Page **2** of **4**

### Re: TARCOINT

Posted: **Tue Mar 26, 2013 3:48 am**

by **delta_h**

Ok thanks, the exogenous variable influences only the residuals but when i estimate the model with or without the exogenous variable all coefficients and statistics are the same. Nothing change.

### Re: TARCOINT

Posted: **Tue Mar 26, 2013 3:27 pm**

by **trubador**

Try defining your exogenous variable as a group with single series and then enter the name of this group when running the add-in. It seems the code does not read if the type of input variable is series. I'll fix it...

### Re: TARCOINT

Posted: **Wed Mar 27, 2013 1:36 am**

by **delta_h**

It works with the exogenous variable as a group.

Thank you very much.

### Re: TARCOINT

Posted: **Fri Jun 21, 2013 6:18 am**

by **zhangshuai**

if i want to test following asymmetric ECM equation ∆H_t=Aw_(t-1)^+ +βw_(t-1)[b]^- +C∆H_(t-1)+D∆m+u_t ,where w_(t-1) should be divided into postive and negative. However, I dont know which factors should enter into the endogenous varibales and which should enter into the exogenous variables in the threshold adjustment of EVIEWS.

i guess the ∆H and ∆m should be the endogenous variables and w_(t-1)should be exgoneous variable. is that right?

THANKS

### Re: TARCOINT

Posted: **Fri Jun 21, 2013 6:28 am**

by **zhangshuai**

I want to test following asymmetric ECM equation ∆H_t=Aw_(t-1)^+ +βw_(t-1)^- +C∆H_(t-1)+D∆m_(t-1)+u_t where w_(t-1) should be dividen into positive and negative and we want to test the asymmetric sign effect of w_(t-1). however we donot know which variables should be the endogenous variables in the above equation and which should be the exogenous variables in the threshould adjustment of Eviews and how to set the threshold level. Should i treat w(the cointegration term) as exgoneous variable and ∆H and ∆m as endogenous variables?

thanks

### Re: TARCOINT

Posted: **Tue Mar 11, 2014 2:12 am**

by **berbatuan9**

Thanks alot for this useful add-in

Please help me for the TCEM model (Threshold correct error correction)

How can i use it in Eviews? I just completed TAR and MTAR for my data? But I dont know What I do next for TECM? Please help me

### Re: TARCOINT

Posted: **Wed Jul 09, 2014 4:37 am**

by **Lucas**

I'm trying to replicate the results found by enders and siklos (2001) using your eviews add-in and a database available at Estima. However, all the results that I estimate are different from the paper (Johassen cointegration, Engle-granger, TAR and M-TAR).

Did anyone get the same results? It is a problem on the data?

Another question, why the delta default value is 0,6? Do I need to test differents values?

Thanks

### Re: TARCOINT

Posted: **Wed Jul 16, 2014 6:33 pm**

by **ninjagirl**

Hello,

I wanted to find a consistent estimate of the Threshold as per Chan 1993. I can't seem to figure out how I can do that using the add-in. Any help? Im consistently getting threshold as 0.

Please help.

### Re: TARCOINT

Posted: **Tue May 19, 2015 1:07 pm**

by **pitsi**

dear trubador,

i use the add-in tarcoint and i would like to know whether i do the procedure right. i do the dols and then decompose the residuals i have obtained from dols into - and + and then put them as endogenous variables in the tarcoint. is that right?

Thank you for your time

### Re: TARCOINT

Posted: **Sat May 23, 2015 2:33 pm**

by **trubador**

There is nothing technically wrong with running the procedure on the residuals. However, I cannot verify or validate your approach from an econometric standpoint.

### Re: TARCOINT

Posted: **Mon May 25, 2015 4:12 pm**

by **pitsi**

thank you very much for the reply trubador!!!

does tarcoint automatically uses the engle - granger cointegration ? so maybe there is a problem with the dols!? or am i ok?

i prefer the residuals because i thought to perform first a dols cointergration with breaks as exogenous and then decompose the res into + and minus and add them to the tarcoint.

what do you think about this concept?

Thank you once again for your time!

### Re: TARCOINT

Posted: **Tue May 26, 2015 12:05 am**

by **trubador**

Yes, tarcoint uses Engle-Granger specification. I am really not sure if you can (or should) follow that approach. You can, however, apply the same procedure and determine the threshold in your case, if you are willing to write your own code.

### Re: TARCOINT

Posted: **Tue May 26, 2015 3:05 pm**

by **pitsi**

thank you so much trubador!!

### Re: TARCOINT

Posted: **Sun May 31, 2015 7:59 am**

by **pitsi**

Dear trubador,

is it possible to get the aic or bic values?

Thank u!!

### Re: TARCOINT

Posted: **Sun Jun 07, 2015 2:58 am**

by **yingtan**

ninjagirl wrote:Hello,

I wanted to find a consistent estimate of the Threshold as per Chan 1993. I can't seem to figure out how I can do that using the add-in. Any help? Im consistently getting threshold as 0.

Please help.