“…Although these models are merely approximations of reality (Barillas and Shanken (2018), Gospodinov, Kan, and Robotti (2013), Kan and Robotti (2009), and Kan, Robotti, and Shanken (2013)), it is important from an academic and practitioner perspective to know which model provides the best overall description of asset returns. For example, there is ample evidence, both empirical and anecdotal, that portfolio managers most often use the capital asset pricing model and a variety of multifactor models to compute expectations of returns (see, among others, Ang (2014), Brealey, Myers, and Allen (2016), Fischer and Wermers (2012), Gitman and Mercurio (1982), Graham and Harvey (2001), Grinold and Kahn (1995), and Jagannathan and Meier (2002)). In relation to this, Fama and French (2016) compare the performance of the recently proposed 5-factor model of Fama and French (2015), along with models that use subsets of its factors.…”