[R] lag orders with ADF.test

[Pfaff, Bernhard Dr.]

>Hello!
> I do not understand what is meant by:
>
> "aic" and "bic" follow a top-down strategy based on the
>Akaike's and Schwarz's information criteria
>
>in the datails to the ADF.test function. What does a "top-down
>strategy" mean? Probably the respective criterion is minimized

Hello Martin,

are you referring to ADF.test() in package "uroot"? If so, it would be good to adress this question to the package maintainer first, which I have cc'ed.
Different approaches for determining an appropriate lag length for ADF tests are propagated in the literature. One of them is the usage of information criteria; a second one starting from a high lag number and cutting the lag length down by signifcance (top to bottom) or vice versa (bottom to top); or finally inspect the ACF/PACF's of the residuals in the ADF-test regression and choose the lowest possible lag length such that the residuals are uncorrelated.

Best,
Bernhard


>and the mode vector contains the lag orders at which the
>criterion attains it local minima? When the calculation is
>over, the ADF.test function gives info about "Lag orders".
>What are these lag orders? Are they the local minima of the criterion?
>
> I will be very thankful if you clarify this to me. I browsed
>a lot, but I could not find a clear answer.
>
> Thank you for your attention.
> Regards,
> Martin
>
>-----------------------------------------------------------------
>http://auto-motor-und-sport.bg/
>С бензин в кръвта!
>
>______________________________________________
>R-help at stat.math.ethz.ch mailing list
>https://stat.ethz.ch/mailman/listinfo/r-help
>PLEASE do read the posting guide
>http://www.R-project.org/posting-guide.html
>and provide commented, minimal, self-contained, reproducible code.
> 
*****************************************************************
Confidentiality Note: The information contained in this mess...{{dropped}}