For checking the Granger's Causality between two variables of a data-set, lets say to check X granger-causes Y, we create two regression models, a restricted model(containing the lagged values of only Y) and a full model(containing the lagged values of only Y as well as X) for predicting the current value of Y. We compare these two regression model to check if using lagged values of X provides some additional explanatory information. My question is, do these regression models need to linear for the purpose above(with degree = 1 for X and Y) or the models can be polynomial(degree > 1) ?


Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.