[S,DATA,B,COUNT] = SVAR(V,DATA,...)
V
[ VAR ] - Reduced-form VAR object.
DATA
[ struct | tseries ] - Data associated with the input VAR object.
S
[ VAR ] - Structural VAR object.
DATA
[ struct | tseries ] - Data with transformed structural residuals.
B
[ numeric ] - Impact matrix of structural residuals.
COUNT
[ numeric ] - Number of draws actually performed (both successful and unsuccessful) when 'method'='draw'
; otherwise COUNT=1
.
'maxIter='
[ numeric | 0
] - Maximum number of attempts when 'method'='draw'
.
'method='
[ 'chol'
| 'householder'
| 'qr'
| 'svd'
] - Method that will be used to identify structural VAR and structural shocks.
'nDraw='
[ numeric | 0
] - Target number of successful draws when 'method'='draw'
.
'reorder='
[ numeric | empty ] - Reorder VAR variables before identifying structural shocks, and bring the variables back in original order afterwards. Use the option 'backorderResiduals='
to control if also the structural shocks are to be brought back in original order.
'output='
[ 'auto'
| 'dbase'
| 'tseries'
] - Format of output data.
'progress='
[ true
| false
] - Display progress bar in the command window.
'rank='
[ numeric | Inf
] - Reduced rank of the covariance matrix of structural residuals when 'method=' 'svd'
; Inf
means full rank is preserved.
'backOrderResiduals='
[ true
| false
] - Bring the identified structural shocks back in original order after identification; works with 'reorder='
.
'std='
[ numeric | 1
] - Std deviation of structural residuals; the resulting structural covariance matrix will be re-scaled (divided) by this factor.
'test='
[ char ] - Works with 'method=draw'
only; a string that will be evaluated for each random draw of the impact matrix B
. The evaluation must result in true
or false
; only the matrices B
that evaluate to true
will be kept. See Description for more on how to write the option 'test='
.
The structural impact matrices B
are randomly generated using a Householder transformation algorithm. Each matrix is tested by evaluating the test
string supplied by the user. If it evaluates to true the matrix is kept and one more SVAR parameterisation is created, if it is false the matrix is discarded.
The test
string can refer to the following characteristics:
S
-- the impulse (or shock) response function; the S(i,j,k)
element is the response of the i
-th variable to the j
-th shock in period k
.
Y
-- the asymptotic cumulative response function; the Y(i,j)
element is the asumptotic (long-run) cumulative response of the i
-th variable to the j
-th shock.