In this article, we put forward a generalization of the Dynamic Conditional Correlation (DCC) Model of Engle (2002). Our model allows for asset-specific correlation sensitivities, which is useful in particular if one aims to summarize a large number of asset returns. We propose two estimation methods, one based on a full likelihood maximization, the other on individual correlation estimates. The resultant generalized DCC (GDCC) model is considered for daily data on 39 U.K. stock returns in the FTSE. We find convincing evidence that the GDCC model improves on the DCC model and also on the CCC model of Bollerslev (1990).

dynamic conditional correlation, multivariate GARCH
Semiparametric and Nonparametric Methods (jel C14), Time-Series Models; Dynamic Quantile Regressions (jel C22),
Econometric Institute Reprint Series
Econometric Reviews
Erasmus School of Economics

Hafner, C.M, & Franses, Ph.H.B.F. (2009). A Generalized Dynamic Conditional Correlation Model: simulation and application to may assets. Econometric Reviews, 28(6), 612–631. doi:10.1080/07474930903038834