Mean Square Errors of factors extracted using principal components, linear projections, and Kalman filter

Matteo Barigozzi, Diego Fresoli, Esther Ruiz · 2026

Read Mean Square Errors of factors extracted using principal components, linear projections, and Kalman filter by Matteo Barigozzi, Diego Fresoli, Esther Ruiz (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-04087.txt.

Description

Matteo Barigozzi, Diego Fresoli, and Esther Ruiz compare how well two standard tools recover hidden factors from many variables at once. Principal components treat the true factors as fixed, while a Kalman filter treats them as random over time. Knowing the error in recovered factors matters when factors are read directly or used to summarise predictors.

For finite numbers of variables, the three authors show principal component factors carry larger mean square error than Kalman filter factors. The gap holds whether the random factors are independent over time or correlated. They also study what happens when leftover parts are wrongly assumed to share one variance or to be uncorrelated. Simulations show how the errors affect confidence intervals built around the factors.

Catalog record

CREATORMatteo Barigozzi, Diego Fresoli, Esther Ruiz
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE102 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-04087

Browse the full library · About the Tower of Babel