[R-SIG-Finance] copula with rmgarch

[alexios ghalanos]

1. The code in the rmgarch.tests folder is up to date. The code below 
that you quote is an outdated example and not from the rmgarch.tests 
folder but from the help page for cgarchsim (which I should update 
anyway when I find the time). Please only use the examples from the 
folder indicated (there are plenty!).

2. You are free to look inside the returned object or work with the 
extractor functions provided.
e.g.
slotNames(sim1)
names(sim1 at msim)

3. This is the 1-ahead simulation. Why would you expect the covariance 
and correlation to be uncertain? This is a non-bayesian setup which 
means that parameter uncertainty is not taken into account and hence the 
1-ahead conditional covariance is given by the GARCH type dynamics 
without recourse to any uncertainty etc.

-Alexios

On 08/05/2012 17:53, Alex Fei wrote:
> Thank you Alexios for quick reply!! The files in the 'rmgarch.tests' folder
> gave me a lot of help.
>
> Can I do 1-step ahead forecasting using GARCH-Copula with the help of
> rmgarch package? for example I need to get the returns of each assets and
> their covariance at T+1 based on the parameters estimated using in-sample
> data until T?
>
> I followed your Example in the help of cgarchsim. Please correct me if I am
> wrong:
> 1) the (mean) forecast returns should be the simmean1 in your example
> 2) I think rcov(sim1) only report the 1st cov out of 3500 simulations. Then
> how to get the mean forecast cov?
>
> I think the Example in ?cgarchsim has a copula of errors, although none of
> them is serious:
>> spec = cgarchspec(uspec = multispec( replicate(3, uspec) ), VAR = TRUE,
>> VAR.opt = list(lag = 1, lag.max = 4,
> lag.criterion = c("AIC", "HQ", "SC", "FPE"), external.regressors = NULL),
> dccOrder = c(1,1), distribution.model = list(copula = c("mvnorm"), method =
> c("ML"),
> time.varying = TRUE, transformation = "parametric"), start.pars = list(),
> fixed.pars = list())
>
> VAR.opt = list(lag = 1, lag.max = 4,  lag.criterion = c("AIC", "HQ", "SC",
> "FPE"), external.regressors = NULL) need to be lag = 1, lag.max = 4,
> lag.criterion = c("AIC", "HQ", "SC", "FPE"), external.regressors = NULL
>
>> sim1 = cgarchsim(fit1, n.sim = 1, n.start = 0, m.sim = 3500, presigma =
>> tail(sigma(fit1), 1),
> startMethod = "sample", preR = preR, prereturns = tail( as.matrix(Dat), 4),
> preresiduals = tail(residuals(.fitlist), 1),rseed = 1:3500)
>
> if I use prereturns = tail( as.matrix(Dat), 4) , it will report error. So
> instead, I use prereturns = tail( as.matrix(Dat), 1)
>
>> forcmean = round( rgarch:::varxforecast(X = Dat, Bcoef =
>> fit1 at mfit$vrmodel$Bcoef, p = 4,
> 				out.sample = 0, n.ahead = 1, n.roll = 0, mregfor = NULL), 5)
>
> the rgarch package is offline. So can you suggest another way for this?
>
> Thank you again!
>
> --
> View this message in context: http://r.789695.n4.nabble.com/copula-with-rmgarch-tp4616138p4618054.html
> Sent from the Rmetrics mailing list archive at Nabble.com.
>
> _______________________________________________
> R-SIG-Finance at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-finance
> -- Subscriber-posting only. If you want to post, subscribe first.
> -- Also note that this is not the r-help list where general R questions should go.
>