Understanding the Crude Oil Price Value-at-Risk: The Case of Azeri Light
31 Pages Posted: 1 Mar 2014
Date Written: February 28, 2014
Abstract
Oil-dependent governments need a well-defined risk measure for crude oil price returns to allow them to apply proper hedges, absorb market shocks and optimize decisions to increase profitability. In that regard, by employing current techniques to forecast volatility for Azeri Light crude oil prices, this study estimates the Value at Risk for Azeri Light crude oil returns in order to provide a quantifiable model for policymakers and global investors. To construct better performing volatility and risk models for Azeri Light Crude oil prices, we compare out-of-sample forecasting performances of GARCH(1,1)-GED, TGARCH(1,1)-GED and EWMA models using both DM test and other the standard measures. Also, we employ Duration based test (Christoffersen and Pel- letier (2004)) as a statistical backtest method for the VaR models, since it yields higher power compared with the formerly proposed tests, such as Kupiec (1995) and Christoffersen (1998). We find that GARCH(1,1)-GED model produces better results compared with EMWA models regard- ing the in-sample estimations. Besides, there is a considerable volatility clustering and asymmetric leverage effect in Azeri Light daily returns. Lastly, in line with the finance literature, our results favor the VaR model based on GED approach compared to the one based on the standard normal distribution.
Keywords: Azeri Light Crude Oil, GARCH, EWMA, Value-at-Risk
JEL Classification: C52, C53, G17, Q47
Suggested Citation: Suggested Citation