Varieties and Mechanisms of Common Ownership: A Calibration Exercise for Competition Policy

Minority shareholdings have been on the regulatory agenda of competition authorities for some time. Recent empirical studies, however, draw attention to a new, thought provoking theory of harm: common ownership by institutional investors holding small, parallel equity positions in several competing firms within concentrated industries. While critical voices abound, EU and U.S. antitrust agencies closely follow these developments indicating an appetite to act. This article connects the common ownership debate to merger control and explores: i) the aims and scope of legal control as regards partial acquisitions in different jurisdictions; ii) the nature of potential competition effects arising from passive minority shareholding; and iii) the plausibility of common owners’ anticompetitive strategies from a corporate governance perspective. Drawing a distinction between “concentrated” and “diffuse” common ownership, it sheds light on the different supporting mechanisms and varying harm potential of each variety. “Passive influence” mechanisms characterizing “diffuse” common ownership may not only generate plausible and material competition concerns in given circumstances but present challenges for the effective jurisdictional and remedial design of merger law frameworks. Competition policy should stay current by explicitly recognizing these novel insights in enforcement practice and developing guidelines on how to treat common ownership cases in the future.


I. Introduction
Minority shareholdings have been on the regulatory agenda of competition authorities for some time. Recent empirical studies, however, draw attention to a new, thought provoking theory of harm: common ownership by institutional investors holding small, parallel equity positions in several competing firms within concentrated industries. Proponents of the "common ownership thesis" suggest that the indirect structural links between industrial competitors due to 7 legal reality does not allow straight conclusions as to whether common ownership is or may be a European problem too. A number of complicating factors affect the answer to this question.
First, commonly owned U.S.-based corporations often hold firms active in the EU, adding to the direct participation and potential influence of major common shareholders in EU-based companies. 35 Second, the legal analysis is complicated by the fact that although common ownership patterns across EU companies and contexts may diverge, transactions between commonly held U.S. companies, which require merger notification in Europe, "may invite examination for possible relevance of common ownership". 36 Third, considering the analytical insights offered in this article, it is possible that the broader phenomenon of common ownership may take different shape and rely on different mechanisms as evidenced in certain European countries ("concentrated" common ownership) when compared to the U.S. ("diffuse" common ownership). 37 What is also notable about the recent literature on common ownership and the corresponding novel theory of competition harm is that it is singularly linked to the rise of index funds and the Big Three asset managers as the typical common owners. This limited focus on common ownership in the narrow sensedue to the empirical prominence of this type of investors in the U.S. especiallyobscures the fact that it is subset of a broader phenomenon. 38 Indirect partial ownership of industrial competitors by common shareholders-investors that are not active in the same relevant market may take many forms. 39 In principle, common ownership is neither limited to certain institutional investors or to public firms with a dispersed shareholder base as is typically the case in most U.S. listed companies. Common owners of public firms may also be individuals or other types of institutional investors such as hedge funds with more concentrated holdings in several rivals in an industry. Common ownership Commission [1987] ECR 4487; Cases IV/33.440, Warner-Lambert/Gilette and IV/33.486, BIC/Gillette [1993] OJ L 116/21. However, this case law predates the adoption of the EUMR and has effectively become dead letter following its implementation in 1990. See further on this Tzanaki, 'The Regulation of Minority Shareholdings and Other Structural Links between Competing Undertakings' (n 20). 35 Rosati and others (n 26) 7. 36  may also exist in private firms when venture capital, private equity funds or alternative institutional investors seek to diversify their portfolio of investments in entrepreneurial firms that may be actual or potential competitors with each other. 40 To be sure, the type of investor and firm setting is material not only for the assessment of common ownership as an empirical phenomenon but also as a matter of theory. Different forms of common ownership may be associated with different theoretical mechanisms or channels of influence and distinct factual settings under which competition concerns may be likely and substantial. In this sense, scholarly discussion on common ownership has been undisciplined, narrowly focused, and often thwart by communication gaps and misconceptions.
Against this backdrop, the present article offers a unifying framework to organize and provide the right focus to the analysis of common ownership in the narrow and broad sense from an antitrust perspective. By putting forward a novel distinction between two paradigmatic types of common ownership, the "concentrated" and the diffuse", the article illustrates that each variety of common ownership is conceptually associated with different unilateral theories of harm and economic mechanisms potentially giving rise to anticompetitive effects. 41 Accordingly, it is shown that variety I of common ownership is conceptually linked to economic mechanisms of "active influence" (corporate influence) and legal conceptions of control in competition and corporate law (majority control). In contrast, variety II may obtain in the presence of "passive influence" mechanisms, based on economic conceptions of control or influence under industrial organization theory on one hand (strategic influence) and de facto minority control in the governance of the commonly owned firms on the other (actual control).
As explained, the contemporary debate on the competitive effects arising from parallel minority shareholdings in rivals by common institutional investors falls within the paradigm of "diffuse" common ownership. In other words, the distinct driver of the potential anticompetitive effects of diffuse common ownership is diversification and not investor concentration as such, as is the case in instances of concentrated common ownership. Taking the distinctive characteristics of the two varieties of common ownership as well as the circumstances and assumptions under which each is plausible to emerge. I illustrate that while variety I neatly fits within existing paradigms of competition and corporate laws ("controlling" partial acquisitions), variety II is a new phenomenon squarely pushing their boundaries ("noncontrolling" partial acquisitions). 42 The above conceptual distinction also provides analytical support to a broader "effects-based" theory of competitive influence that is flexible yet delimited enough to capture a range of plausible unilateral effects flowing from structural changes in the ownership and governance structure of firms, which may affect their performance and also have implications for the structure and performance of product markets. 43 That is, unilateral competitive effects may arise either due to a formal change of control (concentrated common ownership) or an informal change in incentives (diffuse common ownership). Indeed, U.S. case law has explicitly recognized that anticompetitive effects of partial ownership may be brought about by many means or mechanisms and control over the partially acquired business is unnecessary for an antitrust violation under Clayton Act §7, the U.S. merger statute. This is so because the key focus of the merger inquiry is on the "effect on competition" regardless of its cause. 44 In this light, diffuse common ownership creates a unique challenge for competition law: it may theoretically lead to harmful effects on competition yet via unconventional means. In essence, both mechanisms supporting diffuse common ownership are informal: one is based on pure incentives, i.e., passivity under antitrust law (absence of active influence) and the other on partial minority control, i.e., factual control under corporate law (absence of large, active 42 See section II.B below. 43 According to Moss (n 39), there are "many mechanisms for expanding horizontal ownership and control of economic resources" that implicate merger policy and enforcement such as horizontal mergers, joint ventures and "acquisitions of partial ownership stakes", including common ownership. The U.S. Horizontal Merger Guidelines (2010) address all three scenarios of effecting structural change but note that partial acquisitions "may require a somewhat distinct analysis from that applied to full mergers or to acquisitions involving effective control." 44 United States v. Dairy Farmers of America, Inc., 426 F.3d 850, 859-862 (6 th Cir. 2005): "We, however, do not agree with the district court's conclusion that a lack of control or influence precludes a Section 7 violation. [...]even without control or influence, an acquisition may still lessen competition. [referring also to the du Pont, and Denver and Rio Grande cases] The key inquiry is the effect on competition, regardless of the cause. [...] For example, in du Pont, the Supreme Court found that even though du Pont did not have control or influence over General Motors because it no longer had voting rights, anticompetitive effects could still occur, because a group with similar interests as du Pontits shareholdersheld the voting rights. Likewise, in this case, DFA purportedly cured any potential antitrust problems in the agreement with Southern Belle by giving all of its voting rights to AFLP. This cure, however, ignores the fact that AFLP and DFA have closely aligned interests to maximize profits via anticompetitive behavior." investors). 45 What is more, a second distinguishing characteristic of diffuse common ownership is that its competitive effects arise on a cumulative basis: it is not individual minority shareholding participations in rival firms but rather their combination and aggregate effect due to the diffusion of common ownership across multiple competitors in an industry at the same time that may be problematic.
