Common shareholding exists when the leading shareholders of different corporations overlap. More than two dozen empirical studies have now confirmed that common shareholding alters corporate behaviour. Ten of those empirical studies have confirmed that horizontal ownership often has anticompetitive effects in concentrated markets. These include five market-level studies, a massive cross-market study of hundreds of consumer goods, two national studies across all industries, a new study of horizontal ownership by venture capitalists, and a new study showing that firm entry into the S&P500 creates an exogenous increase in horizontal shareholding that raises rival stock prices.

Despite this, critics have argued that the law should not take any action until we have clearer proof on the causal mechanisms through which common shareholding operates. Some have developed a typology of causal mechanisms, only to argue that each type of mechanism either has not been empirically tested or is implausible. Others go even further and argue that the empirical studies showing that common shareholding affects corporate behaviour should be ignored, because it is implausible that institutional investors would have incentives to try to influence corporate conduct through any mechanism.

This paper, available here, argues that we actually have ample proof on causal mechanisms. However, it also argues that this empirical evidence is of limited use, since antitrust enforcement should focus on anticompetitive market structures, rather than on such mechanisms.

Part I shows that there is ample proof of common ownership causing anticompetitive outcomes.

The causal mechanisms through which common shareholders might influence corporate policy include all the ordinary mechanisms by which managers are incentivised to act in the interests of their shareholders: shareholder voting, executive compensation, the market for corporate control, the stock market, and the labour market. For decades, corporate law and economics’ scholarship has argued that this combination of mechanisms ensures that managers are primarily influenced by the interests of their shareholders. When the interests of a firm’s shareholders are changed by common shareholding, these same mechanisms imply that managers will be primarily influenced by those altered shareholder interests.

Horizontal shareholding are relevant because shareholder interests mean that those shareholders will, to some extent, be harmed by competition with rivals. This will lessen firm incentives to compete. The mechanisms through which this can occur are manifold.

  • Shareholding voting for the board creates incentives for board members to promote shareholder profitability over company profitability. This might involve increasing overall industry profitability over that of the company.
  • Executive compensation has been shown to reflect industry performance more than corporate performance in the presence of horizontal shareholding. This is in line with the fact that, the more horizontal shareholders a firm has, the more its shareholders care about industry performance rather than just the firm’s own profits.
  • How horizontal shareholders vote in corporate control contests can directly affect whether the corporation pursues a less competitive strategy. Since managers can anticipate that future control contests may occur, they have incentives to act in ways that please the horizontal shareholders that may be decisive in any future control contest at all times – including by restricting competition when those shareholders have interests in a substantial part of the industry.
  • The stock market is another plausible causal mechanism. Managers might reasonably fear that, if they displease their horizontal shareholders by competing too aggressively, those shareholders might sell their investments, which would depress the stock price and the value of executive stock options that are a major component of their compensation.
  • Yet another plausible mechanism is the labour market. Directors who want additional directorships at other corporations, and executives who want a promotion to their next job at another corporation, will be affected by how favourably disposed the leading shareholders at those other corporations will be towards them. Given the prevalence of horizontal shareholding, the leading shareholders at those other corporations are likely to be the same investors who are horizontal shareholders at their current firm. Directors and executives who want higher odds of gaining directorships or promotions thus have incentives to please those horizontal shareholders with increased returns that result from diminished competition.

Part I also argues that it is incorrect to argue that enforcement should focus on causal mechanisms, rather than on anticompetitive market structures.

The claim that antitrust enforcement requires stronger proof on causal mechanisms is misbegotten. When there are plausible causal mechanisms, it is hard to see why one should ignore multiple statistical correlations between the conduct and serious societal harm (particularly when there are proper controls for other possible reasons for these correlations). As a policy matter, ignoring statistical correlations that have such low odds of being random results entails enduring a risk of social harm that greatly exceeds the risk of harm from regulating the problematic conduct.

Nor is it correct to argue that enforcement should focus on regulating particular causal mechanisms. Given that these causal mechanisms are the same ones used to desirably influence corporations generally to advance their shareholders’ interests, efforts to ban solely the anticompetitive use of any of these mechanisms would be ineffective. This is not only because evidence on that topic will generally be non-public or obscure, but also because of substitution effects across mechanisms.

Instead, enforcement should focus on changing anticompetitive market structures – after all, the Clayton Act bans mergers and stock acquisitions that are likely to have anticompetitive effects regardless of whether the mechanism for those effects is known. Because horizontal shareholding in concentrated markets is a structural problem, the only effective remedy is preventing or undoing that anticompetitive structure.

Part II refutes claims that every type of causal mechanism is either empirically untested or implausible (at least for index funds).

Professors Hemphill and Kahan offer the most thoughtful critique of the literature on causal mechanisms by which horizontal shareholding might cause anticompetitive effects. They focus on a typology based mainly on the various effects that the causal mechanisms might have, and on various claims that such effects are either unproven or implausible, and reach two main conclusions. First, they claim that causal mechanisms that rely on consensus between shareholders, or on common shareholding having generic effects on corporation’s managers incentives to compete, have not been empirically tested. Second, they claim that mechanisms that rely on shareholders trying to actively influencing the board have been tested, but are implausible given the levels of risk and knowledge they entail. Finally, they argue that mechanisms that rely on shareholders failing to press the board to compete are implausible for the index fund families that are major horizontal shareholders.

