Francisco Barillas, Raymond Kan, Cesare Robotti, and Jay Shanken
We show how to conduct asymptotically valid tests of model comparison when the extent of model mispricing is gauged by the squared Sharpe ratio improvement measure. This is equivalent to ranking models on their maximum Sharpe ratios, effectively extending the GRS test to accommodate comparison of non-nested models. Mimicking portfolios can be substituted for any nontraded model factors and estimation error in the portfolio weights is taken into account in the statistical inference. A variant of the Fama and French (2018) 6-factor model, with a monthly-updated version of the usual value spread, emerges as the dominant model.