Abstract
This article provides the first empirical application of the dynamic equicorrelation (DECO) model to a cross-market data set composed of equities, bonds, foreign exchange and commodity returns during 1983–2013. The results reveal that the average cross-market equicorrelation is around 47%, although it is found to be time-varying and mean-reverting. Besides, we display the equicorrelation across markets as a natural way of looking at the DECO dynamics, which overcomes the cumbersome estimation difficulties encountered with multivariate GARCH models. Implications are derived in terms of asset management.
Notes
1 VECH denotes the operator that stacks the lower triangular portion of a symmetric N × N matrix into an N(N+1)/2 × 1 vector of the corresponding unique elements.