VAR estimation for variance-covariance matrix

[yiagos]

I am not very familiar with VAR analysis, so I would appreciate your help.
I am running a VAR of two country daily stock market returns in order to get the variance-covariance matrix(to compute correlations). I am using the two countries interest rate and oil prices as exogenous variables to control for aggregate shocks. My question is should I use the interest rate and oil price as it is or use the first difference?

[alex_hk90]

I would've thought that it would depend on whether you think it is the levels of those variables that affect the (level of?) the daily stock market returns or whether it is the changes (first differences) of these variables which do so. I don't know anything about the stock market so can't really help beyond that, sorry.