This article, available here, examines common ownership through a European lens. The article considers whether the theory of harm flowing from common ownership is sufficiently robust to provide a basis for enforcement, and (if so) whether current European Union competition law tools could be used to that end.
The authors argue that it is premature to draw any conclusions as to whether common ownership concerns justify competition enforcement. In any event, levels of common ownership seem to be lower in Europe than in the US, so it is unclear whether intervention would be justified in the EU even if it were in the US. Until a better understanding of the underlying facts and a broad academic consensus emerge, reform prescriptions that have been advanced will remain a solution in search of a problem.
Section II describes the common ownership theory of harm.
The authors begin by distinguishing between cross-shareholding and common ownership. Cross-ownership arises where one firm acquires a non-controlling minority interest in a competitor. This is a recognised source of potential anticompetitive effect under two theories of harm. First, the acquirer may compete less aggressively with the target, since any gains from the acquirer’s own activities may be offset by a negative impact on the acquirer’s share of the target’s profits. Second, the acquisition may harm competition if the acquirer has some degree of influence or control over the target, thus enabling the acquirer to inhibit competition by the target. Common ownership addresses a different situation, one where a number of institutional investors each have holdings in a set of listed companies that compete with each other, notably where those listed companies are players in a concentrated industry.
The problem with common ownership is that competition between firms with shareholders in common may be weakened due to reduced incentives to compete, even when those shareholders have only non-controlling minority interests (the ‘common ownership theory of harm’). In particular, so the theory runs, institutional investors that manage investment funds with shares in multiple competing firms would prefer to maximise industry as opposed to firm profits, and may therefore—implicitly or explicitly—encourage firms to compete less aggressively. The theory contrasts such investors with a shareholder that has no interests in competing firms and would thus prefer management to maximise firm profits, even if at the expense of industry profits.
A question this raises is how institutional investors with modest interests in multiple portfolio companies might somehow be capable of influencing management to compete less aggressively. Broadly, proponents of the theory of harm point to three factors that confer greater power on institutional investors relative to individual shareholders. First, diversified institutional investors with shares in competing companies are among the largest shareholders, and taken together their holdings may be substantial. Accordingly, the managers of portfolio companies may have an incentive to pivot firm strategy towards the interests of those investors. Second, active influence by diversified institutional investors may not be even required in order for anti-competitive effects to arise: these investors may simply choose to remain detached and not push portfolio companies to compete aggressively, thus allowing managers to enjoy the ‘quiet life’. Third, compared to individual shareholders, institutional investors are better placed to engage actively and to influence management due to their superior organisation, expertise, and extensive monitoring resources.
Section III considers the degree of common ownership in the EU.
To date, the empirical research on the competitive effects of common ownership has essentially focused on the U.S. At the time of writing this article, no comprehensive review had been undertaken of the level of common ownership in the EU. To the extent there have been any studies on this topic in Europe, they have been limited in scope. The CMA looked at three sectors, and found limited evidence of common ownership. The German Monopolkommission found that common ownership could be identified in some industrial sectors, but not others.
The apparently limited extent of common ownership in the EU likely reflects the fact that the equity investor landscape in Europe is fundamentally different from that in the U.S. It follows from the above that the question of whether common ownership is widespread in the EU remains unanswered. If anything, the available evidence suggests the opposite.
Section IV provides an overview of critiques of the common ownership theory of harm.
Several academics have vigorously contested the findings of the papers that identified common ownership as a problem, on both theoretical and methodological grounds. Therefore, a consensus is yet to emerge.
In particular, the empirical foundations of the common ownership theory of harm have been repeatedly called into question. Given the uncertainty about the degree as well as the nature of influence exerted by index funds over corporate governance matters, it is unsurprising that the common ownership theory of harm has been met with great scepticism. The authors explain how concerns about common ownership have been said to rest on flawed assumptions about the asset management industry: notably proponents of the theory of harm overlook the distinction between asset ownership and asset management. Shares in portfolio companies are managed by asset managers on behalf of thousands of clients who are the owners of those shares, and who in many instances retain control over the voting of those shares. This leads to the erroneous treatment of asset managers and clients as unitary entities. Additionally, the theory of harm rests significantly on the premise that holdings of different investors may legitimately be counted together because the clients share an adviser—a point that has been as vigorously contested. Further, evidence in relation to ‘voice’ is weak and anecdotal.
Drawing conclusions on these issues is beyond the ambition of this article; instead, the more limited goal here is to show that the empirical evidence in support of the theory of harm remains hotly disputed, such that it is altogether premature to draw any conclusions from the ongoing debate, still less to propose solutions.
Put simply, more empirical work is needed to credibly establish a link between common ownership and diminished competition.
Section IV also evaluates some proposed interventions to address the harms of common ownership.
