demand and supply shocks

Questions and discussions on Vector Autoregressions
sachin
Posts: 7
Joined: Fri May 08, 2009 11:53 am

demand and supply shocks

Unread post by sachin »

hey,

I got confused how to graph demand and supply shocks and how to get the exact single value of demand and supply shocks under the Blanchard-Quah decomposition. Earliest reply will be appreciated.
TomDoan
Posts: 7814
Joined: Wed Nov 01, 2006 4:36 pm

Demand and supply shocks

Unread post by TomDoan »

You need the structural residuals. These can be computed using the @StructResids procedure:

Code: Select all

estimate(resids=resids)
*
* Compute the Blanchard-Quah factorization of the covariance matrix
*
compute bqfactor=%bqfactor(%sigma,%varlagsums)
@structresids(factor=bqfactor) resids %regstart() %regend() sresids
With the BQ factor, the supply shocks are in the first component (sresids(1)) and the demand shocks in sresids(2).

Check the Blanchard-Quah AER 1989 topic. That was added since this was originally written
sachin
Posts: 7
Joined: Fri May 08, 2009 11:53 am

Re: demand and supoply shocks

Unread post by sachin »

hey,
thanks for the reply with commendable help. I did what u said, but i got such results, which i have hereby posted. I may be stupid, but i am still unable to figure out demand and supply shocks from this result. Could you also let me know codes for making graphs in rats for demand and supply shocks as well?
I am very new for RATs. So, your suggestions will be highly appreciative. Once again, thanks for the reply. Looking forward to see reply for this very soon.
compute bqfactor=%bqfactor(%sigma,%varlagsums)

procedure StructResids
(01.0005)
(01.0005) source structresids(factor=bqfactor) resids %regstart() %regend() sresids
(01.0022)
(01.0022)
(01.0022)
(01.0022)
(01.0025)
TomDoan
Posts: 7814
Joined: Wed Nov 01, 2006 4:36 pm

Re: demand and supoply shocks

Unread post by TomDoan »

StructResids already exists. Don't try to write a new one. Just use the @StructResids... as I wrote it above.

If you want to graph the supply shock (after doing the @StructResids line), use something like:

Code: Select all

graph(footer="Supply Shock") 
# sresids(1)
The demand shock is

Code: Select all

graph(footer="Demand Shock") 
# sresids(2)
sachin
Posts: 7
Joined: Fri May 08, 2009 11:53 am

Re: demand and supoply shocks

Unread post by sachin »

hey,

Thank you very much for reply. But it does not work to me. I am confused. I coded in the following way.
Could you let me know where it went wrong. I am using Rats 6.2 version. I have been struggling and struggling to solve it. Please help me once again.

Code: Select all

cal 1970
allocate 2007:1
open data in.txt
data(format=free,org=obs) / yyy ingdp incpi usgdp


seed 123456

smpl 1970:1 2007:1
set lningdp = log(ingdp)
set lincpi = log(incpi)
set lusgdp = log(usgdp)



smpl 1970:1 2007:1
set dingdp = lningdp -lningdp(t-1)
set dincpi =lincpi-lincpi(t-1)
set dusgdp =lusgdp-lusgdp(t-1)

system(model=mywork) 
variables dingdp dincpi 
lags 2
det constant dusgdp{0 to 1}
end(system)

estimate(resids=resids)
*
* Compute the Blanchard-Quah factorization of the covariance matrix
*
compute bqfactor=%bqfactor(%sigma,%varlagsums)
 
@StructResids(factor=bqfactor) resids %regstart() %regend() sresids

graph(footer="Supply Shock")
# sresids(1)

graph(footer="Demand Shock")
# sresids(2)

end(system)
However, the rats asked to open StructResids and could not find the file name. then, it gives such a message"
"## CP18. STRUCTRESIDS is not the Name of a PROCEDURE. (Did you forget to SOURCE?)"

Once again, awaiting your reply.

I really appreciate your help
TomDoan
Posts: 7814
Joined: Wed Nov 01, 2006 4:36 pm

Re: demand and supoply shocks

Unread post by TomDoan »

sachin wrote:

Code: Select all

system(model=mywork) 
variables dingdp dincpi 
lags 2
det constant dusgdp{0 to 1}
end(system)
I think you want that to be

Code: Select all

lags 1 2
Post Reply