The number of variable data

I have a question about the number of variable data.
If I have more than 10 variables, and only three variables have data, is it ok?
if not, why the results can be out?

[Exogenous shocks in Bayesian estimation)