0

Is it possible to have no causal relationship between two variable? Example: X does not granger cause Y and Y does not granger cause X. Are these results okay or there are might be an error in the regression?

Silverfish
  • 23,353
  • 27
  • 103
  • 201
Luo
  • 1
  • 1
    Welcome to the site! As it's presently written, your question is a little confusing because there are clearly examples of variables that are not causally related. It's also true that there are variables that would not be indicated to be related by Granger precedence, for example variables that co-occur or that are cyclic with different frequencies. – John Madden Jan 31 '23 at 19:36
  • 4
    Let $X_t$ and $Y_t$ be two independent sequences which are furthermore independent of one another (e.g. let $X_t,Y_t\overset{i.i.d.}\sim \mathcal N(0,I_2)$). This is a perfectly well defined sequence of random variables (and not even very complicated!) and seems to fit your situation. – stats_model Jan 31 '23 at 19:37
  • @stats_model, your comment is an answer. Why not post it as such? – Richard Hardy Feb 01 '23 at 07:39

0 Answers0