Review of Fuel Price Forecasting

In partnership with Heriot Watt University, we investigated the suitability of using a customised copula to capture the interdependency between three correlated fossil fuel prices (coal, gas, oil). The copula is used by DECC to generate correlated samples around the central fossil fuel price projections for use in Monte-Carlo modelling to understand the uncertainty around fuel price projections based on historic data.

The project required modelling of copulas in R and enabled the client to understand the underlying cause of the observed phenomena and to reach conclusions over the best procedures to be used for dealing with uncertainty in DECC’s 2015 Energy and Emissions Projections. The study provided a customised copula to capture the interdependency between fossil fuel prices. The copula will be used by the UK Department of Energy and Climate Change (DECC) to generate correlated samples around the central fossil fuel price projections for use in Monte-Carlo modelling to understand the uncertainty around fuel price projections based on historic data. DECC subsequently commissioned a further study into the suitability of Monte Carlo sampling approaches.