Abstract:
Purpose: The study focused on modelling the volatility of energy markets spot prices using
GARCH models and estimating Value-at-Risk.
Methodology: The conditional heteroscedasticity models are used to model the volatility of
gasoline and crude oil energy commodities. In estimating Value at Risk; GARCH-EVT model is
utilized in comparison with other conventional approaches. The accuracy of the VaR forecasts is
assessed by using standard statistical back testing procedures.
Results: The empirical results suggests that the gasoline and crude oil prices exhibit highly
stylized features such as extreme price spikes, price dependency between markets, correlation
asymmetry and non-linear dependency. We also conclude that the EGARCH-EVT model is more
robust, provides the best t and outperforms the other conventional models in terms of forecasting
accuracy and VaR prediction. Generally, the GARCH-EVT model can be used to plays an integral
role as a risk management tool in the energy industry.
Unique contribution to theory, practice and policy: In light of the research findings, the study
recommends that organizations should leverage modern technology as a basis of realizing
efficiency, effectiveness, and sustainability of projects. The study likewise recommends that
organizations should build capacities to enhance labour productivity. In addition, the study
recommends that organizations should adopt transformational leadership approaches as a basis of
enhancing performance. The study recommends the need to revise the legal framework with a view
to ensure that it reflects the changing needs of the project requirements.
Keywords: Back testing, extreme value theory (EVT), Peak-over-threshold (POT), GARCH-EVT
model, Value-at-Risk (VaR).
Description:
Journal of Statistics and Actuarial Research
2518-881X (online)
Vol. 2, Issue 1, No. 1, pp 1-32, 2019