Taming the Factor Zoo

The article discusses a new methodology for testing and evaluating potential new factors for asset pricing, which is important due to the large number of potential factors (referred to as the "factor zoo"). The authors propose a framework to select from among the many factors found in the literature and to discipline the proliferation of factors. They use a double-selection LASSO approach, which draws from machine learning, to systematically select the best possible control model out of the existing, known set of factors. They apply this methodology to a large set of factors proposed in the past 30 years to evaluate the marginal contribution of new factors proposed in the past five years. The results show that several new factors make a significant contribution to the explanatory power for expected returns and that the evaluation of the usefulness of these significant factors is robust across different samples. This methodology can significantly reduce the number of factors deemed significant.

Cochrane (2011)

Harvey et al. (2015)

McLean and Pontiff (2016)

Hou et al. (2017)

Cochrane (2009)

Chernozhukov et al. (2015)

Belloni et al. (2014b)