# Variance decomposition

**URL:** <https://forum.dynare.org/t/variance-decomposition/1564>\
**Category:** Dynare help (legacy posts)\
**Created:** [28 July 2010 07:18 UTC](https://forum.dynare.org/t/variance-decomposition/1564 "2010-07-28T07:18:37Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![matte](https://forum.dynare.org/letter_avatar_proxy/v4/letter/m/7cd45c/32.png) [@matte](https://forum.dynare.org/u/matte)\
**Post date:** [28 July 2010 07:18 UTC](https://forum.dynare.org/t/variance-decomposition/1564/1 "2010-07-28T07:18:37Z")

</div>

Dear,  
how I can see variance decomposition for different periods?  
For instance, I need to to obtain the variance decomposition for, say, 4, 10 and 20 periods ahead. I try to program in the following way:  
stoch\_simul(irf=4) var1 var2 var3;  
stoch\_simul(irf=10) var1 var2 var3;  
stoch\_simul(irf=20) var1 var2 var3;  
but I get the same numbers for the three cases. What is the way to do it right?  
Many thanks in advance!

---

<div class="post-metadata">

**Author:** ![AssiaEzzeroug](https://forum.dynare.org/letter_avatar_proxy/v4/letter/a/b19c9b/32.png) [@AssiaEzzeroug](https://forum.dynare.org/u/AssiaEzzeroug)\
**Post date:** [28 July 2010 07:56 UTC](https://forum.dynare.org/t/variance-decomposition/1564/2 "2010-07-28T07:56:48Z")

</div>

Hi,

that’s normal you get the same results because for now, Dynare can only compute the variance decomposition for infinite horizons.

Best

---

<div class="post-metadata">

**Author:** ![matte](https://forum.dynare.org/letter_avatar_proxy/v4/letter/m/7cd45c/32.png) [@matte](https://forum.dynare.org/u/matte)\
**Post date:** [28 July 2010 09:15 UTC](https://forum.dynare.org/t/variance-decomposition/1564/3 "2010-07-28T09:15:28Z")

</div>

Thanks a lot for your answer!  
Do you know any possible way to compute the var. decomposition for different horizons? Any program for Matlab?