Distributional modeling and forecasting of natural gas prices

Abstract

We examine the problem of modeling and forecasting European day‐ahead and month‐ahead natural gas prices. For this, we propose two distinct probabilistic models that can be utilized in risk and portfolio management. We use daily pricing data ranging from 2011 to 2020. Extensive descriptive data analysis shows that both time series feature heavy tails and conditional heteroscedasticity and show asymmetric behavior in their differences. We propose state‐space time series models under skewed, heavy‐tailed distributions to capture all stylized facts of the data. They include the impact of autocorrelation, seasonality, risk premia, temperature, storage levels, the price of European Emission Allowances, and related fuel prices of oil, coal, and electricity. We provide rigorous model diagnostics and interpret all model components in detail. Additionally, we conduct a probabilistic forecasting study with significance tests and compare the predictive performance against literature benchmarks. The proposed day‐ahead (month‐ahead) model leads to a 13% (9%) reduction in out‐of‐sample continuous ranked probability score (CRPS) compared with the best performing benchmark model, mainly due to adequate modeling of the volatility and heavy tails.

An Alternative to Index-Based Gas Sourcing Using Neural Networks

“…Alternatively, instead of applying a neural network for forecasting purposes, one might opt for a more classic approach. Berrisch and Ziel [27] proposed a state-space model which is especially focused on the distributional properties. Note that this range of models is only a small excerpt of possibilities suggested by the literature.…”

Energy markets restructure beyond 2022 and its implications on Qatar LNG sales strategy: Business forecasting and trend analysis

“…A concrete understanding of the pricing systems adds to building a robust portfolio to maximise profitability and/or minimise risks. Time series analysis has been a widely used empirical approach to understand the factors driving price behaviour [ [104] , [105] , [106] ], the interconnection between the pricing systems [ [107] , [108] , [109] ], and to predict future values [ [110] , [111] , [112] ]. Overall, investigating the pricing systems for reshaping policies or liberalising local pricing hubs from the perspective of importers has been a dominant theme in former studies.…”

Forecasting Natural Gas Prices with Spatio-Temporal Copula-based Time Series Models

“…The Autoregressive moving average model with external regressors and absolute value, threshhold generalized autoregressive conditional heteroskedasticity model, henceforth denoted by ARMAX-AVTGARCH (closely related to the model from [6]).…”

“…The models performances are examined in a forecasting study. A model closely related to the model from [6], which was shown to outperform other popular models, is considered as benchmark. We find that the spatio-temporal copula time series modeling with ANN-augmented point forecasts are competitive for natural gas and related commoditiy prices forecasting.…”