Seeking to test a longitudinal panel using an IV. Would it be correct to run an ARMA and then use the predicted values as the endogenous variable which can then be conditioned using the (single) IV. If so would it remain valid to use the standard tests for validity and significance? There are about 20 (potentially) endogenous variables in the ARMA, 60 time periods and 10 observations per time period. Grateful for any help!
-
Login or Register
- Log in with
