A MIXED DATA SAMPLING (MIDAS) APPROACH TO MODELING CRUDE OIL PRODUCTION IN NIGERIA: ACCOUNTING FOR MIXED—FREQUENCY DATA

Abstract

This research work attempt to establish an efficient method of forecasting Nigeria’ Crude Oil
Production and the Nigeria Gross Domestic Product applying Mixed Data Sampling approach
(MIDAS) from 2010 to 2022. It combined data set of different frequencies; quarterly and
monthly in the same regression. It was observed that based on the Root Mean Square Error the
MIDAS Almon (PDL) regression model provided a better model estimation than the MIDAS
Step weighting and the MIDAS Beta. The monthly Crude oil production has a positive effect on
GDP as the slope coefficient is statistically significant 0.849895 (Prob. 0.0000). It is therefore
important for appropriate policy formulation and implementation of such policies to encourage
and boost these variables for effective management of Crude oil production in Nigeria. Hence,
direct relationship between Crude oil production and GDP is needed to diversify the economy
base to enhance productive activities in Nigeria and better crude oil production.