Forecasting oil and gas Prices: A Model Combination Approach
Working paper
Permanent lenke
https://hdl.handle.net/11250/3173264Utgivelsesdato
2024-12-31Metadata
Vis full innførselSamlinger
Sammendrag
This paper advances energy price forecasting by extending state-of-the-art econometric models for oil markets to address the distinct characteristics of fragmented natural gas markets. Using a forecast combination framework, we integrate diverse modeling approaches with various global economic indicators. Our findings show that forecast combinations consistently outperform traditional benchmarks, with constant equal weighting providing the most accurate oil price forecasts (Brent, WTI, and U.S. refiners’ acquisition cost) and performance-based weighting delivering superior results for natural gas prices (Henry Hub spot price). These results underscore the importance of developing tailored forecasting methodologies that account for the heterogeneous dynamics of energy commodities, thereby offering a robust framework for advancing predictive accuracy across a range of horizons.