A0333
Title: The nature of ownership and stock returns
Authors: Daniel Weagley - Georgia Institute of Technology (United States)
John Kim - Emory University (United States)
Soohun Kim - KAIST (Korea, South) [presenting]
Abstract: Investment strategies vary in their reliance on characteristics versus intangible information to make investment decisions. We propose a simple machine learning methodology for estimating a strategy's reliance on characteristics or intangible information in making stock selection decisions. Using this methodology, we find stocks held by active mutual funds employing strategies more reliant on intangible information outperform stocks held by funds more reliant on characteristics. A long-short strategy based on the nature of stock ownership earns a CAPM-alpha of 4.8 pps per year (t-statistic of 3.54), and exhibits similar outperformance relative to other factor models. We find especially strong out-performance (19.4 pps per year, t-statistic > 9) within those stocks experiencing relative increases in mutual fund ownership breadth.
