Historical decomposition

[indrani_5]

I am attempting a historical decomposition with my VAR model and in that context trying to understand the references in your user guide pages UG-239 and 240. The solid line in the figure on page UG 239 is the actual GDP. Which one is the base forecast? Is it the dashed line? What does the dotted line indicate? How did you reach the conclusion that entire difference between actual gdp (the solid line) and the base forecast (?) can be attributed to interest rate shocks? Is it because the dotted line aligns with the dashed line uptil 2004?

Now my second set of questions is regarding my model. I used the code given in user guide page 240 (I am not posting the code lest it violates copyrights) and just changed the name of the model. It produced some graphs but I dont understand what it indicates. I am attaching the graphs though. My questions?

1) Is the black line the actual variable?
2) Is the blue line base forecast?
3) What does the green line indicate?

Thanks.

[TomDoan]

Regarding copyrights, excerpts are fine---just not the whole thing.

Black is the actual data.
Blue is the base forecast.
Green is the base forecast plus the cumulative effects of just the one set of shocks (all others zeroed). So, for instance, with the NPA data, the Y and NPA shocks do effectively nothing, while COMM_NE explains most of the overall drop and INT_RATE most of the shorter-term movements.