# BVAR with Rolling Sample

**URL:** <https://forum.dynare.org/t/bvar-with-rolling-sample/18189>\
**Category:** DSGE-VAR\
**Created:** [28 May 2021 20:34 UTC](https://forum.dynare.org/t/bvar-with-rolling-sample/18189 "2021-05-28T20:34:45Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![sahil](https://forum.dynare.org/letter_avatar_proxy/v4/letter/s/f6c823/32.png) [@sahil](https://forum.dynare.org/u/sahil)\
**Post date:** [28 May 2021 20:34 UTC](https://forum.dynare.org/t/bvar-with-rolling-sample/18189/1 "2021-05-28T20:34:45Z")

</div>

Hi,

In dynare, I know that choosing first\_obs=[period1:period2] and keeping a rolling-window of size nobs constant triggers rolling-window estimation.  
I was wondering if there is a way to estimate BVAR in similar fashion?

Thank you

Best  
S

---

<div class="post-metadata">

**Author:** ![jpfeifer](https://forum.dynare.org/user_avatar/forum.dynare.org/jpfeifer/32/5044_2.png) [@jpfeifer](https://forum.dynare.org/u/jpfeifer)\
**Post date:** [29 May 2021 18:01 UTC](https://forum.dynare.org/t/bvar-with-rolling-sample/18189/2 "2021-05-29T18:01:46Z")

</div>

One could probably use the macro-processor to run such a loop, but storing the results in that case is not straightforward. That would have to be done from Matlab.
