A1681
Title: Multiple testing for asset pricing factor models
Authors: Florian Richard - Universite Laval (Canada) [presenting]
Lynda Khalaf - Carleton (Canada)
Abstract: An empirical assessment of the current beta-pricing literature is offered, using non-nested tests for multivariate models and a model confidence set (MCS) approach. Both methods can be used to assess either: (i) the statistical significance of a newly proposed non-nested model, or (ii) the statistical equivalence of their predictions, in the sense of equal predictive ability. The MCS procedure of a prior study with the empirical approach is reconciled. It is found that the non-nested test rejects many models empirically, while the MCS approach favours the Fama and French model. Notably, models based on machine learning algorithms are rejected by both methods. The findings suggest that further improvements are necessary in the effort to price the cross-section of expected returns, as the models formed from the winning factors in the literature still show evidence of misspecification.
