VAR models have been an indispensable tool in empirical macroeconomic research.  It has been noted that adding a time-varying variance (i.e. heteroscedasticity) to a regression is empirically very important with economic data, as it helps in making better forecasts and in obtaining more accurate measures of uncertainty. Although there is previous work that constructs the VAR model with heteroscedasticity, this literature has two main problems.

1) Estimates are sensitive to the order of the variables in the dataset,

2) The estimation method becomes impractical when the number of variables in the VAR is large.

The objective of this project is to introduce heteroscedasticity in the VAR model in such a way that these two problems are solved.