Hi all,
I hope all is well.
Please, I am studying if some of the variables affect firm instrument issuance. Assume that I have these variables (X1 X2 X3 ). Well, in the logit model , the results show that the firm with lower of X 1 are more likely to issue this security. In the second model, I have used fixed effect regression to see the impact of the instrument issuance on firm X1, the results show that issuing this instrument has no significant impact in reducing X1 as well as no increasing in firm X1 . Well, My professor say that the result of X1 in logit is contradict the findings of the second result of x1 in the fixed model. He say that perhaps there is reverse causality ? i know and I have my insight to justify this issue by different ideas but please, what dose this mean (here is reverse causality) ? and how can I test the reverse causality in stata ? Thank you so much