Tue Dec 01, 09 10:53 AM (NEW!) Hi, I have a proble with DCC garch of USCD toolbox,I tried to use dcc_mvgarch and dcc_mv_likelihood to estimate a correlation process between the two series in serie.mat.I think there are something wrong, beacuse don't look like a correlation process.Do you know what is the problem?Can you help me please?%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%clear allclose allload serie.matserie=data;serie=price2ret(serie);p=1;q=1;[Coeff,Errors,LLF,Innovations1,Sigmas,Summary] = garchfit(serie(:,1));[Coeff,Errors,LLF,Innovations2,Sigmas,Summary] = garchfit(serie(:,2));residui(:,1)=Innovations1;residui(:,2)=Innovations2;[parameters, loglikelihood, Ht, Qt, stdresid, likelihoods, stderrors, A,B, jointscores]=dcc_mvgarch(residui,1,1,1,1);[logL, Rt, likelihoods, Qt]=dcc_mvgarch_likelihood(parameters, stdresid,1,1);%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%Thank youStefano