Page 1 of 1

LM test for ESTAR/LSTAR effects

Posted: Sun Aug 05, 2012 7:13 am
by guo
Dear RATS forum team,

Always thank you very much for your previous help. I really learnt lots.

Concerning on LM test for ESTAR/LSTAR effects:

Code: Select all

@StarTest( options ) y start end

Options

D=delay on threshold variable [1]
P=lags in the autoregression [1]
How can I determine the value for D and P? Because the significance and F-statistics of H12 relies on the values of D and P.

I really appreciate your precious time and attention.

Thank you very much.

Best Regards

guo

Re: LM test for ESTAR/LSTAR effects

Posted: Sun Aug 05, 2012 8:44 am
by TomDoan
P is usually picked by looking at simple AR's. The D is typically picked by looking for the largest test statistic. See, for instance, the Terasvirta replications (http://www.estima.com/forum/viewtopic.php?f=8&t=1586)