Essays on the empirical analysis of energy risk
Energy markets have become increasingly sophisticated, requiring modelling techniques of analogous calibre. This thesis deals with models of changing regime for the petroleum complex. Modelling the conditional distribution of energy prices as a regime switching process is motivated by the market-specific characteristics of oil: different market conditions, such as backwardation and contango, involve different dynamics. The first empirical part examines the very short-end of the futures curve volatility. To address in a realistic way the potential diverse response of oil volatility to fundamentals across high and low volatility regimes, augmented regime volatility models are employed. Results indicate that volatility can be decomposed to a highly persistent conditional volatility process and a relatively short-lived non-stationary process. Apart from evaluating the size of price risk, risk managers must also design a framework for mitigating their exposures. This is the focus of the second empirical part which estimates dynamic hedge ratios. Linking the concept of disequilibrium with that of uncertainty across high and low volatility regimes, a state-dependent error correction model with timevarying second moments is introduced. Finally, the third empirical part, examines the information content of the dependence structure between correlated petroleum futures curves. Term structure is decomposed into level, slope and curvature shocks. Introducing a multiregime framework, these factors are utilised to study inter-commodity and inter-market spreads. Results suggest markedly different state-dependent speeds of mean reversion and volatility/correlation dynamics across regimes. Overall, the employed models provide superior forecasting performance and indicate that state-dependent dynamics may provide significant benefits to market participants. The findings of this thesis have important implications for energy market trading and risk management, as well as energy market operations, such as refining and budget planning, by providing valuable information on the oil price volatility dynamics and the ability to predict risk.