Common ownership—the simultaneous ownership of equity of two competing companies by the same investor—has been rising steadily. This trend has been largely triggered by increasing institutional investor participation, including index funds owned by asset managers such as Vanguard and BlackRock. As highlighted by Oxera Partner, Professor Julian Franks, and Oxera Associate, Professor Vikrant Vig, the rise has been accompanied by an important debate on the potential anticompetitive effects of common ownership
The debate has been fuelled by a growing concern among policymakers, such as the European Commission and the US Federal Trade Commission, that common ownership may constrain competition and reduce social welfare. In a recent decision, the Commission stated:
it remains that large shareholders have a privileged access to the companies’ management and can therefore share their views and have the opportunity to shape the companies’ management’s incentives accordingly.
Common ownership among controlling shareholders has always captured regulatory attention, because it is easy to identify the channel by which the controlling shareholder might influence the boards of two competing companies. It is less easy to identify the channel by which shareholders holding non-controlling stakes as small as 5% can exercise a similar influence. However, it is the latter form of common ownership, with a group of asset managers each owning (say) 3–5%, that is currently drawing the interest of some regulators.
The objective of this article is to critically revisit the current evidence on common ownership, and to spell out some challenges for policymakers going forward. We specifically address common ownership by asset managers, rather than common ownership by large shareholders or cross-ownership as in the Ryanair–Aer Lingus case. Nevertheless, there will be important differences among asset managers, such as between those that are widely diversified and those with concentrated holdings in a few companies.
Common ownership and finance theory
How should we think about the economics of common ownership? Finance theory encourages investor diversification and therefore common ownership. Because the returns on stocks do not move in lockstep, investors are encouraged to hold stakes in large numbers of stocks to maximise diversification and risk-reduction. Thus, all investors irrespective of their risk appetite should hold highly diversified portfolios. One consequence is that shareholders will own stocks in competing listed companies. There is some irony in the fact that finance academics have regarded investor diversification and thereby common ownership as one of the mantras of finance. This mantra has fuelled the growth of index investing. By holding a slice of the index funds, both individuals and institutions could access well-diversified portfolios at significantly lower costs. While participation in index funds allows for investor diversification, it also creates common owners in competing companies.
But common owners have existed for a long time. What has changed? For one, investing through asset managers into diversified portfolios, including index funds, has become widespread. In addition, there are concerns about the increase in concentration of the asset management industry, with the total share of the assets of mutual funds and exchange-traded funds managed by the top five largest firms rising from 36% in 2005 to 50% in 2017. This has suggested to some commentators that there are potential costs of such levels of concentration.
Common ownership and antitrust enforcement
One such cost of concentration is the potential for limiting competition. There are three channels by which shareholders or their agents could influence management actions:
- encouraging cartels, with companies directly colluding to fix prices and outputs, etc.;
- creating a ‘hub-and-spoke’ cartel, where asset managers pass on information on one company to another, as a conduit for collusion;
- softening competition, where companies and individuals, while avoiding any collusive behaviour, make decisions that have the effect—even if unintentional—of constraining competition.
Competition authorities have the tools to deal with the first two channels. They both relate to collusion, with instances of individuals conspiring to worsen competitive conditions in the market to benefit all industry participants.
However, the third is far more subtle—since it does not require an agreement (explicit or implicit) between companies, it acknowledges the fact that there is less incentive to compete intensively among companies that share common ownership. It could involve shareholders guiding management to engage in less competitive behaviour by constructing compensation plans that explicitly benchmark industry rather than company profitability. It could also involve shareholders (or their agent, the asset manager) discouraging, or at least not encouraging, particular actions that might lead to greater competition, such as a significant capacity expansion or a new R&D project that has the potential to affect competition in the industry. An example of such behaviour might involve arguing against an expansion of an aircraft fleet, justified on the grounds that the expansion is too risky. The same result could also be achieved by increasing the dividend payout, or adopting higher leverage through share buybacks, so as to restrict the amount of cash available to invest and as a result induce less competition.
Such constraints on management discretion would present competition authorities with a much greater challenge, since there would not need to be explicit collusion or information sharing. Common shareholders might simply vote in such a way as to maximise their joint surplus; since all common shareholders have the same incentive structure or commonality of purpose, they are likely to vote the same way.
Several important (and controversial) academic papers have pointed out the potential influence of institutional investors with significant—albeit small minority—blocks. Appel, Gormley and Keim (2016) find that index funds, which rarely individually have more than 5% of a company’s equity, influence the governance and performance of firms, and their evidence ‘suggests that a key mechanism by which passive investors exert their influence is through the power of their large voting blocks (i.e. voice)’.
