Mulivariate Reduced Rank Regression in Non-Gaussian Contexts, Using Copulas

CORE Discussion Paper 2004/32

Posted: 4 Jan 2005

See all articles by Andréas Heinen

Andréas Heinen

University of Cergy-Pontoise - THEMA

Erick W. Rengifo

Fordham University - Department of Economics - Center for International Policy Studies (CIPS)

Date Written: 2004

Abstract

We propose a new procedure to perform Reduced Rank Regression (RRR) in non-Gaussian contexts, based on Multivariate Dispersion Models. Reduced-Rank Multivariate Dispersion Models (RR-MDM) generalise RRR to a very large class of distributions, which include continuous distributions like the normal, Gamma, Inverse Gaussian, and discrete distributions like the Poisson and the binomial. A multivariate distribution is created with the help of the Gaussian copula and stimation is performed using maximum likelihood. We show how this method can be amended to deal with the case of discrete data. We perform Monte Carlo simulations and show that our estimator is more efficient than the traditional Gaussian RRR. In the framework of MDM's we introduce a procedure analogous to canonical correlations, which takes into account the distribution of the data.

Keywords: multivariate dispersion model, multivariate statistical analysis

JEL Classification: C35, C39

Suggested Citation

Heinen, Andréas and Rengifo, Erick W., Mulivariate Reduced Rank Regression in Non-Gaussian Contexts, Using Copulas (2004). CORE Discussion Paper 2004/32. Available at SSRN: https://ssrn.com/abstract=640322

Andréas Heinen (Contact Author)

University of Cergy-Pontoise - THEMA ( email )

33 boulevard du port
F-95011 Cergy-Pontoise Cedex, 95011
France

Erick W. Rengifo

Fordham University - Department of Economics - Center for International Policy Studies (CIPS) ( email )

United States
0017188174061 (Phone)
0017188173518 (Fax)

Register to save articles to
your library

Register

Paper statistics

Abstract Views
665
PlumX Metrics