A structural factor-augmented vector error correction (SFAVEC) model approach: an application to the UK

Abstract

This note presents a new structural factor-augmented vector error correction model approach to solve the limited information problem present in traditional vector error correction models. We apply this approach to the UK and obtain a reasonable characterization of the long-run equilibrium concerning real activity, taxation, inflation and the rate of interest.

Acknowledgements

An earlier version of this article was prepared while I was visiting scholar at the Department of Economics, University of Copenhagen, for the ESS (2005), where I have received many useful suggestions. I wish to thank Professors Bruno Bises, Marco Lippi and Pasquale M. Sgro for their constructive comments, and seminar participants at CREI, University of Roma Tre. All remaining errors are mine.

Notes

¹ These models were first introduced by Engle and Granger in 1987.

² The idea of partitioning a large panel of variables into subsets of similar variables and then estimating factors by static or dynamic factor analysis was traditionally applied only to stationary variables (Forni et al., 2003; Belviso and Milani, 2006). Bai and Ng (2004) stress, however, that there is nothing that restrict the factors to be all I(1) or all I(0).

³ Based on data inspection and formal unit root tests.

⁴ Results are available on request from the author.

⁵ We couldn't reject the Null hypothesis with a p-value of 0.33.

⁶ We couldn't reject the Null hypothesis of weak-exogeneity with a p-value of 0.50 and 0.30.

⁷ For the definitions of α and β, see Juselius (2005).