Events and Meetings of Italian Statistical Society, Advances in Latent Variables - Methods, Models and Applications

Font Size: 
Macroeconomic forecasting through regularized reduced-rank regression
Gianluca Cubadda, Emmanuela Bernardini

Last modified: 2013-06-14

Abstract


This paper proposes a strategy to detect and impose reduced-rank restrictions in large multivariate time series models. In this framework, Cubadda and Hecq[8] have recently shown that Canonical Correlation Analysis (CCA) does not perform well. We suggest to use proper shrinkage estimators of the variance-covariance matrices that are involved in CCA, thus obtaining a method that is asymptotically equivalent to CCA, but it is numerically more stable in finite samples. Simulations and empirical applications document the merits of the proposed approach.

Full Text: PDF