Granger Causality Test and Lags

[slacknoise]

Hello,
I have several Questions concerning Granger Causality.
I have performed a Granger Causality Test and figuered the optimal Number of lags to be in this case 5. How do i deal with it now? Do I need to construct a new variable with a lag of 5 then in order to estimate a better regression model?

Why does eviews has as a standard a lag of 2?

Most of the time i can reject the Null Hypothesis, what does this imply for my regression model?

Thank you very much in advance.