# Variable form levels or percentage as GDP

**URL:** <https://forum.dynare.org/t/variable-form-levels-or-percentage-as-gdp/16892>\
**Category:** ML/Bayesian estimation\
**Created:** [6 November 2020 20:27 UTC](https://forum.dynare.org/t/variable-form-levels-or-percentage-as-gdp/16892 "2020-11-06T20:27:36Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![nbarone](https://forum.dynare.org/letter_avatar_proxy/v4/letter/n/f475e1/32.png) [@nbarone](https://forum.dynare.org/u/nbarone)\
**Post date:** [6 November 2020 20:27 UTC](https://forum.dynare.org/t/variable-form-levels-or-percentage-as-gdp/16892/1 "2020-11-06T20:27:36Z")

</div>

I am trying to stimate a model for the colombian economy. I’ve been following a paper from BIS. On page 14 you can find some long-term relations that I guess works as initial values. However, I’m not sure how to introduce data, for instance I don’t know if data should be in levels or as percentage of gdp as it is in page 14. In page 27 you can find information about data.

[colombia bis.pdf](https://forum.dynare.org/uploads/short-url/kwZyCMQ8Ffy3juryJwtoBiLnUPh.pdf) (1.2 MB)

---

<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:** [7 November 2020 09:09 UTC](https://forum.dynare.org/t/variable-form-levels-or-percentage-as-gdp/16892/2 "2020-11-07T09:09:59Z")

</div>

The important sentences are on page 14/15:

> The data used in the estimation is expressed in quarterly growth rates and is assumed to have a  
> measurement error

So they use first differences for the variables listed after that sentence. That deals with non-stationary variables.  
For the stationary ones after

> Additionally we included  
> they simply use the level.

---

<div class="post-metadata">

**Author:** ![nbarone](https://forum.dynare.org/letter_avatar_proxy/v4/letter/n/f475e1/32.png) [@nbarone](https://forum.dynare.org/u/nbarone)\
**Post date:** [9 November 2020 14:12 UTC](https://forum.dynare.org/t/variable-form-levels-or-percentage-as-gdp/16892/3 "2020-11-09T14:12:55Z")

</div>

oh thank you, the table distracted me. Two last questions  
1.About the inflation rate and interest rate. In page 12 set a value for the long-term parameters related to those variables. So, for the international interest rate is 2.5% and for the inflation rate is 3%. However, in table 1 page 13 they set up different values, for instance, for the international interest rate is 1.0062 and for the inflation rate is 1.0074. I don’t know why.  
2. In page 25 equation 78 there is an expected value in t minus one about one variable in t. How can I introduce that kind of equation in dynare.

Thank you

---

<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:** [9 November 2020 16:05 UTC](https://forum.dynare.org/t/variable-form-levels-or-percentage-as-gdp/16892/4 "2020-11-09T16:05:26Z")

</div>

1. The model is quarterly while the targets are reported annualized.
2. See [End of period convention and decision lag](https://forum.dynare.org/t/end-of-period-convention-and-decision-lag/6014)

---

<div class="post-metadata">

**Author:** ![nbarone](https://forum.dynare.org/letter_avatar_proxy/v4/letter/n/f475e1/32.png) [@nbarone](https://forum.dynare.org/u/nbarone)\
**Post date:** [9 November 2020 16:45 UTC](https://forum.dynare.org/t/variable-form-levels-or-percentage-as-gdp/16892/5 "2020-11-09T16:45:40Z")

</div>

Thank you for your answers