Victor DeMiguel, London Business School, will discuss his paper.
Abstract
Identifying outperforming mutual funds ex-ante is a notoriously difficult task. We use machine learning to exploit numerous fund characteristics and construct portfolios of equity funds that earn out-of-sample annual alpha of 4.2% net of costs. We show that such performance is the joint outcome of both exploiting multiple fund characteristics and allowing for flexibility in the relation between characteristics and performance. We demonstrate that even retail investors can benefit from investing in actively managed funds. The performance of our portfolios has declined over time, however, consistent with increased competition in asset markets and diseconomies of scale at the industry level.