The theory of diffuse common ownership exposes an enforcement gap in competition law and a tension between "substance" and "formalism" that still today permeates antitrust doctrine in different and subtle ways. 46 The effects of diffuse common ownership go beyond a single business entity or investor and fall short of established legal conceptions of control. As such, the identified gap is partly due to legal and partly due to economic formalism 47 as they both shape rules and theoretical constructs relating to the boundaries of the firm. For one, legal concepts fundamental to the analysis of mergers and cartels such as inter-firm "control" may occasionally be under-or overinclusive in capturing effects of partial ownership, or realistically assessing whether legally separate firms are independent market actors or comprise a single entity of affiliated companies. 48 As illustrated in this article, fixation on control may be inapposite and distracting from undertaking a proper competitive assessment in cases of diffuse common ownership. As a result, rival firms that are formally separate entities as a matter of corporate form and are not part of the same business group based on majority corporate control ("single economic entity" doctrine) may still be interrelated due to structural links such as 45 Commenting on the U.S. Dairy Farmers case, scholars have criticized the unilateral effects theory of harm based solely on competitive incentives and a diversion (cost-benefit) analysis as too open-ended, over-inclusive and not accounting for complicating real-world factors. Still, they accept "a more limited application of the theory in cases where some control is evident." See Brendan J Reed, 'Private Equity Partial Acquisitions: Towards a New Antitrust Paradigm' (2010) 5 Virginia Law and Business Review 303, 325. The theory of diffuse common ownership portrayed here clearly fulfils this additional delimitation criterion. 46 Barak Orbach, 'The Durability of Formalism in Antitrust' (2015) 100(5) Iowa Law Review 2197, 2206 ("Competition-law rules that downplay competitive effects appear to run afoul of the goals of antitrust and, as such, antitrust formalism is counterintuitive [...] reliance on the legal form may distort antitrust analysis."). 47 ibid 2209-2210 ("antitrust formalism combines both legal and economic formalism and both suffer from similar vulnerabilities."). See also David F Shores, 'Economic Formalism in Antitrust Decisionmaking' (2004) 68 Albany Law Review 1053. 48 Orbach (n 46) 2206-2207. Orbach explains how the U.S. Supreme Court "replaced one formalistic rule [the "intraenterprise conspiracy doctrine" finding that firms "affiliated or integrated under common ownership" were capable of conspiring in violation of section 1] for another, holding that a firm and its wholly owned subsidiaries constitute a single entity for antitrust purposes (creating the so-called "Copperweld immunity" [or "single economic entity" doctrine in the EU])". Yet, "the Copperweld standard did not provide lower courts with guidance for common business relationships, such as those among sister companies, between agent and a principal, and in situations of partial ownership." common minority shareholdings, debt holdings or interlocking directorates that could potentially influence their conduct even if these links do not give rise to a merger. 49 Diffuse common ownership sharply showcases the potential breakdown of fundamental economic assumptions such as firm independence and own profit maximization as a universal objective of the firm. Under common ownership in oligopoly, "atomistic" firms and shareholders cannot be assumed, 50 which in turn leads to theories about an altered objective function of the firm (portfolio value maximization) and unilateral competitive effects (acrossfirm internalization of profits). 51 Once externalities between formally separate firms enter the firm objective function the "black box" view of the firm collapses with significant implications for antitrust analysis. 52 Market concentration and market control reflecting firm interactions given diffuse common shareholdings are no longer clearly a function of the nominal number of firms active in the market. 53 These insights further point to the need for affording flexibility to the interpretation of the law and infusing realism to the application of economic analysis while devising creative solutions for effectively addressing this new phenomenon to the extent it is considered to be an empirically significant problem. 49 Federico Cesare Guido Ghezzi and Chiara Picciau, 'The Curious Case of Italian Interlocking Directorates' [2020] Bocconi Legal Studies Research Paper No. 3661733 4 (referring to "common majority shareholding" in horizontal competitors controlled by the same parent company, [which] would give rise to "a corporate group"); On the "single economic entity" doctrine that provides antitrust immunity to anticompetitive agreements between companies within the same corporate group, and in the EU it also imposes intragroup liability on parent companies for any antitrust violations of their subsidiaries based on 'control' (which is presumed for almost wholly-owned or controlled subsidiaries), see Nada Ina Pauer, The Single Economic Entity Doctrine and Corporate Group Responsibility in European Antitrust Law (Kluwer Law International 2014); Carsten Koenig, 'An Economic Analysis of the Single Economic Entity Doctrine in EU Competition Law' (2017) 13 Journal of Competition Law & Economics 281. See also the recent Case C-595/18 P, The Goldman Sachs Group Inc. v European Commission, ECLI:EU:C:2021:73 (extending the "decisive influence" presumption for imputing parental liability to private equity firms holding only a minority stake but all the voting rights in non-wholly owned portfolio companies). On the broader interaction between competition law and corporate governance drawing the external and internal boundaries of the firm and their continuing evolution given modern and ever complex organizational forms such as structural links, see Florence Thépot, The Interaction Between Competition Law and Corporate Governance -Opening the 'Black Box' (Cambridge University Press 2019). 50 Romano (n 8) 394 ("Economists used to have an atomistic view of the world, with firms and markets considered as 'isolated atoms.' This conceptualization might have been an adequate heuristic in a pre-institutional-investors world, but it misrepresents the modern U.S. economy."); De La Cruz, Medina and Tang (n 25) 18 ("no jurisdiction systemically features the kind of atomistic dispersed ownership structure that still influences much of the corporate governance debate. This article is organized in five sections. Part II below defines the two varieties of common ownership, the "concentrated" and the "diffuse", presented through the lens of merger control. Then it explores the purpose and limits of merger control, illustrating gaps in the law of various jurisdictions regarding partial minority acquisitions, particularly involving "diffuse" common shareholdings. Part III discusses the economic mechanisms associated with each variety of common ownership both from the perspective of competition economics ("internalization mechanisms") and corporate governance ("transmission mechanisms"), with special emphasis on the particularities of "diffuse" common ownership. Part IV explores implications for theory, competition policy and enforcement practice, and reflects on potential policy responses. Part V closes with an epilogue.

