A1089
Title: A Knockoff filter approach to asset allocation
Authors: Arman Hassanniakalager - University of Bath (United Kingdom) [presenting]
Abstract: The purpose is to investigate using the Knockoff filter as a novel to optimise long-only equity portfolios by controlling false discoveries and cross-sectional dependence among US equities. Three conventional strategies of value investing, growth investing, and investing in companies by largest market capitalisation are considered benchmark strategies. Using the FamaFrenchCarhart factor analysis, the benefits of the Knockoff regression in generating excess returns alpha and mitigating factor-related risks embedded in common investment strategies are quantified.
