From the Financial Analysts Journal, Volume 80, 2024 - Issue 2:
Abstract
Using simulated historical backtests, we study the impact of stock exclusions on the performance of passive and active portfolios. We find that at low to moderate numbers, stock exclusions have very little influence on passive portfolios. Their effects on active portfolios vary by the factor in consideration and the portfolio construction method, but the magnitudes are much smaller than suggested by the percentage of stocks being excluded. We find similar patterns with industry-concentrated exclusions. Overall, our results suggest that investors should feel comfortable excluding a fairly large number of stocks before experiencing any significant deterioration in their investment performance.
Introduction
Technological advancements coupled with lower trading costs have ushered in a new era of equity investing through direct indexing. The term “direct indexing” refers to a portfolio that directly holds the underlying stocks of a published index rather than owning a comingled index fund, such as a mutual fund or an exchange-traded fund (ETF). Initially introduced within larger separately managed accounts (SMAs), the recent enablement of trading fractional shares has allowed even smaller retail accounts to participate in direct indexing through large brokerages such as Schwab and Fidelity.The benefits of directly owning a basket of stocks are not truly realized until one implements “personalized indexing” or custom portfolios, which is the next generation of direct indexing.Footnote1 Personalization allows individual investors to deviate, potentially materially, from the benchmark indices that direct indexing replicates. In theory, the degrees of customization are no less broad than for any equity portfolio being built from the ground up. In current practice, personalization typically enables investors the following selections: (1) tax loss harvesting (TLH) in taxable accounts, (2) factor enhancements with the intent of improving the portfolio’s return or risk characteristics, (3) socially responsible investing (SRI) restrictions, and (4) single-stock exclusions.
Three out of four of these areas of customization have a commonality. When enabled, they restrict a portfolio from owning individual stocks that it might have otherwise owned if unconstrained. The wash-sale rule restricts owning a stock for 30 days after harvesting a loss, assuming the investor seeks to capture the realized loss. Socially responsible investing restrictions prevent ownership of potentially large sets of stocks that meet certain predefined criteria. Single-stock exclusions allow investors to restrict ownership of individual names for regulatory reasons or due to personal preference.
What impact do such restrictions have on a portfolio’s expected performance? In terms of direction, it is hard to dispute the math and the associated intuition. A binding restriction prevents a portfolio from owning what it would have owned unrestricted. That cannot help in expectation. If the restriction does not bind (you would not have purchased the unrestricted stock anyway), then it will not hurt or have any impact whatsoever. Otherwise, the restriction’s expected impact is negative. Asness (Citation2017) vigorously articulates this case when disputing the claims many ESG portfolio managers make about their values-based restrictions not hurting and potentially even improving portfolio performance.
A binding restriction must hurt expected performance along some dimension (risk, return, or some other objective that a portfolio manager has chosen to optimize). But how bad is the hurt? Is it a papercut or a fatal wound?....
....MUCH MORE
Also at the FAJ:
Rob Arnott on "Smart Rebalancing"
