How to interpret the outputs of DCC Multivariate GARCH

50 vues (au cours des 30 derniers jours)
Esmail
Esmail le 29 Mai 2015
Commenté : Zhitao Zhou le 28 Juil 2018
Hello everybody,
I want to run a dcc.m code of the MFE Kevin Sheppard toolbox by giving the following code:
[PARAMETERS,LL,HT,VCV,SCORES]= dcc(DATA,[],1,0,1)
I've got 4 variables (see attached file). By runing the code, everything goes well and I get the estimated parameters in "PARAMETERS". But I'm having difficulties in understanding the outputs "VCV" and "SCORES". The explanations in the code are not sufficient.
I know that a DCC Multivariate GARCH is designed as follows:
DATA=H(t)^1/2*epsilon(t)
H(t)=D(t)*R(t)*D(t) and R(t)= diag[Q(t)^1/2] * Q(t) * diag[Q(t)^1/2]
where R(t) peresents the conditional correlation matrix.
I want finally to plot the dynamic correlations, in other words I have to plot the values of the R(t) matrix. But where is the R(t) in the outputs of this code?
Can anybody help me please?

Réponses (3)

Lorenzo Orlando
Lorenzo Orlando le 12 Juin 2017
Modifié(e) : Lorenzo Orlando le 21 Juin 2017
VCV is the correlation matrix of parameters, needed to calculate standard errors as sqrt(diag(VCV)).

Weidong Lin
Weidong Lin le 12 Jan 2018
Actually the Rt is in the output of the function dcc_likelihood(), which is used inside of the dcc(). If you want to display the Rt, just simply add 'Rt' into the output option, i.e., [parameters, ll ,Ht, Rt, VCV, scores, diagnostics]=dcc(...)
Hope it works for you.
  1 commentaire
Zhitao Zhou
Zhitao Zhou le 28 Juil 2018
Hi, I am doing CCC GARCH model forecasting. Do you know How could I do the one-step-ahead forecasting after I fitting the model with ccc_mmvgarch.m file?

Connectez-vous pour commenter.


RP
RP le 13 Jan 2018
Hi,
I am using matlab 2017 version. I have 1 query that is MFE-toolbox additionally installed in matlab 2017 version?
Please suggest.
  2 commentaires
Weidong Lin
Weidong Lin le 13 Jan 2018
Just download the toolbox from Kevin's page. Then open your Matlab and type 'pathtool' in the command window, add the folder and subfolder of the MFE toolbox into the path.
Always check the path every time you see any errors when you use the toolbox.
RP
RP le 27 Jan 2018
Modifié(e) : RP le 27 Jan 2018
Thank you. It works.

Connectez-vous pour commenter.

Catégories

En savoir plus sur Conditional Variance Models dans Help Center et File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by