ON MODELLING TAIL RISK OF ELECTRICAL ENERGY PRODUCTION LEVEL

Fitting the correct type of model to a particular set of data is key to proffering solution to bugging issues. Electricity production in Nigeria has been faced with various challenges for over ten decades since the first production and supply. Consequently, there is a need to measure electricity production risk. Extreme Value Theory (EVT) is considered sufficient in measuring such risk by modelling tails of the distribution. Adopting EVT, there is a need to measure Value-at-risk and Expected Shortfall which can be adequately done with Generalized Pareto Distribution (GPD); one of the models for extreme events. The preference for GPD is because it models the distribution of exceedances over a high threshold rather than the individual observations. In this study, diagnostics tests were carried out in order to determine the suitability of GPD for fitting the data, and GPD was found adequate modelling future risk of electricity production for the given data. The GPD was then used to fit the electricity production data in Nigeria at 1%, 0.5%, and 0.1% probability. Following the result, measures to avoid electricity production risks were recommended