The RATS Software Forum

Dear Doan
I'm using Rats version 7.0 and I've already used our LS unit root procedure in my paper. Those test gave controversial results so I also decided to use Lumsdaine and Papell (1997) unit root test to make sure my robustness of my results.
Do you have also code for the following paper?
Lumsdaine, R. and Papell, D. (1997) Multiple trend breaks and the unit root hypothesis, Review of Economics and
Statistics, 79, 212–18.
ps. In some papers including unit root with break tests, Such as Zivot and Andrews, Lumsdaine and Papell they are also giving exact critical values instead of fixed values and those critical values are changing the variable used even if variables have the same number of observations. How can be this possible? Is it possible to compute also those values with RATS?
Thanks for your interest

nazif wrote:I'm using Rats version 7.0 and I've already used our LS unit root procedure in my paper. Those test gave controversial results so I also decided to use Lumsdaine and Papell (1997) unit root test to make sure my robustness of my results.
Do you have also code for the following paper?
Lumsdaine, R. and Papell, D. (1997) Multiple trend breaks and the unit root hypothesis, Review of Economics and
Statistics, 79, 212–18.
ps. In some papers including unit root with break tests, Such as Zivot and Andrews, Lumsdaine and Papell they are also giving exact critical values instead of fixed values and those critical values are changing the variable used even if variables have the same number of observations. How can be this possible? Is it possible to compute also those values with RATS?
Thanks for your interest

This page has a procedure for doing the Lumsdaine-Papell test or anything of a similar type with breaks one and up. (one break is Zivot-Andrews). The p-values that you're talking about are, so far as I know, bootstrapped.

http://www.estima.com/forum/viewtopic.php?f=7&t=1121