gmm1 Generalized Method of Moments gmmCcapmJ

Grocer >> Generalized Method of Moments > gmmAndMon

gmmAndMon

Andrews-Monahan HAC estimators

CALLING SEQUENCE

[band,uout,D,eflag] = gmmAndMon(gmmopt,e,M,Z)

PARAMETERS

Input

• gmmopt tlist from GMM estimation which revelant arguments are

  * gmmopt('prt') = printing (0=none)     [1]

  * gmmopt('aminfo')= a list with fields:

       - aminfo('p') = autoregressive lags (p<=1 for arma models) [1]

       - aminfo('q') = moving average lags [0]

     - aminfo('vardum')= 1 for VAR(p) (overrides q) [0]

      - aminfo('nowhite') = 1 suppresses pre-whitening [0]

      - aminfo('diagdum') =1 for diagnostic graphs [0]

• e = matrix of residuals (we want its spectral density)

• M = Jacobian (derivative of moments)

• Z = instruments

Output

• band = bandwidth determined from the data

• uout = matrix of whitened residuals

• D = matrix to 're-color' spectral density of whitened resids

• eflag = 1 if an error in estimating ARMA model, else 0

DESCRIPTION

Calculates objects needed by gmmS for Andrews-Monahan HAC estimate.

EXAMPLE

[band,u,D,amerr] = gmmAndMon(gmmopt,e,M,Z);
//Example taken from gmmS:

AUTHOR

Emmanuel Michaux 2007
gmm1 Generalized Method of Moments gmmCcapmJ