Discretionary Investing by ‘Passive’ S&P 500 Funds

Peter Molk is the John H. and Mary Lou Dasburg Professor of Law at the University of Florida Levin College of Law and  Adriana Z. Robertson is Donald N. Pritzker Professor of Business Law at the University of Chicago Law School.  This post is based on their recent paper. Related research from the Program on Corporate Governance includes The Agency Problems of Institutional Investors (discussed on the Forum here) by Lucian Bebchuk, Alma Cohen, and Scott Hirst; Index Funds and the Future of Corporate Governance: Theory, Evidence, and Policy (discussed on the Forum here) by Lucian Bebchuk and Scott Hirst; and The Specter of the Giant Three (discussed on the Forum here) by Lucian Bebchuk and Scott Hirst.

So-called “passive” index funds, which track a pre-specified underlying index, manage over $12 trillion in assets. It is widely assumed that the managers of these funds cannot select portfolios that deviate from the index’s holdings. In Discretionary Investing by ‘Passive’ S&P 500 Funds, we show this assumption is false as a matter of both law and empirical fact. We analyze funds that track the S&P 500 – the most widely used index – and find their holdings regularly diverge from the index’s, by between 1.7% and 7.5% in the fourth quarter of 2022 alone. These deviations amount to over $60 billion in discretionary investment decisions, roughly equivalent to Target Corporation’s entire market capitalization. Our findings complicate the standard narrative around index funds and weaken many of the criticisms traditionally levied against these funds.

The manager of an actively managed investment fund is expected to buy and sell securities in line with the fund’s overall investment strategy. When she does so, her goal is typically to outperform a benchmark index or to provide investors with some specialized investment strategy. In contrast, the paradigmatic passive index fund seeks simply to track some pre-determined index like the S&P 500. Academics, commentators, and the popular press widely assume that to track their underlying indices, these index funds must hold the assets of that index with, at most, trivial flexibility to deviate from the index’s holdings. For example, it has been asserted that these funds “buy stock in every company indiscriminately”[1] and that they “are locked into the portfolio companies they hold. They cannot exploit mispricing or other informational advantages through trading, nor can they follow the ‘Wall Street Rule’ and exit from underperforming companies the way traditional shareholders, particularly active funds, can.”[2]

This strait-jacketed characterization of index investing has led many to view index funds with skepticism. Some have argued that index funds are ineffective stewards of corporate governance, because their programmed investment mandate means they lack financial incentives to monitor and hold portfolio companies’ management accountable to investors. Others argue that investment in securities price discovery is reduced because index funds cannot overweight underpriced companies, leading to market inefficiencies from distorted stock prices. Some have even called for banning broad-market index funds as an investment instrument.

These concerns all stem from a common assumption that index funds must robotically replicate the holdings of their tracked index. We demonstrate that, at least with S&P 500 index funds – seen as the quintessential “passive” funds – this assumption is false.

As a matter of theory, funds face at most very loose constraints on their investment decisions. Federal securities laws, like prohibitions against material misrepresentations or the “Names” rule, either do not reach funds’ investment decisions, or else proscribe just the most egregious cases, like an “S&P 500 fund” with holdings that are nowhere close to the index. More moderate exercises of investing discretion do not fall within the scope of these prohibitions. In addition, funds do not voluntarily adopt particularly binding constraints in their prospectuses. These documents typically make similar, limited, commitments that the fund will invest “at least 80% of assets in common stocks included in the S&P 500,”[3] but not whether those assets will mirror the S&P 500’s weights, whether all S&P 500 stocks will be bought, or how the remaining 20% of assets will be invested.

As a matter of empirics, we find funds exercise their discretion to depart meaningfully from S&P 500 holdings. Departures from the index are most pronounced among smaller S&P 500 index funds. Systematic differences across S&P 500 funds imply that these funds are far from homogeneous. And while departures for the largest funds are smaller in percentage terms, they still average about 3% of assets under management. These departures occur in multiple ways: funds attaching different weights to S&P 500 companies than does the index; funds dropping S&P 500 companies from their portfolios; funds trading in advance of, or far after, index reconstitutions; and funds’ holding non-S&P 500 companies in their portfolios. We also find that, at least given the discretion that these S&P 500 funds currently exercise, investors do not respond to these departures by withdrawing capital from these deviating funds.

The implications from these findings are significant. Much of our evidence suggests that these titans of asset management have a more pronounced role to play in corporate governance than some currently think. We also discuss how our findings complicate current debates on universal ownership, which has been seen as a promising way to internalize negative corporate externalities through private markets. Moreover, if index funds track underlying indices voluntarily, and deviate from those indices at will, then there is a potential for investor confusion. Finally, our evidence raises new theoretical concerns. We highlight the difference between tracking error—differences between a fund’s returns and that of the index it tracks—and differential holdings between the fund and the index. As our results demonstrate, these two measures can and do differ substantially.


1https://www.bloomberg.com/opinion/articles/2021-06-22/everything-still-might-be-securities-fraud (go back)

2https://www.jstor.org/stable/45389496.(go back)

3https://www.actionsxchangerepository.fidelity.com/ShowDocument/ComplianceEnvelope.htm?_fax=-18%2342%23-61%23-110%23114%2378%23117%2320%23-1%2396%2339%23-62%23-21%2386%23-100%2337%2316%2335%23-68%2391%23-66%2354%23103%23-16%2369%23-30%2358%23-20%2376%23-84%23-11%23-87%230%23-50%23-20%23-92%23-98%23-116%23-28%2358%23-38%23-43%23-39%23-42%23-96%23-88%2388%23-45%23-32%23-112%23-4%23-65%23-3%2375%23102%23-104%23-74%235%23-89%23-105%23-67%23126%2377%23-126%23100%2345%23-44%23-73%23-15%238%23-21%23-37%23-17%23-14%23-98%23123%23-18%2345%23-59%23-82%2367%2383%23112%2317%2370%23-78%2378%23-50%2336%23-86%23-90%2381%23-21%23-119%23-30%23120%2349%2328%23-98%2333%2351%23-78%23-119%23-16%2350%23-58%2350%23102%2348%23-17%2352%23-99%23(go back)

Both comments and trackbacks are currently closed.