Fund Flows and Market States
Review of Financial Studies, 2017, DOI:10.1093/rfs/hhx015
82 Pages Posted: 13 May 2013 Last revised: 19 Jun 2017
Date Written: October 6, 2016
This paper establishes a new empirical fact: mutual funds' flow-performance sensitivity is a hump-shaped function of aggregate risk-factor realizations. Explanations based on extant theories can only explain a fraction of the pattern. We thus develop a new parsimonious model. It assumes Bayesian investors who are uncertain about the degree to which fund returns are exposed to systematic risk. Fund performance is then less informative about manager skill when factor realizations are larger in absolute value. The data also support the out-of-sample prediction that the hump shape is more pronounced for funds with more uncertain risk loadings.
Keywords: Bayesian learning, parameter uncertainty, mutual funds, flow-performance, Kalman filter, beta
JEL Classification: G00, G20
Suggested Citation: Suggested Citation