Eviews- how to detect and remove heteroskedasticity

preview_player
Показать описание
Please check the built-in method of how to detect and remove heteroskedasticity in eviews from the link given below:
Рекомендации по теме
Комментарии
Автор

what if View/Residual Tests. not showing on? I mean I have Eviews 9 and don't have heteroskedasticity menu on tab view, please help

zuko
Автор

What about autocorrelation in log model

jhabindrapokharel
Автор

is the result still reliable/robust?
as we can see, the result is different, i.e Log(EX) is not significant, meanwhile EX has negative impact and significant.

rezkierdian
Автор

what happen if i already log all of the variables but still have heteroskedasticity problem? how can i solve it?

asrafulnazrin
Автор

Hi there, Thank you so much for your lectures. I am working on a data set of stocks return which I have already converted into log returns in Excel then imported it in the Eviews. I am facing the problem of heteroskedasticity even after changing it the estimation option at Huber-white and then Breush pagan test which shows improved values but still heteroskedastic. I am confused with if I can use log log model or not because the data has already been transformed into log model. however if I use Harvey method my results seems to have homoskedacity. waiting for your reply. Thank you so much for the good work.

minuteFin
Автор

Sir can I convert a variable into log form before doing any test like unit root test and multicollinearity? Can I convert log form for dependent variable as well?

brishtichakraborty
Автор

One more question,   sir can I convert one or 2 variable into log form only if I want ? Or
Should i convert all of them if I want to log form ?? Sir please reply

brishtichakraborty
Автор

how d o you remove heteroskedasticy in a VAR model?

uncleben
Автор

What to do with non positive values.?
I skipped that varibles but still the problem of hetero exists
Keep guide me

saniasaeed
Автор

Hi Sir, how about my model shows 'non positive number'?? what mean by that

kirawei