Page 1 of 1

log condition for ARDL

Posted: Tue Oct 14, 2014 6:00 am
by Em jay
hylo everyone. i want to solve one of quarry about my econometric research please
i have applied ARDL model in my dissertation but i haven't take log nor go for impulse response .. my results are consistent.. can anyone tell how can i defend my this point with logic that it isn't necessary to take log and impulse response function?

Re: log condition for ARDL

Posted: Sat Nov 08, 2014 12:04 am
by nishantvats12
Taking log is not at all necessary unless you are interested in:

a. Analyzing the elasticity
b. removing heteroskedasticity (if it exists)

Impulse response functions are needed for a specific kind of analysis only if you are not interested in it, you might not show it.

Re: log condition for ARDL

Posted: Tue Nov 11, 2014 11:43 am
by Em jay
thanks alot