II. Varieties of common ownership and merger control
A mainstay in competition policy is merger control. Modern enforcement practice and merger control guidelines focus on market concentration and unilateral effects theories of harm. Yet, Stigler has pointedly remarked that "outright merger" is the most comprehensive form of collusion, in the sense that merged firms permanently abandon their independence and jointly determine outputs and prices. 54 Competition laws employ rigid behavioural rules (per se) to deter horizontal price fixing and cartels across markets whereas more flexible rules are used to scrutinise horizontal mergers and structural changes only in concentrated industries. This difference in legal treatment is in tune with economic theory and principles. While it is well recognized that merger is one way in which "competitors may be able to reduce the level of competition among themselves", notably by reducing firms' incentives for competitive pricing, they can also create important efficiencies. 55 Merger laws and policy generally aim to catch and scrutinize structural changes in corporate ownership and control that may result in lasting changes in the structure of product markets and control of industries. 56  ) 2372 (noting that the antitrust laws are designed to address either "exclusion" or "collusion" [broadly defined], the latter category mainly concerned with horizontal price fixing [cartels] and mergers). 56 Tzanaki, 'Common Ownership and Minority Shareholding at the Intersection of Competition and Corporate Law: Looking Through the Past to Return to the Future?' in Marco Claudio Corradi and Julian Nowag (eds), The Intersections between Competition Law and Corporate Law and Finance (Cambridge University Press, forthcoming) 4 (fn 11). This describes the EU position, but a similar rationale applies to other jurisdictions. stark contrast to anti-cartel laws. While the laws more or less agree on the treatment of full mergers and majority share acquisitions, the discrepancies are striking when considering minority shareholding transactions. 57 The discrepancies are all the more remarkable given that there is general agreement that certain minority transactions lack the integrative efficiencies associated with controlling acquisitions. 58 Why such variance if the aims of merger control policy are common? It appears that different regimes place different emphasis on legal or economic conceptions of control (corporate versus industry control or influence) to base merger control scrutiny.
On the one hand, the legal criterion of control is used in some legal systems to create a strict dichotomy in the merger review of minority shareholdings (controlling vs non-controlling). On the other hand, notions of economic control (substance) do not fully overlap with legal definitions (form) while competition and corporate theories of control inform different but interrelated questions (ownership structure-firm performance, industry structure-market competition). Transactions involving "minority" ownership and "partial" control point to distinct issues and challenges for corporate governance and industrial organization (agency costs, partial integration). What is more, common minority shareholdings have progressively come to "fall between the cracks" of corporate and competition laws as each field specialized on separate problems inside or outside the boundaries of firms ("internal affairs" of firms and principal-agent problems, "market power" of firms and their interaction in product markets). 59 In this light, puzzling questions arise as to the purpose and scope of merger review: Why "control" as a jurisdictional criterion? Is there an "ownership threshold" that may clearly indicate (the absence of) control or competitive harm? Should merger control exclusively target controlling acquisitions involving rival firms (formal integration) or rather any tempering of their "competitive independence" due to structural changes? In other words, is competitive harm to be proxied under merger control based on a permanent change in control or also in incentives? Answers to these questions affect not only merger policy as regards traditional mergers and acquisitions but also the treatment of common ownership as a new type of structural change across firms and markets. 57

A. Varieties of common ownership compared to a full merger
In this article, I propose a basic distinction between two "varieties" of common ownershipthe "concentrated" and the "diffuse"viewed through the lens of merger control. This classification reveals the different theoretical attributes, competitive harm potential and underlying economic mechanisms for such harm to arise associated with each, 60 as well as the likelihood that these may be effectively captured under existing merger laws. 61 It is shown that the two varieties of common ownership are conceptually and analytically "separate animals".
In order to animate the subsequent discussion, it is useful to draw a graphical image of "concentrated" and "diffuse" common ownership, by comparison to the case of a full merger.
All three scenarios may be perceived as special cases of "common ownership", which directly derives from the theory of "partial ownership". 62 Common ownership, or "horizontal shareholding" as has been dubbed in legal scholarship, may be defined as the simultaneous holding of (part of the) shares of competing firms by the same set of third-party investors. 63 The critical difference is that in cases of common ownership the rival firms are typically having "partially" overlapping shareholders-owners whereas in a merger they are "completely" overlapping by definition. What further distinguishes "concentrated" from "diffuse" common ownership is that the distribution of the partially overlapping shareholding interests across the rival firms and the concentration of control within them may be rather asymmetric (variety I) or almost symmetric (variety II).
More specifically, in a "full" merger or "complete" acquisition a (set of) common shareholderinvestor(s) comes to fully own and control post-merger the two firms that were previously independent. Said differently, the merged firms have the same common owner(s) post-merger.
In case of a "complete union" of the rival firms, the common owner-controller also happens to be a "sole owner" (100% ownership) of each firm, as shown in Figure 1 64 We focus on the case of a single common owner-controller of the merging firms for simplicity of exposition and for easier comparison to the attributes of the concentrated and diffuse varieties of common ownership. However, the case of multiple common owners-controllers in a merger is analogous. The only difference is that post-merger they share the full ownership and control of the previously rival firms. Yet, since the common shareholders are fully overlapping, their financial interests are identical. So, the analysis for practical purposes is the same. By definition, common ownership typically refers to cases of partial (less than 100%) ownership in at least one of the commonly held firms (otherwise we would simply have a full merger). 65 In the case of "concentrated" common ownership shown in Figure 2, the common shareholder(s) is depicted to have (up to) full ownership and control (as a 100% sole owner and sole controller) over one of the commonly held firms yet a totally passive (non-controlling) interest in the other competing firm. 66 In theory, it is also possible that the common shareholder(s) may have some control rather than a passive stake in the second rival firm. 67 Yet, the critical attribute of concentrated common ownership is that the proportion of ownership interests and level of control in the two commonly held firms is asymmetric. 68 65 Daniel P O'Brien and Steven C Salop, 'Competitive Effects of Partial Ownership: Financial Interest and Corporate Control' (2000) 67 Antitrust Law Journal 559, 563: "unlike most merger analysis, a central part of the analysis of partial ownership is an assessment of which owners have what type of control over the corporation and how this control translates into management decisions." 66 We assume the non-controlling stake is of <10% level based on the legal analysis of merger control regimes developed in section II.B. Also, the 100% assumption regarding the controlled firm is for illustrative purposes (an assumption of >50% majority ownership and control would also typically do) to make the analysis more tractable by having full ownership and control over one firm (sole owner) and a small ownership stake with no formal control over the other rival firm (passive owner), for reasons that will become clear in section III.A. 67 We focus on the simple case of a single common owner of two commonly held firms for ease of exposition and because it has been more extensively treated in existing literature. See David Gilo, 'The Anticompetitive Effect of Passive Investment' (2000) 99(1) Michigan Law Review 1; O'Brien and Salop (n 65). Gilo analyzes the case of a single common owner being "sole controller" over firm 1 and "passive owner" in firm 2 whereas O'Brien and Salop examine a broader range of combinations of partial ownership and control including that of a single common owner being a "sole owner-sole controller" of firm 1 and a "partial owner-partial controller" of firm 2. In what follows, we employ Gilo's paradigm as the baseline for defining and analyzing "concentrated" common ownership, in order to illustrate more sharply the differences in comparing its qualities against the "diffuse" variety of common ownership that consists of "passive" minority shareholdings across all invested firms. 68 The same rationale applies to the case with multiple common owners under concentrated common ownership, but the analysis is more complex. By contrast, in the case of "diffuse" common ownership shown in Figure 3, the common shareholders have minority ownership and control (below 50%) in both commonly held firms at the same time. 69 That is, ownership and control of diffuse common owners in all the rival firms is partial and formally "non-controlling" on a standalone basis. 70 It follows that there are multiple common owners that share such partial ownership and control across the rival firms, 71 and there is no "sole owner" or "sole controller" (either as a common or non-common shareholder) in any of the commonly held firms. 72 It also follows that any common control of 69 O'Brien and Waehrer (n 12) 731: "While it is widely accepted that common ownership can have anticompetitive effects when the owners have control over at least one of the firms they own (a complete merger is a special case), antitrust authorities historically have taken limited interest in common ownership by minority shareholders whose control seems to be limited to voting rights. Thus, if the empirical findings [...] in the emerging research are correct and robust, they could have dramatic implications for the antitrust analysis of mergers and acquisitions. The findings could be interpreted to suggest that antitrust authorities should scrutinize [...] also situations in which all of the common owner's shareholdings are small minority positions."; Rosati and others (n 26) 15. 