Vector Autoregressive Model and Analysis

Abstract

The aim of this study is to explain vector autoregressive (VAR) models and Granger causality. VAR is an econometric model that generalizes univariate autoregressive (AR) models. VAR is a regression model that can be considered as a kind of hybrid among univariate time series models. VAR models are generally defined as alternatives to structural models of large-scale simultaneous equations. All variables in the model are treated symmetrically with an equation for each variable explaining the development of the variable, depending on the lags of the variable in the model and the lags of all other variables. The method is simple. It is not necessary to specify which variables are endogenous or exogenous. VAR models are generally better than traditional structural models. Granger causality test developed by Granger is a test used to determine the direction of causality of the relationship in the presence of delayed relationship between two variables. Granger causality is really just a correlation between the present value of one variable and the past values of others; the movements of one variable do not mean that it causes the movements of another. For the VAR model and Granger causality, the variables that affect the consumer confidence index is analyzed by using Eviews. © The Authors 2021. All rights reserved.