A Dynamic Conditional Score Model for the Log Correlation Matrix

Seminar abstract

This presentation proposes a new model for the dynamics of correlation matrices, where the dynamics are driven by the likelihood score with respect to the matrix logarithm of the correlation matrix.

In analogy to the exponential GARCH model for volatility, this transformation ensures that the correlation matrices remain positive definite, even in high dimensions.

For the conditional distribution of returns it is assumed a student-t copula to explain the dependence structure and univariate student-t for the marginals with potentially different degrees of freedom. 

The separation into volatility and correlation parts allows two-step estimation, which facilitates estimation in high dimensions.

The researchers derive estimation theory for one-step and two-step estimation. 

In an application to a set of six asset indices including financial and alternative assets we show that the model performs well in terms of various diagnostics and specification tests.

Booking

This seminar is free to attend with no need to book in advance. We warmly invite you to share with your friends, colleagues and classmates.

Join this Seminar on Wednesday 10 February 2021 at 1pm

Speaker bio

Christian Hafner is a Professor of Econometrics at Université catholique de Louvain (UCL).

He serves the editorial board of Studies in Nonlinear Dynamics and Econometrics, Computer Statistics, Banking and Finance Review and International Econometric Review.

Christian is a prolific researcher. His research has been published on top econometrics/economics journals, including;

• Journal of Econometrics
• Journal of Business and Economic Statistics
• Econometric Theory 
• Journal of Applied Econometrics 

as well as many more.