Page **8** of **8**

### Re: Time varying SVAR

Posted: **Tue Apr 03, 2018 11:18 pm**

by **rosh**

thanks, it was due to old version of Eviews, I just need to ask that my cointegration test show some longterm relations, if still I can apply TVSVAR, when i checked for unitroot, all variables are stationary except the interest rate, I am using the growth rates for each variable..

### Re: Time varying SVAR

Posted: **Fri May 11, 2018 6:57 am**

by **binhpham79**

Hi Dakila,

I attempt to replicate fig 4-7 of Primiceri (2005). The add-ins does not permit access to its estimates and matrix coeffs. Could you please provide the way replicate the paper fully?

My understanding is that the author computed irf for each point in time (it could be 193 x 4 irf tables), say irf with the 60-quarter horizon. Then collect the irf(0), irf(10), irf(20) and irf(60) of each irf set to plot fig 5. I just confuse what the term permanent shock means? Could you point me out this point?

Anyway, to compute we need the irf matrix or series generated. So, the add-ins should allow for that.

Many thanks.

### Re: Time varying SVAR

Posted: **Mon Jun 11, 2018 6:37 am**

by **mkz86**

IMMEDIATE

Hi everyone,

My work consists of current account (dependent variable) and inflation, exchange rate, interest. My time series are 2008q1-2018q3. What sholud be the code that I need to write. And also if I use dialog box, what should be entries?

If you will help to me, I appericiate

My regards

### Re: Time varying SVAR

Posted: **Sun Jan 06, 2019 12:17 am**

by **mwxhappy**

hi, I see your TVSVAR add-in has the date selection vector to get a certain date's impulse result, but if I want to get some lag periods impulse results, what should I do?

for example, I want to get 4-period lag, and 8-period lag impulse results.what should I do?