Measuring Risk for WTI Crude Oil: An application of Parametric Expected Shortfall

Oil is the most traded commodity in the world and is an important part in the global economy. The change in the price of oil has an effect on all sectors of the economy, and the ability to capture its risk is an important research topic. This study calculates the risk of one benchmark crude oil (West Texas Intermediate) over the period 1986-2015 by estimating the Value-at-Risk (VaR) and the Expected Shortfall (ES) on daily spot returns. More specifically, this is done by using a GARCH (1, 1) model with the normal distribution, the t-distribution, and the Generalized Error Distribution (GED). The study uses a rolling window to estimate these risk measurements creating 7125 estimates for each distribution in each tail. The normal distribution was the worst performing distribution on both ES and VaR according to the backtests. The t-distribution performed good ES estimates; however it was not as accurate when calculating VaR. The GED performed the best when calculating VaR but constantly underestimated ES. The main conclusion is that both GED and the t-distribution are needed when estimating the risk for WTI.