70 The common thread weaving together both varieties of common ownership and full mergers is their conceptual link to the theory of partial ownership. See O'Brien and Salop (n 65). Accordingly, the two core parameters determining the scope and magnitude of unilateral competitive effects in all these cases consist of: i) a degree of collective financial interest; and ii) some measure of common control, produced by the simultaneous investments in rival firms and the resulting (partial) shareholder overlaps across firms. The novelty of "diffuse" common ownership is that both the common interest and the common control in all the interlinked firms is partial. 71 Unlike "concentrated" common ownership that may also exist in the presence of a single common owner of two rival firms, having no control over the one (passive investment) and total control over the other (either under complete (100%) ownership as in a full merger, or majority (>50%) ownership as in a controlling acquisition). 72 This means that the "sole owner"-"sole controller" paradigm (that may fit the "concentrated" common ownership variety because formal control, either under complete or majority ownership, may be established in at least one of the commonly held firms) is not appropriate for assessing "diffuse" common ownership that rests on a lack of large asymmetric blockholders. Given the dispersed shareholder structure in situations of "diffuse" common ownership, note that 100% (full) "sole ownership" (or equivalently majority ownership and sole control) of all commonly held firms is not possible but also that "no control" whatsoever (passive ownership) across all commonly held firms is equally impossible (as corporate control has to lie with some shareholder(s) eventually). diffuse common owners in individual firms is informal (i.e., factual). Thus, the truly novel and distinctive characteristics of diffuse common ownership are that, depending on the context that it arises (i.e., dispersed ownership of large public corporations, widespread portfolio diversification by passive institutional investors such as index funds): i) the common shareholders' interests in the rival firms may be parallel and similar or symmetric (roughly equal, which makes not to matter what or how low the level of the individual common shareholdings may be 73 ), whereas ii) corporate control is typically proportional (to share ownership based on the standard "one share-one vote" corporate norm), 74 relative (to other shareholders rather than absolute) and equally shared among the common shareholders (de facto joint control absent larger non-common shareholders). 75 As a result, "diffuse" common ownership entails partial ownership and partial control precisely because of the (corporate ownership and governance) context in which it arises. 73 From the perspective of common shareholders, what matters is their total portfolio profits and relative financial interest in the rival firms given the size of their individual common shareholdings. If such shareholdings are symmetric, the level of ownership participation or control becomes irrelevant as individual firm profits are equally internalized and control is equally shared among the common owners in the absence of more powerful undiversified shareholders. See Julio J Rotemberg, 'Financial Transaction Costs and Industrial Performance' [1984] MIT Sloan School of Management Working Paper No. 1554-84 (considering the case of ex ante identical diversified shareholders with equal stakes in all [symmetric] firms). 74 Backus, Conlon and Sinkinson, 'Common Ownership in America' (n 16) 7-8. 75 O'Brien and Waehrer (n 12) 741: "a common owner's influence over the manager rises as the other owners' shareholdings become more diffuse.". Absent a large dominant shareholder in firm governance (with majority ownership and control) and given the corporate law "one-share-one-vote" principle in the absence of special asymmetric governance structures (e.g., dual class shares), the competitive effects of common minority ownership are estimated based on a "proportional control" assumption, which essentially implies for any positive level of common ownership among rival firms there is some potentially anticompetitive effect even if produced by small, minority shareholdings. More generally, the magnitude and likelihood of the effect depends on particular corporate governance and control assumptions. Azar, Schmalz and Tecu (n 1) (using the "proportional control" assumption in empirical research to calculate the MHHI; but also testing alternative control scenarios e.g., by using Banzhaf indices of voting power [defined as the probability that a shareholder is pivotal in an election] and finding that the proportional control assumption is not driving the baseline results). There is a double rationale for the choice of terminology for the distinction between the two varieties of common ownership. "Concentrated" common ownership suggests: i) concentration of ownership (up to 100%) in a single firm relative to the other commonly held firms that are partially owned by means of lower shareholding participations (less than 100%, or in any event less than the shareholding held in the first firm) -in which case the common owner's "relative financial interest" in the linked firms is clear); ii) concentration of control in a single dominant shareholder ("sole control" -a sole 100% owner being an extreme case of concentrated common ownership, but sole control may also be established in case of "partial" majority ownership  79 For these reasons the analysis is more complex and needs to be case-specific. For instance, it may depend on the context and facts of the case to determine whether the diffuse institutional common owners are the "real" owners (residual claimants) and "actual" controllers (exercising governance power and management influence) diversification, especially due to the "enormous success" of "passive" index funds. 80 Diversification has rendered "corporate ownership both diffuse and collectivized at the same time" meaning that "many small shareholders are partial "co-owners" not only of a single but several competing corporate enterprises" given their parallel (often symmetric) common shareholdings within an industry. 81 Accordingly, diffuse common owners are interested in their total portfolio profits (rather than individual firm profits), thus having the incentive to internalize any competitive externalities among their invested rival firms (interest in industrywide returns rather than "cut-throat competition") while they follow "passive investment" and "portfolio-wide" (rather than firm-specific) governance strategies that maximize portfolio value. 82 As such, this new type of institutional and index fund common ownership has particular "hidden" properties: diffuse common shareholdings may lead to "invisible" rival profit internalization that transcends the boundaries of firms (cumulative unilateral or network-like effects) and "latent" shareholder control concentration that operates below formal legal definitions of corporate control (de facto joint control). 83 In other words, the anticompetitive mechanisms that underpin the diffuse common owners' incentives and ability to generate competitive harm do not rely on active influence or individual firm governance due to "big" and "active" equity holdings in at least one of the commonly held rival firms, as in cases of concentrated common ownership. 84 The atypical threat of diffuse common ownership by passive, diversified investors and index funds is its potential aggregate effect on competition by a web of individually "small" and "passive" but multiple and parallel common shareholdings in rival firms in oligopolistic industries, 85  shareholder power in the governance of the commonly held firms in the hands of the potentially relatively more prominent common shareholders. 86 What is more, for certain types of investors, typically diversified and passive, and certain types of firms, typically publicly listed firms with a dispersed shareholder base, corporate law and governance principles (such as rules on fiduciary duties) need not necessarily obstruct or constrain such potential effects. 87 That is, diffuse common ownership may not be constrained by corporate governance actors (noncommon shareholders, managers) and institutions. 88 In fact, latest corporate law scholarship encourages the "systematic" or "portfolio-wide" governance style of passive institutional investors, precisely for the same reason that their presence may lead to competitive harm: their ability to internalize externalities given their collective interest in their diversified portfolio of invested firms. 89 It follows from the above that the appropriate benchmark for assessing the competitive effects of the two varieties of common ownership is very different. This is because the most anticompetitive harm potential is reached: i) in case of "focused" majority ownership and concentrated shareholder power under a model of "targeted" governance and "total" sole control over at least one commonly owned firm (asymmetric common ownership and control) for "concentrated" common ownership, 90 while ii) in case of perfectly "symmetric" (i.e., equal) and parallel ownership of all the commonly held firms due to diversification (rather than within-firm shareholder concentration as such) and "portfolio" governance that entail aligned economic incentives and shared minority control among several common shareholders (symmetric common ownership and control) for "diffuse" common ownership, other things 86 It is noted that the index fund (and other individual fund) shares are voted centrally at the fund family level in a coordinated manner while separate large institutional investors with similarly diversified interests may also vote in a similar (and possibly coordinated) way or act in congruence in their governance activities, although such coordination need not be explicit. being equal. 91 Consequently, a "sole owner"-"no agency cost" paradigm that may fit well when analyzing concentrated common ownership cases with a single dominant shareholder disciplining corporate management may be rather unsuitable or misleading in evaluating diffuse common ownership's harm potential. 92 Given the presence of many common owners in many rival firms at the same time with "mirroring" diversified portfolios, 93 a "perfect symmetry" benchmark is more apt to capture aggregate effects that given the size of individual shareholding and level of standalone control may seem implausible or improbable at first sight.
