The U.S. competition agencies issued a Notice of Proposed Rulemaking (NPRM) that advises of proposed updates to the Hart-Scott-Rodino (HSR) notification rules. Bloomberg celebrated this news as an indication that the FTC is about to scrutinize common-ownership positions of index funds, which have aroused antitrust concerns in recent years.[1] Insiders from the agency also suggested I should view the NPRM as a success of the common ownership research agenda to date. After reviewing the proposed rules, I am a bit more cautious in my optimism than previous commentators about the rule change, and indeed worried about particular aspects of it. In this post, I explain why.
The backdrop to the proposed rule change are multiple lawsuits against institutional investors that held industry competitors and engaged with them on matters that would ultimately require shareholder approval, for example regarding replacing directors and a potential merger with the commonly owned competitor. Such involvement falls outside the scope of the “investment-only” exemption in the HSR rules, and led the FTC to clarify that ”investment-only means just that”. The suit also led observers to wonder why the rules appear to be selectively enforced against hedge funds, but not against much larger mutual funds that “routinely” engage in similar ways as the rule-breakers with their portfolio firms — but on a much larger scale. Amongst the enforced cases, however, the government also sued investors that violated the existing notification rule, as they held sufficiently high fractions of stock of the issuer to trigger the filing requirement, but were not even common owners. Does it make sense that the existing rules require such investors to file?
Industry and practitioners have long advocated for an amendment to ease the filing burden of such investors and make the rule less over-inclusive. Indeed, it at first seemed intuitive to me as well that regulators should remove apparently overbroad filing requirements for non-common owners. But on second thought, it becomes clear why my initial intuition was mistaken.
Before I point out the problems I perceive, I wish to note that the proposed rule change contains several good ideas. A perhaps incomplete enumeration is:
- The proposed rule covers significant common ownership links between significantly vertically related firms.[2]
- The proposed rules recognize that common ownership links in private markets are potentially a lot more frequent, and economically even more important than in public markets; the proposed rule explicitly covers private-equity owned competitors.[3]
- The proposed rule recognizes that governance interventions tend to happen at the asset manager level rather than at the fund level. The FTC is aware that “Treating these non-corporate entities as separate entities under HSR is often at odds with the realities of how fund families and MLPs are managed.“ Indeed, not only Private Equity firms, but also large mutual fund companies, including sponsors of index funds and ETFs, should have to file at the appropriate level of aggregation.[4]
Other proposed elements appear sensible on first sight, but on second thought give reason for concern. Whereas it is intuitive to relieve non-common owners of filing requirements if the goal is to get a comprehensive view of common ownership, that intuition is misguided. The reason is that the effect of common owners’ holdings depends on the holdings of everyone else as well, including the holdings of non-common owners. To illustrate, a common ownership position between Facebook and Google by an institutional investor is unlikely to significantly affect the two firms’ conduct, because large individual investors, including Mark Zuckerberg, hold the majority of control. Indeed, common ownership indexes used in research cannot be computed without having a set of ownership and control records that also includes all significant non-common owners. Yet, according to the proposed rules, investors holding less than 10% of the issuer are exempt from making HSR filings — unless they also hold at least 1% of a competitor. Having gaps in the ownership structure, e.g. due to granting filing exceptions to holdings that don’t have a “competitively significant relationship”, makes the resulting dataset unusable to study competitive effects of common ownership. Therefore, the FTC should want to obtain a full picture of the significant owners of all firms, and not just of common owners’ holdings.
What concerns me most about the proposed rule, however, is the lack of an amendment of the rules regarding what qualifies as an exemption to the filing requirement on grounds of holding the stock “solely-for-investment”[5]. The proposed rule continues to grant an exemption to investors who, at the time of the asset purchase, have no “intention” to influence the business decisions of the issuer.[6] This rule is hugely problematic for reasons the literature on common ownership has reiterated since the 1980s: an intention to influence basic business decisions is not a necessary condition for common ownership to have anticompetitive effects — the governance mechanism translating common ownership to anticompetitive conduct may simply be that management is “looking out for their shareholders”.[7] Conversely, an intention to actively push firms to maximize their own value may be necessary for firms to aim to do so — which is the basic assumption underlying textbook models of competition. Therefore, broadly exempting those investors who don’t have such an intention does not help to address the economic harm the rule is meant to help target. Instead, the overly broad exemption is likely to lead the agency to overlook the potentially most harmful investor holdings. The agency should consider tightening the solely-for-investment exemption, so that only investors who are entirely passive in matters of governance – including not voting their shares – don’t have to file.
The idea to require all significant and not entirely passive investors to file all holdings meets with some valid and some less valid critiques.
