0
$\begingroup$

I have a var model :

y=dummy + other variables where dummy =1 if the firm is having a negative return on stock and 0 otherwise. Y is the return on stock. Is it appropriate to use the VAR model to study the impact of the dummy on y and other variables in this context?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
1
$\begingroup$

You might want to look at threshold VARs (TVAR). Lo and Zivot, 2001 They allow you to estimate seperate Var equations depending on the state of the economy/firm. Examplesource

Here is variable c the threshold variable. In principle, an infinite number of boundaries are conceivable. In practice, however, it is usually no more than 2-3, which is also due to the limited number of degrees of freedom.

There are also TVEC Models, which can deal with cointegrated data.

Improve this answer
$\endgroup$
1
  • 2
    $\begingroup$ This is a little brief as an answer by our standards; you could improve it by saying a little about what TVARs are and why you think they would help in this case. $\endgroup$
    – Silverfish
    Commented Aug 20, 2018 at 18:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.