Title: Application of extreme value theorem in modelling oil consumption of organisation of petroleum exporting countries
Authors: Masha Ahoba Buah; Victor Amoako Temeng; Samuel Kwesi Asiedu Addo; Yao Yevenyo Ziggah
Addresses: Department of Mathematical Sciences, Faculty of Engineering, University of Mines and Technology, Ghana ' Department of Mining Engineering, Faculty of Mining and Mineral Resource Technology, University of Mines and Technology, Ghana ' Department of Mathematics Education, University of Education Winneba, Ghana ' Department of Geomatic Engineering, Faculty of Mineral Resources Technology, University of Mines and Technology, Ghana
Abstract: In this study, oil consumption of organisation of petroleum exporting countries (OPEC) from 1973 to 2018 is modelled using extreme value theorem (EVT). The main objective of this paper was to employ both the block maxima method and peak over threshold method to determine the best fitting distribution to extreme oil consumption of OPEC. In both approaches, the maximum likelihood estimation was applied to determine the distribution parameters. Findings from the study revealed that the optimal model from the generalised extreme value (GEV) distribution was the Weibull distribution with shape, scale, and location parameter values of -0.63, 6,217.54 and 26,773.21. The optimal model from the generalised Pareto distribution (GPD) was Pareto type II distribution with shape and scale parameters of 0.04 and 666.64 at a threshold value of 34,000. A comparison of the return levels revealed that GPD gave higher return level estimates than GEV. In the study, GEV was chosen over GPD because the difference between the 95% lower and upper confidence intervals and the actual return level for the BMM was found to be lower as compared to the POT approach.
Keywords: block maxima method; BMM; peak over threshold; POT; Weibull distribution; return period.
International Journal of Business Forecasting and Marketing Intelligence, 2022 Vol.7 No.3, pp.259 - 276
Received: 26 Aug 2021
Accepted: 19 Nov 2021
Published online: 16 May 2022 *