Spurious Factor Analysis

This paper draws parallels between the principal components analysis of factorless high‐dimensional nonstationary data and classical spurious regression. We show that a few such absorb nearly all variation. The corresponding scree plot suggests contain factors, which is corroborated by standard panel information criteria. Furthermore, Dickey–Fuller tests unit root hypothesis applied to estimated “idiosyncratic terms” often reject, creating an impression factors are responsible for most nonstationarity in data. warn empirical researchers these peculiar effects suggest always compare levels with differences.