Another way to see this is that "concentrated" common ownership cases require a "mergerlike" analysis given the clear yet disproportionate legal control established over one of the commonly held firms (asymmetric solely controlling acquisition) whereas "diffuse" common ownership cases call for a "joint venture-like" analysis given the symmetric yet factual minority control that underpins the parallel (similar if not identical) interests in the many commonly held rival firms (de facto jointly controlling acquisition). 94 It is notable that in the special case of a full merger (or a joint venture), these qualities happen to coincide: full common control and identical financial interests are found in the complete union of the previously independent merging firms. 95 Unlike a full merger (or a joint venture) that is presumed to lead to joint profit maximization post-merger, common ownership may be perceived as an effective "partial merger" presumably leading to "partial" internalization of rival profits. 96 In this context, therefore, the critical difference between concentrated and diffuse common ownership is that this "partial" internalization of rival profits is asymmetric in the first case while more symmetric in the latter.
In this light but seen from a post-merger (ex post legal) perspective, "concentrated" and "diffuse" common ownership may be thought to roughly correspond to situations of partial acquisitions involving "controlling" and "non-controlling" shareholdings (on a standalone basis) respectively. 97 The next section moves on to look at the legal treatment of majority and minority acquisitions under diverse merger control regimes. This comparative analysis aims to specify the extent to which legal conceptions of control and applicable merger thresholds may be able to capture partial acquisitions relating to the two varieties of common ownership.

B. The purpose and limits of merger control
Merger review is the "one-off", usually ex ante, "process to determine whether a more durable combination of previously independent assets is likely to materially change incentives as to how the assets are used in the competitive process". 98 The aim of merger control is thus to target and scrutinize transactions such as full mergers or acquisitions of ownership and control that are "sufficiently material", in terms of size of the parties or the transaction or shareholding, and "may harm competition" through structural changes in the market that may create durable market power. 99 The concept of "control" is a key foundation both for the legal definition of a notifiable merger transaction and also for the economic theories of harm associated with mergers and acquisitions. The underlying economic logic is that "in most horizontal mergers, two competitors come under common ownership and control, completely and permanently eliminating competition between them". 100 Accordingly, legal jurisdictions generally agree that full mergers and "majority" acquisitions be subject to their merger control rules.
In contrast, the treatment of "minority" acquisitions varies greatly under different merger control regimes. Certain forms of "controlling" or competitively "influential" minority shareholdings are to different degrees captured by merger control statutes. On the other hand, although the potential anticompetitive effects of "non-controlling" or purely "passive" shareholdings have long been recognized, the harm potential is often not considered likely, ibid 5-6. Merger control regimes use different "objective" (numerical) criteria and/or more "economic" criteria (open-ended standards) to single out M&A transactions for review. The first category is used to specify (ownership) percentage thresholds for share acquisitions in the target, while the second category aims to select potentially problematic transactions (e.g., "by focusing on whether a transaction will enable a firm to acquire the ability to exercise some form of influence over a previously independent firm"). 100     The most conservative is the EU approach that employs a "decisive influence" test to determine which transactions fall within its merger control regime. Under the EU Merger Regulation ("EUMR"), 106 the "possibility of exercising decisive influence" may be established either on a standalone basis ("sole control") 107 or jointly with other shareholders ("joint control").
Acquisitions of below 50% of voting shares may lead to a finding of effective control ("de facto control") if the remaining shareholder base is very dispersed and the acquirer has in practice the largest (minority) stake in the target, which effectively means that it is "highly likely to achieve a lasting majority of the votes cast at the shareholders' meetings, given the evidenced presence of shareholders at past meetings and the voting patterns in previous 106 Council Regulation ( years". 108 "Joint control" may be found on the basis of : i) "equal voting or board representation rights", i.e., equality of two parent companies in a joint venture; ii) "strategic veto rights", i.e., power to block strategic decisions, when a supermajority of votes is required, resulting in deadlock situations (de jure blocking power); iii) "joint exercise of voting rights" or "stable coalitions" between minority shareholders, i.e., if they act together as a group and are able to jointly achieve an ex ante certain and stable majority in corporate decision-making (majority voting bloc); or iv) "strong common (strategic) interests", i.e., if they are expected not to act against each other in exercising their rights and are required to cooperate in practice (de facto collective action). 109 "Non-controlling" minority shareholdings or "changing (voting) coalitions" are not captured by the EUMR. 110 However, in recent reform proposals the European Commission considered of extending the EUMR to shareholdings giving rise to a "de facto blocking minority" (as in some Member States, below a legal threshold of 25%) or shareholdings above 5% that combined with "additional factors" may establish a "competitively significant link". 111 EU Member States such as Germany and Austria apply lower control thresholds than "decisive influence". In Germany, any acquisition of shares of 25% is automatically subject to its merger control rules, whereas shareholdings below 25% are reviewable if they give rise to a "competitively significant influence", 112 which is construed as a position of de facto influence comparable to that of a shareholder of 25% shares or voting rights. 113 In practice, this latter test only rarely will capture minority interests below 10%, 114 although, in theory, there is no "safe harbor". 115 Furthermore, "competitively significant influence" may be found to be exercised jointly by several companies when a "joint possibility of influence" is established on a de facto basis or a "common interest which goes beyond the joint participation as such" is evident. 116 Similarly, UK merger control rules may apply to acquisitions of minority shareholdings that confer the ability to exercise "material influence" over the target. 117 Such influence is presumed for shareholdings with voting rights above 25% but may also be found for shareholdings of 15% or more (e.g., based on the "acquirer's ability to influence the target's policy through exercising voting rights at shareholders' meetings", together with "any additional supporting factors"). 118 Exceptionally, shareholdings below 15% may attract scrutiny. The CMA has wide "discretion" in applying the "material influence" test and in theory there is no minimum shareholding threshold that excludes such influence. 119 The U.S. merger control regime is the most far-reaching. In the United States, any acquisition of stock is subject to scrutiny and may be challenged under Section 7 of the Clayton Act, where the effect "may be substantially to lessen competition". 120 The Horizontal Merger Guidelines include a section on "partial acquisitions" 121 that provides the circumstances under which these are analysed as mergers, i.e., if they result in "effective control", or pursuant to a distinct analysis considering "any way they may affect competition". In the latter case, the US agencies focus on "three principal effects": i) the acquirer's ability to influence the competitive conduct of the target; ii) a reduction in the acquirer's incentive to compete; iii) the acquirer's access to the target's non-public, competitively sensitive information. 122 The Hart-Scott-Rodino ("HSR") Act 123 requires premerger notification for acquisitions of "voting securities" above certain thresholds, irrespective of obtaining any control or influence 124 or the existence of a "competitive link" between the parties. 125  10% or less that are made "solely for the purpose of investment" 126 (filing exemption) or analogously acquisitions of 15% or less made by "institutional investors". 127 Although the US regime provides for a similar "solely for the purpose of investment" exemption from liability under Section 7 of the Clayton Act (substantive exemption), this is inapplicable in case of "any influence", e.g., by "passive" institutional investors who are "active owners" in their governance activities, or "actual anticompetitive effects" even after completion of the acquisition. 128 Interestingly, the U.S. antitrust agencies have recently proposed two amendments to the HSR Act reporting rules: 1) requiring aggregation of holdings of all "associates" 129 within the more-broadly defined acquiring "person"; and 2) introducing a new "de minimis exemption" for acquisitions 10% or less "without an examination of intent", given that "they are unlikely to violate the antitrust laws", unless: i) the acquiring person has a "competitively significant relationship" with the issuer (cross-ownership); or ii) "the acquiring person (and its associates) hold more than 1% in a competitor of the issuer on an aggregate basis" (common ownership); or iii) "someone from the acquiring person is an officer or director of the issuer or a competitor of the issuer" (interlocking directorates). 130 The second exception to the proposed exemption seeks to "ensure the Agencies receive filings that provide insights into the influence of holdings in competitors" given the current common ownership debate.

