Electronic International Standard Serial Number (EISSN)
This article analyzes the tail behavior of energy price risk using a multivariate approach, in which the exposure to energy markets is given by a portfolio of oil, gas, coal, and electricity. To accommodate various dependence and tail decay patterns, this study models energy returns using different generalized hyperbolic conditional distributions and time-varying conditional mean and covariance. Employing daily energy futures data from August 2005 to March 2012, the authors recursively estimate the models and evaluate tail risk measures for the portfolio's profit-and-loss distribution for long and short positions at various horizons and confidence levels. Both in-sample and out-of-sample analyses applied to different energy portfolios show the importance of heavy tails and positive asymmetry in the distribution of energy risk factors. Thus, tail risk measures for energy portfolios based on standard methods (e.g. normality, constant covariance matrix) and on models with exponential tail decay underestimate actual tail risk, especially for short positions and short time horizons.