Pairwise and Block Exogeneity Wald Test

[lalaland]

I have 2 variables: Oil price, Exchange Rate. First I open 2 series as group and run Pairwise Granger Causality Test with 5 lags. The p-value for "Exchange Rate does not Granger cause Oil price" is: 0,0041. Then I run unrestricted VAR with 5 lags anh run VAR Granger Causality/Block Exogeneity Wald Test. The p-value for in case of dependence variable is Oil price is: 0,0023 and of dependence variables is Exchange Rate is 0,0586. Why is it different? I try to omit the constant c but the p-value is also different!

[EViews Gareth]

Degrees of freedom are different. I believe the group includes all variables (so 5*2+1 in your case) in the DF calculation.