III. Mechanisms of Common Ownership
In order to fully appreciate the subtleties of the basic distinction between concentrated and diffuse common ownership proposed in this article, it is useful to understand the precise mechanisms and modes of operation under which each variety of common ownership: (A) may alter incentives to compete for common shareholders and their commonly held rival firms and produce competitive harm (internalization mechanisms), and also (B) may influence corporate 126 15 U.S.C. § 18a(c)(9). The exemption applies if the [holder or acquirer] "has no intention to participate in the formulation, determination, or direction of the basic business decisions of the issuer" (16 C.F.R. § 801(1)(i)). 127 16 C.F.R. § 802.64(b). 128 Elhauge, 'Horizontal Shareholding' (n 10) 1305-1312 (analyzing the scope of the substantive and filing "passive investment" exemptions, and suggesting a change in the US agencies' interpretation so that filing is required "whenever a set of large shareholders plans to vote shares that, in aggregate, are more than 10% of the stock in multiple competing corporations". Although merely voting does not automatically exclude "passive intent", the filing exemption is currently inapplicable when, among others, there are: i) interlocking directorates; or ii) cross-ownership of rivals). 129 Previously, reporting was required only for "associates" with controlling or minority interests in entities active in the same line of business as the target. 130 FTC Notice of Proposed Rulemaking (n 22). management and firm behavior within the governance of the commonly held firms (transmission mechanisms). 131 The following sections closely examine these mechanisms.

A. Competition effects and internalization mechanisms
The above merger control analysis creates the impression that minority acquisitions that involve small shareholdings (e.g., below 10%) or afford less intense degrees and kinds of influence ability (e.g., "non-controlling", "passive" or de facto "influential" shareholdings 132 ) may produce limited and ad hoc effects on competition and welfare. Hence, their generally privileged treatment under merger laws. While such views are common in the literature, they only represent an overgeneralization of the competitive implications of partial acquisitions, and diffuse common ownership in particular, that is not justified in principle. For instance, this assessment may be accurate in some cases when applied to individual minority shareholdings seen in isolation. Yet as a general matter, the economic analysis is more complex (e.g., in case of multiple, widespread common shareholding links among rivals in an oligopoly) and needs to be granular (i.e., case and context specific). This in turn means that the economic and legal or institutional context and specific details in each individual case may matter to determine the impact of partial acquisitions on both industry and firm performance.
More specifically, the potential competition effects of partial acquisitions are the product of three particular factors: i) market structure, ii) ownership structure, and iii) governance structure. 133 For instance, in (almost) perfectly competitive markets the effects of such acquisitions between actual or potential competitors may be unlikely or negligible as aggressive product market competition (and the competitive constraints posed by other independent competitors) is expected to discipline any anticompetitive instincts of the partially linked rival firms. Furthermore, in corporate settings where there is a discernibly dominant shareholder with total control of the firm (or a homogenous group of shareholders with majority control over the board of directors and corporate management), the likely competitive threat or impact of minority shareholding acquisitions is also considered to be insignificant or 131 See n 19 above. 132  inconsequential in terms of its implications for business strategy or general firm governance.
Under these conditions, minority shareholdings albeit in rivals can be safely considered "passive" in the antitrust sense (no market power motive or effect) in that they are not prone to produce material competitive concerns. 134 In this light, a public policy supporting the more lenient legal treatment of small, noncontrolling minority acquisitions and justifying the current merger law structure in certain jurisdictions may implicitly rest on the following double premise and default assumptions: i) profits and minimize cost and managerial slack. 137 In fact, intense product market competition may as a general matter reduce "private benefits of control", meaning any kind of corporate "agency costs" regardless of whether their source is management (usually in publicly listed, widely held firms) or controlling blockholders (in firms with concentrated ownership) being in control of the firm. 138 Against this backdrop, it may be accurate to state that the minority shareholdings' threat to competition is not continually present or substantial. 139 Also, the line drawn between controlling and non-controlling acquisitions and the interpretation of the Horizontal Merger Guidelines in this regard becomes meaningful. That is, merger policy generally recognizes that unlike full mergers or controlling acquisitions where independent competition is "completely and permanently" substituted by common control, 140 partial minority acquisitions may not "necessarily or completely" eliminate competition between the parties to the transaction. 141 Said differently, "minority" ownership and "partial" control may prove problematic only in certain but not all market and corporate settings. This further suggests that when the ownership or control acquired is not complete, the absolute size of the ownership or control stake in isolation is not a good proxy for (the lack of) competitive harm. Indeed, the key driver of any is not to suggest that they will always or usually be present or substantial.
