in pricing, resembles the Joseph and Noah effects in terms of persistence, expected duration and probability of transition. CBR describes a serial process in pricing from ‘bull’ to ‘bear’ and then to ‘transitory’ state. Differences in asymmetry and high price premiums, conditionally on market states, are detected among products. Matching probabilities show robustness of results.