Grant detail

GAUK (852013/2013)-Modeling time series dependence with dynamic copulas and high frequency data

Principal investigator: Mgr. Krenar Avdulaj Ph.D.
Collaborators: doc. PhDr. Jozef Baruník Ph.D.
Description: Modern risk management techniques require implementation of more complex models than the ones assuming linear correlations. The latter has been the foundation of risk models for many decades. However, it has been shown that linear correlations are not enough to describe dependence structure when the multivariate distribution of variables of interest is non-elliptical. Thus, the use of copula based dependence measures like rank correlation and tail dependence, are more realistic measures for a sound model. The methodology we employ has been developed recently and it extends the theory of copulas allowing for conditioning variables. It measures the dependence by the use of a two-step procedure: first modelling the marginal distributions and then utilizing the dynamic copulas. This project contributes the current state of literature by researching the multivariate dependence using the high frequency data and dynamic copulas. We will aim at explaining multivariate distribution of stock market returns utilizing the increasing availability of high frequency data and the flexibility of copula functions. Employing dynamic copulas will allow us to study the dynamic dependence as well.
Work in grant:
Web link:
Finance: Financed by the Grant Agency of the Charles University (GAUK) starting from year 2013.
End date: 2013

Can we still benefit from international diversification? The case of the Czech and German stock markets


2nd IES Economic Meeting: Modelling in Macroeconomics and Finance

7th International Conference on Computational and Financial Econometrics (CFE 2013)




Patria Finance