Too good to be true? Fallacies in evaluating risk factor models

• 成果类型: Article
• 署名作者: Gospodinov, Nikolay; Kan, Raymond; Robotti, Cesare
• 署名单位: Federal Reserve System - USA; Federal Reserve Bank - Atlanta; University of Toronto; University of Warwick
• 刊物名称: JOURNAL OF FINANCIAL ECONOMICS
• ISSN/ISSBN: 0304-405X
• DOI: 10.1016/j.jfineco.2018.10.012
• 发表日期: 2019
• 页码: 451-471
• 关键词: Asset pricing; Spurious risk factors; Unidentified models; Model Misspecification; Maximum Likelihood; Goodness of fit; Rank test

摘要

This paper is concerned with statistical inference and model evaluation in possibly misspecified and unidentified linear asset pricing models estimated by maximum likelihood. Strikingly, when spurious factors (that is, factors that are uncorrelated with the returns on the test assets) are present, the model exhibits perfect fit, as measured by the squared correlation between the model's fitted expected returns and the average realized returns. Furthermore, factors that are spurious are selected with high probability, and factors that are useful are driven out of the model. While ignoring potential misspecification and lack of identification can be very problematic for models with macroeconomic factors, empirical specifications with traded factors (e.g., Fama and French, 1993; Hou et al., 2015) do not suffer from the identification problems shown in this study. (C) 2018 Elsevier B.V. All rights reserved.