Modeling of Economic Data using Bivariate MGARCH Models

Abstract

Aims: The aim of this study is to examine Economic data using the multivariate GARCH model. Study design: The study used monthly data of Nigerian crude oil prices (dollar Per Barrel) and Consumer price Index Methodology:  This work covers time series data on crude oil price and consumer price Index rural obtained from   Central bank of Nigeria (CBN)   from 2000 to 2019. To achieve the aim of the study, bivariate VECH and BEKK model were applied. Results: The results confirmed that returns on economic data were correlated. Also, diagonal multivariate VECH model confirmed one of the properties that it must be ‘positive semi-definite’ and the BEKK also confirmed the volatility spillover effects among the economic data. Conclusion: From the results obtained, it was confirmed that conditional variances depends only on own lags and own lagged square returns and conditional covariances depends only on own lags and own lagged cross products of returns. As for cross-volatility effects, past innovations in crude oil price have greatest influence on future volatility of returns on economic data. It was also confirmed that time varying covariance displays among these economic data and lower degree of persistence and based on Model selection criteria using the Akaike information criteria (AIC) diagonal VECH model is better fitted than the BEKK model.