Forecasting of a time series can be done using various models. All the models, on the basis of number of variables used can be divided into two sets: univariate and multivariate forecasting tools. VAR is a multivariate method to forecast the time series. There are several strengths and weaknesses of this model. It is a restrictive model and can be used under a few typical conditions. It only uses pre-determined variables and no contemporary terms are part of the model specification. It uses more than two variables in the systems of equation format and helps in understanding impact on one variable due to the change or shock in another variable by using impulse response function and variable decomposition tools. In toto, it is a good tool to do the multivariate forecasting of the time series. It is difficult to predict which is a better model given so many models in the market, which includes statistical as well as machine learning tools. However, VAR has a unique place in the multivariate forecasting world.
VAR, Forecasting, Time-series, Multi-variate, Inflation
Rahul Singh Gautam , Jagjeevan Kanoujiya "Multivariate Inflation Forecasting: A Case of Vector Auto Regressive (VAR) Model" Iconic Research And Engineering Journals Volume 5 Issue 12 2022 Page 11-14
Rahul Singh Gautam , Jagjeevan Kanoujiya "Multivariate Inflation Forecasting: A Case of Vector Auto Regressive (VAR) Model" Iconic Research And Engineering Journals, 5(12)