On Selection of Cross-Section Averages in Non-stationary Environments

TL;DR

This paper investigates the effectiveness of information criteria in selecting cross-section averages in factor models, revealing that they underperform in non-stationary environments despite prior assumptions of their robustness.

Contribution

It provides a formal analysis and simulation evidence showing the limitations of information criteria in non-stationary factor models, challenging previous claims of their general applicability.

Findings

IC underselects in non-stationary environments

Simulation results confirm severe underselection issues

Challenges assumptions of IC robustness in non-stationary settings

Abstract

Information criteria (IC) have been widely used in factor models to estimate an unknown number of latent factors. It has recently been shown that IC perform well in Common Correlated Effects (CCE) and related setups in selecting a set of cross-section averages (CAs) sufficient for the factor space under stationary factors. As CAs can proxy non-stationary factors, it is tempting to claim such generality of IC, too. We show formally and in simulations that IC have a severe underselection issue even under very mild forms of factor non-stationarity, which goes against the sentiment in the literature.