<< irf_mc1 VAR estimations rolvar >>

Grocer >> VAR estimations > rolirf

rolirf

Calculates Impulse Response Function for rolling VAR

CALLING SEQUENCE

res=rolirf(results,S,arg1,...,argn)

PARAMETERS

Input

* results = results tlist returned by VAR

* S = scalar for number of periods in IRF

* arg1,...,argn = optional arguments:

  - 'mres=x' where x = chol1 (cholesky decomposition) x = chol2 (triangular factorisation) x = original (original residuals) (default = chol1)

  - 'meth=x' where x = asym (asymptotic formula) x = mc1 (Monte-Carlo simulations using draws from the coefficients) (default = asym)

  - 'niter=x' where x= # iterations for the Monte-Carlo simulations (if any; default=1000)

  - 'size=x' where x = significance level for the confidence band (default =0.05)

 

Output

* res = a results tlist with

  - res('meth') = 'rolling irf'

  - res('rolling var results') = the rolling Var results tlist entered as input to rolrif

  - res('mres') = decomposition method

  - res('T') = # of periods represented

  - res('IRF') = (nrolling x (S+1) x T) impulse response functions (with nrolling = # of rolling VAR estimated)

  - res('IRF_LOW') = (nrolling x (S+1) x T) lower range of impulse response confidence band

  - res('IRF_UPP') = (nrolling x (S+1) x T) upper range of impulse response confidence band

  - res('PHI') = (nrolling x N*p x T) matrix of coefficients

  - res('msg') = message inidicating the nature of the decomposition

  - res('size') = size of the confidence band

  - res('ans_var 1 to_shock 1') ... until res('ans_var n to_shock n') = a (nrolling x N*p x T) matrix collecting all irf of one variabel at a time to a shock at a time

DESCRIPTION

Calculates Impulse Response Function for rolling VAR.

EXAMPLE

global GROCERDIR;
load(GROCERDIR+'\data\france_maunf_survey.dat')
bounds()
resrolvar=rolvar(6,'endo=fra_past_output;fra_exp_output','nper=120','start_firstbound=1990m1','end_firstbound=2004m1')
resirf=rolirf(resrolvar,20)
// Example taken from function rovar_d; after estimating rolling VAR estimates with France past output balance and France expected output balance
// with estimations starting in 1990m1 for 120 periods, then in 1990m2 for 120 periods ... until 2004m1 for 120 periods
// and 6 lags, calculates the corresponding IRFs until the 20th period

AUTHOR

Éric Dubois 2014

Report an issue
<< irf_mc1 VAR estimations rolvar >>