A less valid rationale is the notion that “index funds’ investment strategy means they are broadly diversified and therefore should be exempt.” The following example illustrates why thinking so would be wrong on factual and logical grounds. JETS is a “passive” ETF that tracks an index composed of competing airline companies. That is to say, JETS “passively” specializes in establishing common ownership links (whereas “passive” practically means the portfolio choice is implemented by a computer code). Surely one would not want to exclude funds that specialize in establishing common ownership links from the HSR filings, merely because human-programmed computer establishes the common-ownership positions rather than a human. Note JETS is not particularly diversified, despite being labeled a “passive” index fund.
If we don’t want to exclude JETS from the antitrust rules, then surely we wouldn’t want to exclude much larger index funds either, such as the very large S&P500 tracker SPY: SPY’s contribution to common ownership of some airlines is much greater still than JETS’s contribution. “Because these entities base their investments on an index” is therefore not a valid reason to be excluded from antitrust scrutiny.
Another misguided rationale would be to exclude funds based on the level of cheap diversification they offer to investors. A reduction in the cost of diversification is the fundamental reason for anticompetitive effects of common ownership in the theory of Rotemberg (1984), who, incidentally, concludes: “it may well be that the funds which concentrate on specific industries and those whose portfolio is very broad do the most harm.“[8]
A valid reason for complaints regarding the filing requirement on behalf of investors would be the apparent complexity of the filing process. It appears that this process could be simplified significantly and made nearly effortless for sophisticated investors that command modern IT systems. But not only is filing HSR notifications burdensome – it also comes with pecuniary costs. A single filing costs up to US$280,000! If the FTC aims for greater transparency, it should consider reducing the complexity and cost of filings to investors, so as to increase the supply of information. If the filing costs were reduced to zero, any remaining opposition by the industry would have to be interpreted as opposition to transparency.
Reducing the cost of filings has additional benefits, next to increasing transparency and the ability to research and monitor competitive effects of common ownership with a more complete ownership data set. If being active in corporate governance is taxed by the government (e.g. via filing fees for required filings), economic theory would predict that fewer investors will choose to be active. Discouraging active ownership probably reduces both the quality of governance and competition. The reason is that when fewer shareholders push firms to maximize their own value, to less productive firms, and thus higher consumer prices (though not necessarily higher margins). As such, reducing the government’s tax on active shareholders is important to maintain a competitive economy – and can be achieved without sacrificing transparency.
Reducing the cost and complexity of the filling system has additional benefits both for the FTC and investors: investors can be asked to report all their holdings, without either investors or the FTC needing to scrutinize whether the portfolio spans horizontal competitors or vertically related firms. Uncertainty whether all necessary filings have been made and the associated very significant legal liability would then be taken off the table for investors, and the FTC could focus its resources on other areas (such as the analysis of the data) if the rule required investors to report all holdings (other than those subject to a narrow investment-only exemption), irrespective of whether they pertain to related firms or not.
Taken together, the FTC should have individual or institutional investors that hold significant stakes in large private or public firms (e.g. 0.1% in publicly traded firms) report ownership and control rights of all their holdings, electronically, at zero cost, aggregated to a meaningful level, with the only exception of holdings that are entirely passive.
To sum up, the proposed rules contain several good ideas. In particular including private equity is likely important. At the same time, the competition authorities should keep in mind that measuring the entire ownership structure of the firms is necessary to be able to study common ownership and its effects. The agencies should aim to make the filings cheap for investors, and to not offer opt-outs for index funds or other investors based on their intentions at the time of acquisition of the asset. They should amend the investment-only exemption to make it as narrow as the statute allows. Greater inclusiveness of the filing requirement not only makes the collected data more useful, but also reduces the cost and complexity of the HSR system — both for investors and the FTC.
[1] A paper on airline competition under common ownership first circulated in April 2014 and published in the Journal of Finance in 2018 inspired hundreds of follow-on papers; I review that literature here and here.
[2] §802.15(b)(5)
[3] Research has shown competitive effects of common ownership of VC-owned firms, see e.g. Li-Liu-Taylor (2020).
[4] The proposed rule achieves this by expanding the definition of a person and an associate in §801.1(a)(1) and §801.1(d)(2).
[5] 15 USC §18a(c)(9)
[6] Large common owners may make such claims in the context of the antitrust liabilities their holdings create, irrespective of their statements elsewhere, where they emphasize that they do have the intention and power to influence firms to the effect of mitigating climate change, or the intention to make drug companies “put qualms about competition aside” and stop enforcing patents in the COVID-19 pandemic.
[7] Rotemberg (1984). For the same reason, the HSR filings should also not be streamlined with the SEC’s 13D filings.
[8] Notwithstanding, preventing common ownership at the asset manager level does no need to imply a reduction in diversification at the ultimate owner / household level because households can diversify across asset managers; see also Elhauge (2021).
Leave a Reply