SVAR for monetary spillover effect

[Thomas Lee]

Hi everyone,
I'm working on my research about monetary spill over effect of the US on emerging market. I try to use svar to structure the contemporaneous effect between countries. However, when dealing with panel data i have separate to 11 variable ie. y_us, r_us, y_ja, r_ja, y_uk,r_uk. The problem is that it seems is too heavy for the impulse response part then it keep not responding when i run the do file. Do you have any suggestion ? Is there any way that can form 2,3 variables into a vector and regress on the vector ? Is that true that better computer can compute somethings that i cannot right now. For a better view, please look at my script. Thank you so much.

set mem 15m
set matsize 5000
import excel using "svar1", first
gen quarter = tq(1980q1) +_n-1
tsset quarter, quarterly
replace quarter = quarterly("","YQ") if quarter == quarterly("2018q1","YQ")

matrix A = ( 1,0,0,0,0,0,0,0,0,0,0 \ .,1,0,0,0,0,0,0,0,0,0 \ .,.,1,0,0,0,0,0,0,0,0 \ .,.,.,1,0,0,0,0,0,0,0 \ .,.,.,.,1,0,0,0,0,0,0 \ .,.,.,.,.,1,0,0,0,0,0 \.,.,.,.,.,.,1,0,0,0,0 \ .,.,.,.,.,.,.,1,0,0,0 \ .,.,.,.,.,.,.,.,1,0,0 \ .,.,.,.,.,.,.,.,.,1,0 \ .,.,.,.,.,.,.,.,.,.,1)
matrix B = (.,0,0,0,0,0,0,0,0,0,0 \ 0,.,0,0,0,0,0,0,0,0,0 \ 0,0,.,0,0,0,0,0,0,0,0 \ 0,0,0,.,0,0,0,0,0,0,0 \ 0,0,0,0,.,0,0,0,0,0,0 \ 0,0,0,0,0,.,0,0,0,0,0 \ 0,0,0,0,0,0,.,0,0,0,0 \ 0,0,0,0,0,0,0,.,0,0,0\ 0,0,0,0,0,0,0,0,.,0,0 \ 0,0,0,0,0,0,0,0,0,.,0 \ 0,0,0,0,0,0,0,0,0,0,.)
svar pi_us y_us r_us pi_jap y_jap r_jap e_jap pi_kor y_kor r_kor e_kor, lag(1/2) aeq(A) beq(B)
matlist e(A)
matlist e(B)
irf cre order1, step(10) set(myirf1,replace)
irf graph sirf, xlabel(0(4)20) irf(order1) yline(0,lcolor(black)) byopts(yrescale)

[Igor Paploski]

Hi Thomas,

Whenever I hit issues that I believe are due to them being "too much" for the computer to handle, I try to simplify the problem and then increase it's complexity in steps, so I can detect when the software stops working (a lot of times I find an easier way to write my command while attempting this, so in the end I can solve the original problem with an improved code).

Also, please note that while people here are sympathetic for the situation of others, hard to understand and replicate issues are much more unlikely to be unanswered. Please don't bump the thread - it is better to improve a question than just to ask the same one over again.

I would suggest you trying to re-write your question in clearer fashion while providing more data (please use dataex so others can replicate your database and try running your commands on it). Also, please use the code delimiters (the # box on the right of your screen when you are typing a response on the forum) whenever you post codes - this makes it easier to other users to identify the beginning and end of your code.