Applied risk management series: Venturing beyond VAR

In this article, Carlos Blanco and José Ramón Aragonés review the historical simulation methodology used to estimate value-at-risk and expected tail loss, while including adjustments to traditional assumptions that can help improve risk forecasts for energy and commodities portfolios

Value-at-risk and expected tail loss (ETL) are the pre-eminent ways of measuring financial market risk. The three main approaches used to estimate VAR and ETL are variance-covariance (also know as delta-normal, the RiskMetrics variance model, parametric or linear), Monte Carlo simulation and historical simulation (HS).

Many early adopters in the energy markets first implemented the variance-covariance method, primarily due to the issues of computational speed encountered with the other