Anmerkungen:
Nach Informationen von SSRN wurde die ursprüngliche Fassung des Dokuments December 15, 2022 erstellt
Beschreibung:
Kozak and Nagel (2022) theoretically show how GLS factors, slopes of monthly cross-sectional GLS regressions of returns on characteristics, perfectly span the mean-variance frontier. We provide an empirical design to recover feasible GLS factors clustering the covariance of returns at the industry level. In line with theory, our feasible GLS factors achieve a better MV spanning than the Fama and French (2020) OLS factors, slopes of monthly cross-sectional OLS regressions of returns on characteristics, and their heuristic hedging improvement proposed by Daniel, Mota, Rottke and Santos (2020). Moreover, we show the same performance can be more parsimoniously achieved with less number of factors extracted