Yet, despite the debate being nascent and ongoing, a number of measures have already been proposed to address the alleged competitive harms of common ownership. The US antitrust agencies have been called to apply a standard rule whereby transactions undertaken by intra-industry diversified institutional investors would be automatically investigated under the Clayton Act if (i) they result in a modified HHI delta of over 200, and (ii) take place in a market with an modified HHI in excess of 2500. Others would impose a hard limit on common ownership that would restrict investors in firms in defined oligopolistic markets to a stake of no more than 1 per cent of the total size of the industry or shareholdings in a single ‘effective firm’ per market. This approach would require antitrust agencies to compile an annual list of industries to be designated as oligopolies.
Leaving aside the question whether any regulatory intervention is warranted at all in the absence of proof of the anti-competitive effects of common ownership, both proposals are inherently problematic from an enforcement and compliance standpoint. Extending the application of the Clayton Act to common ownership may give rise to significant uncertainty. Antitrust agencies may have to dedicate significant resources to undertaking a comprehensive annual market definition exercise. Finally, some of these proposals go against duties imposed by corporate law and securities’ regulation.
In short, regulators need to proceed with caution given the state of the ongoing academic debate.
Section V examines whether the common ownership theory of harm can be accommodated in the current EU competition law framework.
There have been suggestions that EU law could be used to challenge common ownership. The authors think that such calls have very limited support.
Calls for the EUMR to be interpreted so that institutional investors’ common shareholdings can be reviewed as an acquisition of de facto joint control is predicated on the assessment of institutional investors’ share acquisitions as a single operation that aggregates their shareholdings in the relevant undertaking. However, a series of acquisitions of shares by institutional investors is incapable of being characterised as a single concentration for the purposes of the EUMR. Even assuming—quod non—that uncoordinated share acquisitions by multiple funds could be considered to constitute a single concentration, the shareholdings of institutional investors must be capable of conferring joint control in order for jurisdiction to be triggered under the EUMR. This is exceedingly difficult, requiring proof of (i) links between institutional investors, (ii) a commonality of interests that stems from mutual dependency, and (iii) a stable coalition that is capable of consistently garnering the majority of votes. None of these requirements is met by mere common ownership; and without all of them being met there is no valid basis in fact or in law to establish de facto joint control.
It also has been argued that Article 101 TFEU can be extended to common ownership, since it involves contractual agreements between institutional investors and competing corporations that give rise to anti-competitive effects. However, for an infringement to be found, those contractual arrangements must involve something akin to control, and the result must be akin to coordinated action (e.g. by requiring the companies to consider each other’s interests when determining their respective commercial policies) – the production of anticompetitive effects as a result of the acquisition of financial stakes would not suffice. As a result, arguments that horizontal shareholding amount to anticompetitive agreements ignore the requirement of the existence of a concurrence of wills between at least two parties to achieve an anticompetitive effect or object. Finally, the qualification of horizontal shareholding as a concerted practice is of doubtful validity, since it would require a form of coordination whereby the shareholders would knowingly substitute practical cooperation between them for the risks of competition by means of direct or indirect contact between such operators. Common ownership does not seem to meet such a threshold.
Finally, as regards Article 102 TFEU, it has been argued that it applies to (i) acquisitions of minority shareholdings by a dominant shareholding, and (ii) common ownership is caught by Article 102 TFEU where it gives rise to collective dominance that results in excessive pricing. As regards the former, it is doubtful that multiple passive investors can each be said to have a dominant position. Further, case law supporting this type of abuses refer to exclusionary strategies to prevent an effective competitor from emerging by purchasing the company in which the dominant firm had invested. Regarding collective dominance, the authors note that the idea that institutional shareholders could be held liable for infringements of their portfolio companies flies in the face of the principle of personal responsibility; and that the types of links created by common ownership are not sufficient to give rise to collective dominance.
Section VI discusses the Commission’s recent invocation of the theory of harm in its assessment of mergers.
The common ownership theory of harm featured in the Commission’s review of the Dow/DuPont and Bayer/Monsanto mergers. Although the theory of harm was not dispositive to the outcome of the review in either case, it was taken into account as an ‘element of context’ in the assessment of unilateral effects arising in relation to ‘innovation competition’. Annex 5 of the Dow/DuPont decision, which runs over 20 pages, is entirely dedicated to an ‘assessment of the effects of common shareholding on market shares and concentration levels’. The Commission found that (i) industry shares under-estimated the expected non-coordinated effects of the transaction due to significant levels of common shareholding between the principal players, and (ii) high levels of common shareholding ‘provide indications that innovation competition in crop protection should be less intense as compared with an industry with no common shareholding.’ Despite the ongoing academic debate, the Commission appeared to fully endorse the common ownership theory of harm, and the decision is replete with declarations that common ownership is likely to soften competition.
