Page 2 of 2

Re: LSUnit—Lee-Strazicich unit root test with multiple break

Posted: Tue Apr 09, 2019 7:17 am
by smgrilli
Hi Tom,

While I am able to compute the lsunit test, the output does not display the results in my Rats Version 8. Any idea why?

Code: Select all

@lsunit(breaks=2,pi=.10,lags=5,method=gtos,model=break) illiqaut

Code: Select all

Lee-Strazicich Unit Root Test, Series ILLIQAUT
Regression Run From 1991:07 to 2018:12
Observations         330
Trend Break Model with 2 breaks
With 2 lags chosen from 5

Variable     Coefficient T-Stat
S{1}             -0.5261 -7.4733
Constant         -0.1638 -3.9510
D(1998:07)       -0.0666 -0.2100
DT(1998:07)       0.4170  5.6078
D(2005:03)        0.5360  1.6726
DT(2005:03)      -0.5367 -6.5558
Thanks

Stefano

Re: LSUnit—Lee-Strazicich unit root test with multiple break

Posted: Tue Apr 09, 2019 8:07 am
by TomDoan
The test statistic is the t-stat on S{1}.

Your software is seven years old and some of your procedures are as well. You can get the newer version of @LSUNIT---the link is in the first post, but an update will get the newest versions of everything and it is also easier to keep up-to-date than version 8 was.

Re: LSUnit—Lee-Strazicich unit root test with multiple break

Posted: Fri Apr 23, 2021 7:53 am
by sue
Hi Tom,

I have an exchange rate dataset(R) (4790 days of data) and I used LSUnit - Lee-Strazicich unit root test and 2 structural breaks.
I need to look at both the crash model and the break model.
But it takes a long time and I have been waiting for 3 hours.
Why does it take so long? How can I make it take a short while?

@LSUNIT (LAGS=4, MODEL=CRASH,BREAKS=2) R

Re: LSUnit—Lee-Strazicich unit root test with multiple break

Posted: Fri Apr 23, 2021 8:54 am
by TomDoan
That's explained in the description of the procedure. See the second paragraph.