Orthogonalized Impulse Response Functions

[Matt Slaney]

Hi,

After running the var command, I would like to produce orthogonalized impulse response functions. However, using the "irf graph/table oirf" command shows the response (I think) to a 1 standard deviation innovation in the structural shock. Is there a command to show the orthogonalized IRF to a 1 unit innovation, or a simple workaround?

Thanks

[Eric de Souza]

The only document I know of that contains some information on how the responses are standardised is the following:
http://econpapers.repec.org/paper/bocmsug11/03.htm
And indeed it states that for orthogonal shocks, "the magnitude of the shock corresponds to one unit standard deviation"
On edit: a more recent source is http://blog.stata.com/2016/02/18/vec...ence-in-stata/

[Matt Slaney]

Thanks for confirming that. Any idea how I might get the orthogonalized IRF to a 1 unit change instead? In the data I have, it is much more intuitive to talk about a unit change than a standard deviation change.

[Matt Slaney]

As a workaround, would it be correct to divide all coefficients and standard errors from the oirf table by the impact multiplier of the impulse variable on itself?

[Eric de Souza]

Since the IRFs are nonlinear functions, I don't think that just dividing would be a good solution. I think it would have to be programmed. Stata Blog has some other posts on VARs and SVARs where Mata programs are presented. They could maybe serve as a starting point.

[Tareb Tallovi]

For the standard IRF, the innovation is one unit, and for the Orthogonal IRF (“OIRF”) the shock is one standard deviation.