In Bayer/Monsanto, a similar assessment, consisting of an eight-page section, is not as extensive as the analysis in Dow/DuPont. However, the conclusion is the same. In particular, the Commission maintains that concentration measures, such as market shares or the HHI, are likely to underestimate the level of concentration of the market structure and, thus, the market power of the parties; and that common ownership is an element of context in its competitive assessment. However, in this merger the Commission did not go as far as to assert that the presence of significant common shareholding is likely to have an adverse impact on competition, and acknowledges that the debate is ‘relatively recent and not yet entirely settled’.
The Commission was careful to say that it was not relying on the common ownership theory of harm in both mergers. Indeed, it is highly unlikely that a decision predicated on the common ownership theory of harm would survive a judicial challenge. Following Tetra Laval, the economic evidence that is relied on by the Commission to assess a proposed merger needs to be factually accurate, reliable, consistent, and exhaustive. Some have argued that theories of harm that are not consensual are subject to stricter evidentiary requirements; and that, where the anti-competitive nature of a merger is not readily apparent, there is need for more convincing evidence as compared to more easily expected anticompetitive effects. Given that the economics underlying the common ownership theory of harm remain unsettled, as the Commission itself acknowledges, it is highly unlikely that a theory of harm based on common ownership effects would be upheld by the EU courts.
Despite the great respect I have for the authors’ intellectual integrity, I must start by emphasising that the authors are not only lawyers in the same law firm, but also disclose that the Investment Company Institute supported their research. They also disclose that they are lawyers who acted in a number of the cases discussed, including Aer Lingus/Ryanair and Bayer/Monsanto; and that their law firm acts for clients in the asset management industry.
After this disclaimer, I will say that if you want a review of the arguments against common ownership being sufficiently established as a source of anticompetitive effects to justify competition intervention, you should probably read this article. From the way the authors outline the issue, to the manner in which they review the main criticisms against common ownership as a theory of competitive harm, to the explanation of why EU competition law is not well placed to intervene, the exposition is clear and comprehensive. The point that there is limited evidence of the actual causal mechanisms through which common ownership causes harm in individual instances is well argued, and something which – if I understood the literature correctly – has been acknowledged even by proponents of the common ownership hypothesis. As may be apparent from one of my reviews last week, I am also broadly in agreement with the authors’ analysis of how common ownership would be treated under EU law (please note that I read this piece only after I penned that review).
On the other hand, I think that a focus on index funds and the very largest institutional investors is misplaced in a paper that criticises the common ownership theory of harm in general. As should be clear from the papers I reviewed over the last couple of weeks, the literature on common ownership has since moved well beyond a focus on such investors, and so should its critique. The distinction between cross-shareholding and common ownership also strikes me as being of limited usefulness, and for the same reason: the common ownership literature deals with cross-ownership as well as with common ownership as defined by the authors, i.e. it is not limited to the practices of institutional investors. This is not to say that the distinction is without merit, but inasmuch as their focus is solely on institutional investors, the authors are not addressing many of the concerns identified in the common ownership literature. Further, inasmuch as there is empirical evidence of anticompetitive effects in markets where common ownership is present, I have serious doubts that identifying specific causal mechanisms in individual cases is necessary before measures to address such anticompetitive effects are adopted.
Further, in some areas, the article has been surpassed for events. For example, the authors note that there was no survey of shareholding in the EU at the time they wrote the article. That is no longer the case: the European Commission published a study on the topic last year. It contains some stark findings. For example, firms in BlackRock’s portfolio in the EU Oil & Gas market represented roughly 90% of total assets for this market in 2016. Undermining the authors’ argument that common ownership is a US problem, the study found that while 60% of US public firms in 2014 had common shareholders that held at least 5% in the firm itself and also in a competitor, the number of common shareholding with at least 5% participation in Europe involved 67% of the listed companies in 2016. Firms included in the largest portfolios represent a significant proportion of the total value of the market, reaching a coverage of above 80% of total assets and more than 90% of market capitalisation. This means that the top investors not only hold shares in a considerable number of firms, but also typically choose to invest in the largest enterprises.
This report also analyses five sectors in more detail: Oil & Gas, Electricity, Mobile Telecoms, Trading Platforms and Beverages. In these industries, common shareholding patterns more or less mirror the general trends just described. Portfolios of common shareholders are very large in all five sectors, in some cases including 30% to 40% of active companies (i.e. in the Electricity and Oil & Gas sectors). In all studied sectors, the inclusion of firms in the portfolios of common shareholders continues to be based on size, with excluded firms only representing around 10% of industry total assets. Once weighted by the respective holdings, the top investors typically enjoy joint ownership of large portions of the considered industries. The comparison of portfolios of large investors also highlights a wide overlap of strategies, in many cases without evidence of differential investments. The largest funds show almost coincident portfolios, investing in the same companies, with correlations as high as 90% repeatedly over the years in all considered sectors.
The implication of this is not that the arguments of the paper are spurious. However, some of its assumptions – e.g. that common ownership is not as prevalent in Europe as in the US – are now facing additional contrary evidence. The question that remains is whether the legal analysis will also be overtaken by reforms.