More directly and of greater concern, Azar, Schmalz and Tecu (2018) find that such common ownership may have constrained competition in the airline industry. They calculate market power adjusted for common ownership and assert that their estimates are ten times higher than what would be approved by the antitrust authorities. They also conclude that airfares are higher as a result. The authors suggest that a hidden social cost—reduced product market competition—accompanies the private benefits of highly diversified portfolios and active ownership.
Preventing anticompetitive behaviour arising from common ownership
Are there mechanisms that can mitigate or prevent the negative effects of common ownership? To start with, management is incentivised to maximise the value of the firm it manages, not firms that it does not manage. Also, compensation plans for management are usually linked to the company’s performance rather than that of the industry. Moreover, non-executive directors may refuse to countenance measures that constitute anticompetitive behaviour. Finally, shareholders do not often show the commonality of purpose implied by the theory of common ownership. They often disagree with one another, which is reflected in the way they vote on management proposals at annual meetings. Similarly, asset managers do not always have voting mandates from all their clients, and those clients will direct them on how to vote.
There are other factors that may act as a constraint on attempts to restrain competition. For some asset managers with widely diversified portfolios, constraining competition in one segment of an industry may have spillover effects in other segments. Since the objective of the fund would be to maximise the value of the entire portfolio and not just a subset of it, restricting competition may not be in asset managers’ best interests. For example, constraining competition between airlines would adversely affect suppliers of aircraft. If a fund owns shares in an aircraft supplier, this would decrease its incentives to soften competition between airlines.
Finally, the democratisation of equity ownership has meant that many more consumers have become shareholders, and while reduced competition may be in their interests as shareholders, it is not always in their interests as consumers. In theory, shareholders of two airlines should not vote to restrict competition and push up airfares when they themselves will have to pay the higher fares when they fly. Of course, shareholders may gain more from reduced competition than they lose from higher airfares. The pattern of ownership is not the same as the pattern of consumption. The rich own more of the stock market than the poor. Moreover, some may argue that the voting decisions of asset managers do not generally take account of consumer welfare; their job is to improve shareholder returns. However, there is evidence that asset managers take account of the interests of stakeholders other than shareholders. For example, asset managers often involve themselves in decisions relating to corporate social responsibilities (CSR) of the firms they invest in.
In summary, while the potential for common owners to act together to disadvantage consumers is theoretically possible, the reality may be very different. However, if common ownership resulted in costs to consumers then oversight by the regulatory authorities would be required. The intervention should be justified only through empirical evidence.
Proving these costs would be difficult, and the evidence so far is limited. Critics rightly demand to know the channel through which competition is being constrained. However, as we have pointed out, the channel may not be as transparent as some academics have suggested: common owners may vote to maximise joint surplus without colluding, particularly through soft measures that constrain competition. Motives for opposing expansion plans may be difficult to prove, particularly when they are shrouded in arguments over risk. A softening of competition may also occur unintentionally as a result of weaker incentives to compete over market share.
In this case, there may be no obvious culprits to prosecute, although the outcome may be against the public interest. That makes the problem unusually difficult to solve. Solutions that restrict investor diversification, such as stricter merger control, would impose significant costs on investors, since these investors would lose some of the benefits of diversification. Broadening the fiduciary duty of directors to disregard the joint surplus of common shareholders may have unintended consequences. Sometimes, maximising joint surplus might be in everyone’s interests—for example, in a cost-reducing merger. Moreover, legislating against the maximisation of joint surplus would be difficult to enforce, particularly when the intentions of shareholders when they vote are difficult to discern. Restricting shareholder votes when competition issues are at stake would also prove difficult to design and would invite avoidance. None of these solutions are easy to implement, and in some cases they carry high costs.
This leads to the final question: what industries are at risk from common ownership? This is largely an empirical question, and so far the empirical evidence is limited. The usual suspects seem to be those industries that are more prone to collusive activities and are already on the radar of competition authorities. In these cases, common ownership simply makes collusive action easier, and the risk is that what used to be explicit collusion becomes subtle enough that it is no longer detected by competition authorities. The other interesting question is: what can we say about other industries that are not on the radar of the competition authorities, where there is little or no apparent collusion? Could increasing common ownership change the competition landscape?
In situations where competition is driven by large, disruptive investments (in capacity or R&D, for example), common ownership may slow down the pace of displacement, technological progress and competition. However, in industries where the barriers to entry are low, the industry is fragmented, and where competition is dynamic, it is far more difficult to conceive of common ownership presenting competition problems.
At this point, it appears desirable to remain cautious and to await further empirical evidence, particularly on whether there is a causal link between industries with high common ownership and a lower level of competition. Regulators should take this into account in their monitoring of anticompetitive behaviour. While there is the potential for common ownership to influence and constrain competition in product markets, at present the evidence is limited and rather mixed. Any knee-jerk reaction on the policy front may have unintended consequences and could do more harm than good. However, policymakers need to be watchful—even though it may not have had negative effects so far, the potential for future harm is there.
Professor Vikrant Vig