9. Non-stationarity and Unit Root Testing using EViews || Dr. Dhaval Maheta

#econometrics, #timeseries, #regression, #eviews, #stationarity, #unit, #root, #augmented, #dickey, #fuller, #philips, #perron, #kpss, #correlogram
Email: [e-mail]
Twitter: / dhavalmaheta77
LinkedIn: / dhaval-maheta-320200153
Facebook ID: / dhaval.maheta.37
Telegram ID: t.me/DhavalMaheta

[@yesugen4126]

This is very useful. Thank you!

[@azzahassan84]

@DhavalSaifaleeAaryash Thank you for this great efforts. Please if all independent variable stationary at first difference but the dependent variable is stationary at level, should I take the first differences for all of them including the dependent variable?
And if the it is stationary at level when applying KSPP test, but stationary at first difference when using ADF AND PP TESTS should i use the first difference in this case?

[@mohamedabdirahmanabdihamid8257]

Hello, am always confused when testing for stationarity, when to include trend and intercept option, just the intercept or the none option, what determines when including those options?
My second question is, if data is showing non stationary with trend for example but it becomes stationary with intercept assuming that they are on same level of order on integration, can I assume stationarity if the variable you are testing has become stationary with intercept and trend option but non stationary with intercept option when testing for ADF test given that the order of integration is the same or all of those options has to show stationarity for me assume stationarity.

[@vaibhavgadhave2]

In depth analysis very nice sir hope we will have some interactions in future

[@DhavalSaifaleeAaryash]

You can join my whatsup channel
chat.whatsapp.com/HoJ7AqIPnoW7cPSTsYekjZ

[@dhanu3367]

Sir the link for ur WhatsApp channel is not working sir. Kindly reshare the link

[@dhanu3367]

After generating the new differenced series of a variable, should we use that new series only for our regression and further modelling?

[@DhavalSaifaleeAaryash]

Yes

[@EconomiaASAP]

Thank you so much for this great work!!! When applying "genr dgdp=d(gdp, 1)", is the same of applying the ADF test on first difference + the intercept and trend? if not, how can I type the command to do so?

[@DhavalSaifaleeAaryash]

yes correct but we are generating the variable

[@EconomiaASAP]

@@DhavalSaifaleeAaryash how can I use this command to do the same but first, without trend and intercept? and with only the intercept?

[@DhavalSaifaleeAaryash]

@@EconomiaASAP when you take first difference, constant and trend is removed automatically

[@khalilahmadbahadur8878]

what if I got different result from ADF and PP, on which should I rely?
Suppose GDP using ADF is I(2) and with PP is I(1) , inflation using ADF is I(0) and PP - I(1) and all other variable are I(1) using both

[@DhavalSaifaleeAaryash]

Ardl models

[@user-nr5vw3vz1y]

Are you sure for ARDL selecti9n having variable of integrated I2)?