Message-ID: <CAJ3qPLCSk1FhRNG1r9hujusGaVGczoKgN9YNWybcBkccvf3aDA@mail.gmail.com>
Date: 2011-09-16T06:54:44Z
From: Alex Bird
Subject: CCCgarch.MM in PortfolioAnalytics
Greetings,
I have found a function in PortfolioAnalytics called CCCgarch.MM
which computes comoments based on CCC-Garch.
Does any one here has any example of using it in portfolio opimization?
I think it could be used in maximazing some kind of utility functions
like the following one
tcara<-function(r, w, lambda, gamma, theta){
r.mean<-apply(r,2,mean)
r.covar<-covar(r)
r.coskew<-coskew(r)
r.cokurt<-cokurt(r)
uf<-(r.mean%*%w -
1/2 * lambda^2 *(w%*%r.covar%*%w) +
1/6 * gamma^3*(w%*%r.coskew%*%(w%x%w)) -
1/24* theta^4*(w%*%r.cokurt%*%(w%x%w%x%w)))
uf
}
where w - weights vector, r - returns matrix, lambda/gamma/theta -
preference coefficients.
But maybe be there are some other ways to use the moments in optimization.
I also have another thought to use the moment in maximazing Omega
ratio calculated based on Corrado and Su modification which allows to
price options using first four moments.
Any advice would be very helpful.
Thanks in advance!
Kind regards,
Alex