RT Generic
T1 Detecting big structural breaks in large factor models
A1 Chen, Liang
A1 Dolado, Juan José
A1 Gonzalo, Jesús
A2 Universidad Carlos III de Madrid. Departamento de Economía, 
AB Time invariance of factor loadings is a standard assumption in the analysis of largefactor models. Yet, this assumption may be restrictive unless parameter shifts aremild (i.e., local to zero). In this paper we develop a new testing procedure todetect big breaks in these loadings at either known or unknown dates. It relies upontesting for parameter breaks in a regression of the first of the r¯ factors estimated byPCA on the remaining r¯ − 1 factors, where r¯ is chosen according to Bai and Ng’s(2002) information criteria. The test fares well in terms of power relative to otherrecently proposed tests on this issue, and can be easily implemented to avoidforecasting failures in standard factor-augmented (FAR, FAVAR) models where thenumber of factors is a priori imposed on the basis of theoretical considerations.
SN 2340-5031
YR 2011
FD 2011-12
LK https://hdl.handle.net/10016/14000
UL https://hdl.handle.net/10016/14000
LA eng
DS e-Archivo
RD 20 may. 2024