# Constraint on covariance matrix for shocks in VAR representation of DSGE model

**URL:** <https://forum.dynare.org/t/constraint-on-covariance-matrix-for-shocks-in-var-representation-of-dsge-model/12209>\
**Category:** DSGE-VAR\
**Created:** [2 August 2018 09:31 UTC](https://forum.dynare.org/t/constraint-on-covariance-matrix-for-shocks-in-var-representation-of-dsge-model/12209 "2018-08-02T09:31:12Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Jesse](https://forum.dynare.org/letter_avatar_proxy/v4/letter/j/c77e96/32.png) [@Jesse](https://forum.dynare.org/u/Jesse)\
**Post date:** [2 August 2018 09:31 UTC](https://forum.dynare.org/t/constraint-on-covariance-matrix-for-shocks-in-var-representation-of-dsge-model/12209/1 "2018-08-02T09:31:13Z")

</div>

Dear Johannes,  
Thank you very much for your previous reply, I am grateful.  
I have a question about var representation of dsge model in the following paper:

> **[Stock market conditions and monetary policy in a DSGE model for the U.S.](https://www.sciencedirect.com/science/article/pii/S0165188910001582)**
>
> This paper investigates the interactions between stock market fluctuations and monetary policy within a DSGE model for the U.S. economy. First, we des…

  
In this DSGE model, the prior mean of standard errors of shocks is set to 0.01 instead of 1. I have modified this model by replacing measurement error with another structural shock, so the number of observables equals the number of shocks  
I can find a finite VAR representation for the modified DSGE model,  
Define X\_t as observable vector, U\_t as reduced form shock vector, V\_t as structural innovation vector  
\Lambda is lag operator matrix, i have:  
\Lambda X\_t=U\_t=\delta V\_t, in order to identify this finite VAR representation of DSGE model, I impose constraint on variance-covariance matrix of V\_t, Covariance Matrix of V\_t=0.01 I,  
However, in the literature, people impose covariance matrix of V\_t=I, but their associated DSGE shock variance’s prior mean is 1 instead of 0.01  
I have add 0.01 because my DSGE shock variance’s prior mean is 0.01 ,  
My question is my imposed constraint of structural shocks V\_t covariance matrix as 0.01 I right or wrong?  
Thank you very much and look forward to hearing from you.  
Jesse

---

<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:** [2 August 2018 15:11 UTC](https://forum.dynare.org/t/constraint-on-covariance-matrix-for-shocks-in-var-representation-of-dsge-model/12209/2 "2018-08-02T15:11:47Z")

</div>

Sorry, but I do not really understand the question. If you are consistent, you can use any normalization you like as the model is linear.
