In large-scale panel data models with latent factors the number of factors and their loadings may change over time. This paper proposes an adaptive group-LASSO estimator that consistently determines the numbers of pre- and post-break factors and the stability of factor loadings. The data-dependent LASSO penalty is customized to account for unobserved factors and an unknown break date. A novel feature of our estimator is its robustness to unknown break dates. Existing procedures either overestimate the number of factors by neglecting the breaks or require known break dates for a subsample analysis. In an empirical application, we study the change in factor loadings and the emergence of new factors during the Great Recession. Joint with Zhipeng Liao and Frank Schorfheide.

JEL Classication: C13, C33, C52
Keywords: Great Recession, High-dimensional Model, Large Data Sets, LASSO, Latent
Factor Model, Model Selection, Shrinkage Estimation, Structural Break