Variance Decomposition

[hana88]

Hi All,

I just need confirmation with variance decomposition results. First of all, I took the RGDP time series in absolute number (in millions). The variance compo that I've got looks like the file i've attached here. Just wondering do the standard error make sense? I'm having a bit of doubts since the numbers are pretty huge. Is it dependent on the type/unit i'm using (in %, millions etc).

Also, from my wald test using RGDP as the dependent variable, CPI is not significant. But from this result it looks like CPI could explain the variation in RGDP quite substantially. Hence I'm a bit confused on how to interpret this.

Thanks in advance for your help.