[…] Indeed, more lenient treatment might be defended on the ground that the competitive threat is weaker. Unfortunately, there is no formula that can describe the likelihood of such effects for the generality of cases or even for the particular case." This position is similar to the EU's in its 2001 Green Paper (n 101) concluding against a holistic change of the ex ante EU merger control regime to address non-controlling shareholdings. In its 2014 White Paper (n 102) however, the Commission proposed a more flexible "targeted transparency regime" for shareholdings above 5% that qualify as a "competitively significant link" in connection with "additional factors". Areeda and Hovenkamp's solution to the practical administrability concerns (difficulty in proving partial control or influence and quantifying anticompetitive effects) is a "structural presumption". Any partial interest above 5% is considered "substantial" and to be analyzed as if it were a full or controlling acquisition (assuming control), while creating a 5% de minimis safe harbor for thus (conclusively presumed) "passive" financial interests. 140  effects on competition will be the industry structure in combination with the relative ownership or control stake in a rival firm. 142 At the same time, differences in the structure of merger control systems could be in principle justified, at least in part, based on the relative empirical prevalence and potency of the above two forces (vibrant market competition and strong corporate governance) and related institutional and organizational factors (e.g., varieties of capitalism, embracing to varying degrees a shareholder or stakeholder model of governance and corporate regulation, the strength and centrality of market forces and the relative proportion of firms with concentrated or widely dispersed ownership) 143 in each specific country or jurisdiction. 144 Importantly, however, a sweepingly generous policy stance towards non-controlling partial acquisitions is not defensible in other industry (concentrated markets) 145  . 144 Seen in this light, differences between merger control systems may be logically reconcilable. EU regimes using influence-based thresholds for merger scrutiny of minority acquisitions may be the result of path dependence and the relative predominance of concentrated ownership and control structures among continental European firms whereas U.S. merger law capturing any partial acquisitions above and beyond a criterion of control or influence may be understood given the strong presence of external capital markets, the historical absence of blockholders due to legal and political restrictions and the generally fragmented and diffuse ownership structure of U.S. publicly listed firms, factors which thus indicate the increased and realistic possibility that lower levels of shareholding may raise competitive concerns. The hybrid case of the UK regime with influence-based, although flexible, merger control thresholds can be seen as the combination of: i) currently dispersed ownership structures but earlier family-dominated firms with ownership and control structures closer to continental structures than to American ones, and also ii) the greater interaction with continental EU merger control systems. On the roots and potential persistence of country-specific ownership and governance patterns, see Roe, ownership of the Big Three in the U.S. and finding that together they constitute the largest shareholder in 88 percent of the S&P 500 firms); Banal-Estañol, Boot and Seldeslachts (n 29) (finding overall that common investors have gained importance in Europe but non-common investors [governments, individuals, corporations] still remain important, particularly so in certain EU countries). See also Roe, 'The Shareholder Wealth Maximization Norm and Industrial Organization' (n 136), who suggests that greater skepticism towards the shareholder wealth maximization norm in continental Europe may be explained, among others, by their historically and comparatively less competitive product markets and more concentrated industry; and noting that changes in the relative product market concentration (e.g., towards more competitive structures) may increase demand or tolerance for shareholder primacy institutions in Europe. These ongoing shifts in markets and institutions could (partially) rationalize policy discussions in the U.S. to expand the reporting requirements under merger control and in the EU to potentially extend the scope of the EUMR. 145 146 where the afore-mentioned assumptions are not tenable either theoretically or empirically. Specifically, in oligopolistic markets with high entry barriers shareholding links between actual or potential competitors may have clear competitive implications as they are likely to lead to reduced output and higher prices. 147 Indeed, even acquisitions of small and purely "financial interests" ("silent minority shareholdings" or "passive investments") 148 in a rival are expected to alter the acquirer's incentives to competewithout any collusion, communication or control prerequisites -resulting in unilateral priceincreasing effects. The reason is that the acquirer will take into account the financial interest (level of the non-controlling minority shareholding) in the rival while setting its market strategy and the effect its business decisions may have on the profits of the rival that the acquirer now stands to partially internalize as a return on its passive investment.
Therefore, "passive" investments in competitors in oligopolies in the corporate sense (no influence) are not really passive in the antitrust sense (competition effects): 149 the unilateral pricing effects may be quantitatively lower than in case of full or partial mergers, 150 but they are always predicted. 151 In this case, even absent any control or influence over the partially acquired rival the shareholding link induces the acquirer to competitive behavior that is less aggressive (reduced incentives to expand market share or lower price). 152 On the same logic but via different means (and unilateral theories of harm), this price-increasing effect may also be produced indirectly as the shareholding may diminish the intensity of competition by profits to share with a rival firm (via a passive investment) since "competition [drives] price all the way down to marginal cost"; the acquirer "places no weight" in its shareholding in the rival and there are no unilateral effects). 146 As it will be explained in the next section. 147 Reynolds and Snapp (n 94) 142: "the effects are purely structural: they arise not because of increased opportunities for collusion or changes in the concentration of control, but because the linking of profits gives each firm an incentive to compete less vigorously and adopt behavior more conducive to joint profit maximization than otherwise would be the case." 148 All this alternative terminology refers to the same phenomenon: the holding of equity interests in a firm without any corresponding control rights (e.g., non-voting stock affecting strategic variables other than price or quantity. For instance, a non-controlling shareholding may reduce competition when firms compete in non-price dimensions (e.g., innovation or quality) and may also lessen incentives to compete with the partially acquired rival over geographic markets or entail reduced incentives for the acquirer to enter the incumbent firm's market in which it holds a financial interest. 153 In all these cases, existing or potential competitive constraints are effectively reduced.
Essentially the shareholding link similarly to a full merger may produce a softening of competition, due to the "internalization of competitive externalities" 154 it induces and the tempering of the natural "business-stealing" 155 instinct among competitors, to the detriment of consumers. These unilateral effects are purely "structural", 156 i.e., they depend solely on the partial or common owner's incentive structure and not on any further (governance or strategic) action by the acquirer or the partially acquired firm, and as such they are "probabilistic" in nature. 157 Another way to see this is that the financial interest in the rival creates an ("What makes the merger anticompetitive is that it internalizes the rivalry between the merging firms and thereby causes them to alter their actions."). On the "Cournot merger paradox" suggesting that since competition is eliminated between the merging parties and their output is reduced due to the merger, rivals come to benefit from this output restriction in the post-merger equilibrium because they expand and capture all private gains from the merger, and thus we can presume efficiencies (and an increase in total welfare) for mergers actually taking place, see respectively . This is because mergers given partial or common shareholding may be profitable overall for the shareholders of the acquiring firm that are also invested in the rival target firm (and as a result share in its profits and increased value), although the transaction as such may be unprofitable for the acquiring firm. This fact however may put into question the policy presumption about the private profitability of mergers in the presence of common ownership. See Azar and Tzanaki (n 10). 155 Anton and others (n 6). 156 Elhauge, 'Horizontal Shareholding' (n 10) 1270, 1274, 1302 ("The anticompetitive incentive created by this horizontal shareholding is purely structural, changing the price-setting incentive of each firm acting separately. [...] The basic anticompetitive effects arise from the fact that interlocking shareholdings diminish each individual firm's incentives to cut prices or expand output by increasing the costs of taking away sales from rivals."). 157 Gilo (n 67) 31-33. As Gilo emphasizes, the anticompetitive effects of passive investment are "probabilistic in nature" as is prospective merger control review, which is conducted in the U.S. pursuant to an "incipiency test" that requires only "likely" adverse effects on competition; however, partial stock acquisitions that fall within the "opportunity cost" to the acquiring firm of increasing its output or reducing its price. 158 If in turn the acquirer increases output or lowers price, it will divert customers and sales away from the target firm reducing the rival's profits and accordingly its own share in such profits as an investor. 