EViews.com

Hey everyone.

I ran through forum archives but couldn't find exact answer for my question, so decided to make a new post.

Eviews provides generalized generalized impulse response function for one SD innovations after VAR\VECM estimation. The issue for me is receiving generalized IRF for ONE UNIT shock for estimated VECM.
1. Are there any possibilities to get it in Eviews (may be I just didn't look for them properly)?

2. If answer for aforementioned question is negative then is it possible to compute generalized IRF for one unit shock manually?
Pesaran and Shin (1998) "Generalized impulse response analysis in linear multivariate models" states that "the m*1 vector of the (unscaled) generalized impulse response of the effect of a shock in the j-th equation at time t on x_t+n" is inversely proportional to j-th innovation (shock).
So to compute generalized IRF of one unit shock of endogenous variable (say X) to another endogenous variable (say Y) instead of one SD shock one should just take provided by Eviews generalized IRF for one SD shock and normalize it by by X's SD for each future period in IRF. Is that correct?

If it's wrong could you please state an algorithm of receiving generalized IRF for one unit shock in VECM.
Thanks in advance.

Best regards,
s1Lence

You can provide your own shock matrix, with whatever values you want in it, including a "1".

You can provide your own shock matrix, with whatever values you want in it, including a "1".

Thanks for your reply but I still need help here.
The user specified shocks feature in Eviews allows to define impulse values (shocks\innovations values) but doesn't allow to specify IRF type (or I'm wrong?). I mean lets imagine that innovation matrix "shock" is constructed. How is it possible to choose generalized IRF type (or orthogonalized or just regular residual IRF)?

Ah sorry, I didn't read your question fully enough. I'm afraid I don't know enough about IRFs to answer.

Ah sorry, I didn't read your question fully enough. I'm afraid I don't know enough about IRFs to answer.

It's a pity. May be someone could help?