Page 1 of 1
FEVD estimation question
Posted: Fri Oct 10, 2014 6:11 am
by ldevries
I was wondering what kind of shock values eViews uses for the estimations of FEVDs (in the standard option)? For the IRFs it are the 1 SD shocks. Is this the same for the FEVDs? When I check Hamilton (1994) I think it should be 1SD shocks also for the FEVDs. But unfortunately I cannot find this anywhere in eViews. Thank you in advance!
Re: FEVD estimation question
Posted: Fri Oct 10, 2014 1:08 pm
by EViews Chris
I'm not sure I understand exactly what you are asking. The variance decomposition numbers are expressed as percentages of the total forecast error variance within the period. I'm not sure what you are refering to by 'shock values' in this context.
What different possibilities did you have in mind for what we might be calculating?
Re: FEVD estimation question
Posted: Mon Oct 13, 2014 1:06 am
by ldevries
Well, for IRFs 1 SD shocks are used to calculate the effects of one variables on other(s). However, theoretically, you could use 1 SD or 1 unit shocks for the calculation of the FEVDs. My question is then, how are the percentages of the FEVDs exactly calculated/estimated? I hope that my question is now a bit more clear.