Fuel prices scenario generation based on a multivariate GARCH model for risk analysis in a wholesale electricity market

Request a copy

Abstract

This paper presents a fuel prices scenario generator in the frame of a simulation tool developed to support risk analysis in a competitive electricity environment. The tool feeds different exogenous risk factors to a wholesale electricity market model to perform a statistical analysis of the results. As the different fuel series that are studied, such as the oil or gas ones, present stochastic volatility and strong correlation among them, a multivariate Generalized Autoregressive Conditional Heteroskedastic model has been designed in order to allow the generation of future fuel prices paths. The model makes use of a decomposition method to simplify the consideration of the multidimensional conditional covariance. An example of its application with real data is also presented.
<p>This paper presents a fuel prices scenario generator in the frame of a simulation tool developed to support risk analysis in a competitive electricity environment. The tool feeds different exogenous risk factors to a wholesale electricity market model to perform a statistical analysis of the results. As the different fuel series that are studied, such as the oil or gas ones, present stochastic volatility and strong correlation among them, a multivariate Generalized Autoregressive Conditional Heteroskedastic model has been designed in order to allow the generation of future fuel prices paths. The model makes use of a decomposition method to simplify the consideration of the multidimensional conditional covariance. An example of its application with real data is also presented.</p>
<p>This paper presents a fuel prices scenario generator in the frame of a simulation tool developed to support risk analysis in a competitive electricity environment. The tool feeds different exogenous risk factors to a wholesale electricity market model to perform a statistical analysis of the results. As the different fuel series that are studied, such as the oil or gas ones, present stochastic volatility and strong correlation among them, a multivariate Generalized Autoregressive Conditional Heteroskedastic model has been designed in order to allow the generation of future fuel prices paths. The model makes use of a decomposition method to simplify the consideration of the multidimensional conditional covariance. An example of its application with real data is also presented.</p>
Ítem

Información detallada

Materias, derechos, colecciones e identificadores

Keywords

Instituto de Investigación Tecnológica (IIT), Fuels; Monte Carlo methods; Power system modeling; Risk analysis; Stochastic processes

Impact metrics