MCPROCESSIRF (Calculation of bands for IRF's)
Posted: Mon Aug 20, 2012 1:02 pmThis is a revised version of the @MCPROCESSIRF procedure, which takes a set of simulated impulse response functions (done with, for instance, @MCVARDODRAWS) and generates upper, lower and center series as they are for @MCGRAPHIRF (see http://www.estima.com/forum/viewtopic.php?f=7&t=995) without doing the graphs. That way you can report the information as you wish. The input simulated responses have to be in the %%RESPONSES array as described in the comments to the procedure.
@MCPROCESSIRF(MODEL=model used in generating responses,other options)
CENTER=[MEAN]/MEDIAN/INPUT
IMPULSES=RECT[SERIES] with input "center" for IRF's
These determine what is to be used as the point estimate of the IRF. The default is MEAN, which will be the mean across simulations. MEDIAN requests the median. If CENTER=INPUT, you must input the desired point estimates using the IMPULSES option---typically the impulse responses at the original estimates for the VAR.
PERCENTILES=||percentiles for lower and upper bounds|| [||.16,.84||]
STDDEV=# of standard deviations from mean for lower and upper bounds [not used]
STDDEV is used for doing error bands based upon multiples of the sample standard deviations. For instance, STDDEV=1.0 will give upper and lower bounds that are one standard deviation above and below the central response. PERCENTILES is the default.
WEIGHTS=series of (relative) weights on the draws (from importance sampling) [equal weights]
Note that this can dramatically slow down the calculation if you ask for percentiles and have many (5000+) draws.
These are the outputs. IRF is required, the other three are options. Each generates a RECT[SERIES], where x(i,j) is the series of responses of variable i to shock j.
IRF=RECT[SERIES] with the central measure of the impulse responses
VAR=RECT[SERIES] with the variances of the IRF's
LOWER=RECT[SERIES] with the lower bounds
UPPER=RECT[SERIES] with the upper bounds
mcprocessirf.src- Procedure file
- (6.48 KiB) Downloaded 88 times
@MCPROCESSIRF(MODEL=model used in generating responses,other options)
CENTER=[MEAN]/MEDIAN/INPUT
IMPULSES=RECT[SERIES] with input "center" for IRF's
These determine what is to be used as the point estimate of the IRF. The default is MEAN, which will be the mean across simulations. MEDIAN requests the median. If CENTER=INPUT, you must input the desired point estimates using the IMPULSES option---typically the impulse responses at the original estimates for the VAR.
PERCENTILES=||percentiles for lower and upper bounds|| [||.16,.84||]
STDDEV=# of standard deviations from mean for lower and upper bounds [not used]
STDDEV is used for doing error bands based upon multiples of the sample standard deviations. For instance, STDDEV=1.0 will give upper and lower bounds that are one standard deviation above and below the central response. PERCENTILES is the default.
WEIGHTS=series of (relative) weights on the draws (from importance sampling) [equal weights]
Note that this can dramatically slow down the calculation if you ask for percentiles and have many (5000+) draws.
These are the outputs. IRF is required, the other three are options. Each generates a RECT[SERIES], where x(i,j) is the series of responses of variable i to shock j.
IRF=RECT[SERIES] with the central measure of the impulse responses
VAR=RECT[SERIES] with the variances of the IRF's
LOWER=RECT[SERIES] with the lower bounds
UPPER=RECT[SERIES] with the upper bounds