Fitting a DCC-GARCH with more than one external regressor per single process using rmgarch

Hi Alexios,

thanks for answering, yes the problem is the one you wrote.

I just tried to estimate the dcc using (in the sample) the two univariate
estimates, I included in the "fit" command (dccfit(...fit=,...)) a list
which contains the univariate estimated, and the error is the same.

Now, as you suggested, I tried to use a multifit object, so my code is:

spec

$`Electronic Technology`

*---------------------------------*
*       GARCH Model Spec          *
*---------------------------------*

Conditional Variance Dynamics 	
------------------------------------
GARCH Model		: eGARCH(1,1)
Variance Targeting	: FALSE 
Exogenous Regressor Dimension: 2

Conditional Mean Dynamics
------------------------------------
Mean Model		: ARFIMA(1,0,0)
Include Mean		: FALSE 
GARCH-in-Mean		: FALSE 

Conditional Distribution
------------------------------------
Distribution	:  std 
Includes Skew	:  FALSE 
Includes Shape	:  TRUE 
Includes Lambda	:  FALSE 


$`Non-Energy Minerals`

*---------------------------------*
*       GARCH Model Spec          *
*---------------------------------*

Conditional Variance Dynamics 	
------------------------------------
GARCH Model		: eGARCH(1,1)
Variance Targeting	: FALSE 
Exogenous Regressor Dimension: 2

Conditional Mean Dynamics
------------------------------------
Mean Model		: ARFIMA(1,0,0)
Include Mean		: FALSE 
GARCH-in-Mean		: FALSE 

Conditional Distribution
------------------------------------
Distribution	:  std 
Includes Skew	:  FALSE 
Includes Shape	:  TRUE 
Includes Lambda	:  FALSE 

mspec = multispec( spec )
mspec

*-----------------------------*
*     GARCH Multi-Spec        *
*-----------------------------*
Multiple Specifications	: 2
Multi-Spec Type			: equal

r=as.xts(sector.tf.list[[tf]][a,1:2])
multi.fit=multifit(mspec,r)
multi.fit

*----------------------------*
*     GARCH Multi-Fit        *
*----------------------------*
No. Assets :2
GARCH Multi-Spec Type : Equal
GARCH Model Spec
--------------------------
Model : eGARCH
Exogenous Regressors in variance equation:  2 

Mean Equation :
Include Mean :  0
AR(FI)MA Model : (1,d,0)
GARCH-in-Mean :  FALSE
Exogenous Regressors in mean equation: none
Conditional Distribution:  std 

GARCH Model Fit
--------------------------
Optimal Parameters:
        Electronic Technology Non-Energy Minerals
ar1                   0.05294             0.05204
omega                -0.48544            -0.51636
alpha1               -0.01188            -0.00150
beta1                 0.96834             0.96459
gamma1                0.16409             0.16298
vxreg1               -0.00957             0.04071
vxreg2                0.00126             0.01153
shape                 3.66710             3.98666
Log-Lik           81768.21938         76305.84698

multispec=dccspec(mspec,model="DCC")
dccfit(multispec,fit=multi.fit,r)

Error in `rownames<-`(`*tmp*`, value = c("mu", "ar1", "ma", "arfima",  : 
  length of 'dimnames' [1] not equal to array extent

The error is the same.

You could try to run the code below, which reports the same error:

set.seed(422)

r=cbind(rnorm(1000),rnorm(1000))
colnames(r)=c("1","2")

ext=list()
ext[["1"]]=cbind(rnorm(1000),rnorm(1000))
ext[["2"]]=cbind(rnorm(1000),rnorm(1000))

spec=list()

for(i in 1:length(ext)){
  name=names(ext)[i]
 
spec[[name]]=ugarchspec(variance.model=list(model="eGARCH",garchOrder=c(1,1),external.regressors=ext[[name]]),
                                              mean.model = list(armaOrder =
c(1, 0),include.mean = F,external.regressors=NULL),
                                              distribution.model="norm")
}



mspec = multispec( spec )
multi.fit=multifit(mspec,r)
multispec=dccspec(mspec,model="DCC")

dccfit(multispec,fit=multi.fit,r)


Best regards,
Leopoldo Catania.



--
View this message in context: http://r.789695.n4.nabble.com/Fitting-a-DCC-GARCH-with-more-than-one-external-regressor-per-single-process-using-rmgarch-tp4675275p4675284.html
Sent from the Rmetrics mailing list archive at Nabble.com.