Concerning the first type of causal mechanisms, the problem with these conclusions is that such causal mechanisms have been empirically demonstrated, and the authors misunderstand a number of those studies. As regards the second type of causal mechanisms, the risk of conflict between shareholders is much lower than the authors assume. Further, there is evidence that horizontal shareholders are able to obtain the necessary information to try to influence corporate decisions, and many empirical studies do prove targeted effects in particular markets.

Part III shows that horizontal shareholders (including index funds and diversified active funds) have strong incentives to influence corporate conduct in anticompetitive ways.

The last argument by Hemphill and Kahan is that those mechanisms that rely on shareholders failing to press the board to compete are implausible for the index fund families that are major horizontal shareholders. This claim can take two forms. First, it is claimed that any anticompetitive incentives from horizontal shareholdings are negated by those shareholders’ investments in vertically-related corporations. This argument ignores not only the reality that horizontal shareholders (even index funds) generally are not equally invested in vertically-related firms, but also the point that, even when they are, such investments would create two layers of horizontal shareholdings that would compound, rather than negate, anticompetitive effects.

The second form this claim takes is to argue that index funds lack incentives to exert any affirmative effort to increase portfolio value by lessening competition or otherwise. They claim that an increase in portfolio value: (a) cannot make an index fund perform better than other similar index funds, and thus will not induce additional investment flow; and (b) will reap additional index fund fees that they claim are too small to induce any significant effort on increasing portfolio value. However, economic theory indicates that index funds have strong incentives to exert efforts to increase portfolio value by lessening competition because their anticompetitive gains are vast, while the incremental effort costs are generally zero or negative. First, investment flow to an index fund does not come solely from other investment funds. Increasing their profitability is likely to attract more funds to index funds in general. Further, investment funds have legal requirements to incur the costs of voting in an informed manner. Those costs are thus mandatory, and it costs the same to vote to either promote or lessen competition. Indeed, costs are probably negative when it comes to shareholder influence on competitive behaviour: because competing vigorously is hard work for managers, they are less likely to do it unless their shareholders are actively pressing them to compete. Horizontal shareholdings can thus induce less competitive corporate behaviour by incentivising horizontal investors to expend less effort on encouraging greater competition or cost reductions than they would have exerted if they invested in only one of the competing corporations.

In any event, horizontal shareholdings are generally not held by index funds and, even when they are, their shares are voted by fund families that also have active funds. Finally, the argument that index funds lack incentives to exert effort to increase corporate valuations conflicts with copious empirical evidence, which indicates not only that index funds engage in extensive efforts to influence the corporations they hold, but that their efforts are highly effective.

Part IV concludes.

Underlying critiques of calls for antitrust intervention against certain types of common ownership is the fear that antitrust enforcement against horizontal shareholding would either greatly restrict diversification, or discourage desirable institutional investor influence on corporate conduct. This argument is more than a little ironic, given that its premise is that institutional investors can influence corporate conduct, which is inconsistent with the critics’ claim that such influence is unproven or implausible.

In any event, this argument rests on a false premise that tackling the anticompetitive effects of horizontal shareholding requires restricting either diversification or institutional investor influence. To the contrary, the natural remedy would just shift diversification to a different level and increase investment fund influence by having such funds concentrate their shareholdings in one firm per product market.

Comment:

This piece provides an accessible introduction to the current status of the debate on whether common ownership poses antitrust issues. It is naturally one-sided, but it presents – and refutes – arguments that common ownership is not a problem in a fair-handed way that is quite welcoming in this sometimes heated debate. Despite this already being a fairly long paper, there are two additional elements that I think the author could have dealt with, at least for the above purposes (which are, naturally, not those of the author). First, the paper could have included a structured literature review, including a summary of the various empirical studies that purportedly support intervention against common ownership. Second, the paper could contain a discussion of how antitrust should deal with common ownership. On this latter point, the author suggests that a structural solution is appropriate, but it is not obvious to me how a workable solution would look like. Thankfully, this is the focus of a paper that I will review in the coming weeks.

Author Socials A weekly email with competition/antitrust updates. All opinions are mine

3 Comments on Einer Elhauge ‘The Causal Mechanisms of Horizontal Shareholding’ (2021, forthcoming) Ohio State Law Journal 82

  1. Einer Elhauge says:

    Thanks for your analysis. Your concluding comment is “Despite this already being a fairly long paper, there are two additional elements that I think the author could have dealt with…. First, the paper could have included a structured literature review, including a summary of the various empirical studies that purportedly support intervention against common ownership. Second, the paper could contain a discussion of how antitrust should deal with common ownership.” I would just point out that both those additional elements were addressed at length in the companion article, Elhauge, How Horizontal Shareholding Harms Our Economy—And Why Antitrust Law Can Fix It, 10 HARVARD BUSINESS LAW REVIEW 207 (2020). Because the articles would otherwise be too long, I had to split the analysis into two articles, so I just cite to the latter article for those two elements in this one.

    1. antitrustdigest says:

      Thank you for getting in touch. That is a fair comment, and you will note that I mentioned that those two matters are “the focus of a paper that I will review in the coming weeks” – it is the very paper you refer to. Hopefully I will be able to engage with your proposals with a bit more depth than I did with your analysis here.

      1. Einer Elhauge says:

        Thank you. I will also note that an updated version of the causal mechanisms article, which I think is later than the version you reviewed, does also add a summary and some analysis of new empirical studies that came out after my 2020 article was written. So you might want to check that out too. Cheers.

Reply to Einer Elhauge Cancel Reply

Note: Your email address will not be published