On Selection of Cross‐Section Averages in Non‐Stationary Environments
研究了信息准则在非平稳因子环境下选取横截面平均值的表现，发现因子越持久，小样本下性能越差，与现有文献观点相悖。
ABSTRACT Information criteria (ICs) have been widely used in factor models to estimate an unknown number of latent factors. It has recently been shown that ICs perform well in Common Correlated Effects (CCE) and related settings when 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 an excellent performance of ICs under general factors, too. We show formally and in simulations that they remain consistent, but the more persistent factors are, the poorer they perform in small samples, which goes against the sentiment in the CCE/CAs literature.