Energy markets volatility modelling using GARCH

- We investigate the empirical properties of crude oil, natural gas, and electricity price volatilities using a range of univariate and multivariate GARCH models
- We estimate a trivariate VARMA-BEKK model that allows for spillovers and interactions among energy markets
- We evaluate and compare the performance of univariate and multivariate models with a range of diagnostic and forecast performance tests, and present sample forecasts

This paper investigates the empirical properties of oil, natural gas, and electricity price volatilities using a range of univariate and multivariate GARCH models and daily data from wholesale markets in the United States for the period from 2001 to 2013. The key contribution to the literature is the estimation of trivariate BEKK and DCC models that allow us to observe spillovers and interactions among energy markets. We evaluate and compare the performance of univariate and multivariate models with a range of diagnostic and forecast performance tests, and assess forecasting performance and conditional correlation dynamics.