0

Hi TimeSeries and R gurus,

Is there any way to impliment Elliott–Rothenberg–Stock (ERS) with Hannan-Quinn lag selection criterion in R?

Libraries like urca, fUnitRoots and fSeries have commands for Elliott–Rothenberg–Stock unitroot test. However, I could not find any option where I can use HQIC for automatic lag length selection.

For example, here is what is available:

gnp <- na.omit(nporg[, "gnp.r"])
ers.gnp <- ur.ers(gnp, type="DF-GLS", model="const", lag.max=4)
summary(ers.gnp)

I am after something like:

gnp <- na.omit(nporg[, "gnp.r"])
ers.gnp <- ur.ers(gnp, type="DF-GLS", model="const", lag.max="HQIC")
summary(ers.gnp)

Aim is to get something like following produced in Eviews:

enter image description here

Any help in this regard is much appreciated.

Mubashir

M.Qasim
  • 1,827
  • 4
  • 33
  • 58
  • 1
    URCA is using BIC here. As URCA is written in R, it would be very easy to modify the `ur.ers` code to use your own information criteria in place of BIC. – A. Webb Feb 08 '17 at 15:28
  • A. Webb, thanks for your time and suggestion. My questions is if there is any existing package to achieve this. – M.Qasim Feb 09 '17 at 01:23
  • 1
    After spending a lot of time I found there is not any existing R packing to implement Elliott–Rothenberg–Stock (ERS) unitroot test with Hannan-Quinn (HQC) lag selection criterion... – M.Qasim Feb 13 '17 at 00:12

0 Answers0