159 Thus the acquirer will weigh the (own) additional profits versus the (internalized) opportunity costs of a potential price increase in deciding its strategy given the minority shareholding in the rival. 160 In terms of quantitative impact, the output and price effects of a single small, non-controlling shareholding acquisition in a rival may be (the most) modest, compared to a full merger or a controlling acquisition. 161 However, the magnitude of the unilateral effects of multiple minority shareholdings on equilibrium output levels may be significant depending on a number of factors such as: i) the number of firms in the market (market concentration), ii) the number of firms linked (cross-or common ownership), iii) the level of the shareholding links (percentage ownership and control interests), 162 iv) the reciprocity of such links (mutuality of the "solely for investment" exemption, require proof of an "actual" lessening of competition. Gilo critically notes that the unilateral effects and acquirer's strategic motives of such acquisitions are ignored by the case law (that in the absence of active influence over the target presume that the stock acquisitions are passive or harmless) while this is not justified by their economic analysis. Given the difficulty to prove actual effects that he considers "tantamount to a de facto exemption for all passive stock acquisitions", his recommendation is that "such stock acquisition[s] must be scrutinized under the main effects clause of section 7 of the Clayton Act. That is, there must be a full-blown investigation of market conditions to establish whether the stock acquisition, although passive, may (in the probabilistic sense) substantially lessen competition.". Elhauge, 'Horizontal Shareholding' (n 10) 1307-1308, on the other hand, suggests that empirical economic evidence showing (unilateral) anticompetitive effects may satisfy the "actual" effects test applied to presumably purely passive acquisitions, that "would negate the [substantive] passive investor exception and leave the horizontal shareholders subject to challenge under § 7 of the Clayton Act." 158 O'Brien and Salop (n 65) 607. In a sense, the acquirer may in functional terms "inflate" its own "cost structure" via the partial shareholding acquisition. 159  : "Another factor that affects the level of output reductions is the level of the equity ownership. The higher the level of ownership, the higher the incentives of the firms to lower their output given the output of the other firms.". However, control and the relative ability of the acquirer to over-or underrepresent its partial ownership stake in a rival may "disrupt" this linear, progressive relationship between the (nominal) level of ownership acquired and the degree of competitive effects (internalized rival profits internalization as in a full merger or a 50/50 joint venture), 163 and iv) the firm's (and manager's) objective function given the partial minority shareholding. 164 For instance, the unilateral effects may be of limited magnitude when "few firms are linked, and those links are small" but significantly greater "when the links include virtually all the firms in the market". 165 In the limit, "when ownership shares are at the maximum level which is feasible, given the number of firms in the market, the monopoly output level will result regardless of the number of firms". 166 This result suggests that in case shareholding links among competitors are pervasive in an industry, the number of firms operating in the market may indicate an oligopoly structure, but the competitive effect produced by those inter-firm linkages indicates a monopoly outcome, i.e., profit-maximizing firm operation equivalent to that of a monopolist. 167 On the other hand, This assumption is only valid in special cases, however"). Yet, "proportional control" may be justified in certain settings (absent large dominant shareholders in firm governance, and dual class stock or other asymmetric governance structures and given the corporate decision-making norm of "one share-one vote" As a general matter, acquiring additional minority shareholdings in other rival firms in the market or (the controllers of) rival firms simultaneously holding non-controlling shareholdings in further competitors tends to reinforce the unilateral effects as the "network" of partial shareholdings (number of links) in the market will increase. Similarly, if the level of shareholdings held in horizontal competitors increases, the extent of "internalization" of rivals' profits among the linked firms (level of links) will also increase. 169 In addition, the acquisition of more symmetric links (symmetry of links) or the presence of more symmetric ownership structures in the industry corroborate such unilateral effects, by increasing the similarity in equity share positions held by each investor (same percentage shareholding or equal financial at capturing this increase in "effective" concentration and market power due to partial crossor common ownership. 168 See Azar and Tzanaki (n 10). It is also noted that while the overall welfare effects and the general equilibrium effects of common ownership within and across industries may be more mixed or nuanced, competition policy focuses on consumer welfare and competition enforcement is "market-specific" in that only efficiency gains within the same relevant market (and for the same group of consumers) may offset potential anticompetitive unilateral effects (consumer harm) found in that market. Efficiencies associated with common ownership (e.g., improved corporate governance, greater diversification, increased liquidity) may be substantial but are "out-ofmarket" efficiencies and as such generally not credited by antitrust enforcers. Besides, common ownership is unlikely to generate "merger-like" synergies. In addition, as Baker notes, within-industry diversification benefits are generally limited because stock and profits of rival firms in the same industry are highly positively correlated and if common ownership lessens competition, these diversification benefits are further reduced because the positive correlation in profits across firms increases. See Elhauge, 'Horizontal Shareholding' (n 10) 1303-1304; Baker (n 10) 227-231; US Horizontal Merger Guidelines 2010 §13 ("partial acquisitions usually do not enable many of the types of efficiencies associated with mergers"). Other welfare enhancing and competition relevant efficiencies (that exist in the presence of positive spillovers that may be internalized due to common ownership) such as cost reducing R&D investment and innovation, are unlikely to offset anticompetitive harm in industries with high concentration and low levels of spillovers, see López and Vives (n 6 interests) across the linked firms and the uniformity of portfolios held by all common investors (same shareholding positions held in the same set of firms). 170 Estimating the "degree of internalization" of rivals' profits (i.e., the "profit weight") due to partial or common shareholding is a critical starting point for unilateral and competitive effects analysis. 171 This is the weight the acquirer (a firm or a firm's controller) places on the partially 170 Therefore, both the symmetry in equity positions and investor portfolios across firms will affect and increase profit internalization, other things being equal. See Boller and Morton (n 20) 6-7 ("One interesting property of MHHI is its sensitivity to ownership symmetry. If common owners are exactly symmetric in holding the same percentage of the same set of companies, ownership is equal to control, and other owners [retail investors] are atomistic, then in this model the monopoly outcome is achieved. This is true whether the common owners each hold 2% or 20% of the competing companies." 171 Strategic or coordinated effects may also arise from purely financial interests without any additional control or explicit information exchange. First, this is because with a non-controlling shareholding acquisition at period one of a (multi-period) game, the acquirer alters its own incentive structure which in later periods may influence the rival's strategic (re)actions but without directly affecting the rival's profit or objective function, i.e., it does not alter the rival's incentives or opportunities (which could occur in the case with control). In an oligopolistic environment with repeated interaction among rivals in a non-cooperative game where history matters, the investment has "commitment value" and operates as a "sunk cost" that is to the benefit of the acquirer and "preempts" the rival's future choices. The strategic incentives of firms to engage in such "self-manipulation" (incentives to over-or under-invest) will largely depend on the nature of competition among the rivals (e.g., "strategic substitutes" or "strategic complements" The RAND Journal of Economics 304. Furthermore, even a unilateral increase in cross-ownership holdings, by one firm but not by others, that results in a more asymmetric incentive structure will typically facilitate collusion. This is because an increased shareholding in rivals will generally make it "less attractive to deviate from a collusive price" and also "less feasible to escape punishments". In essence, both the incentives to collude and the incentives to deviate will be positively affected by the shareholding acquisition. This result is robust and depends precisely on the fact that such shareholdings are non-controlling, in contrast to full mergers (that if they are asymmetry-increasing, they make collusion less likely). In addition, analysis of cross-shareholdings based on Nash reversion punishment strategies can be misleading. Further, a profit weight exceeding one implies asymmetric internalization and inflated overlapping ownership incentives, i.e., incentives for shifting profits across the interlocked firms and thus expropriation of atomistic, undiversified shareholders, due to the outsized relative control ability of the overlapping owners ("tunneling" or private benefits of control). 176 This result arises in situations of partial ownership with "total control" as pointed out by O'Brien and Salop because the misalignment of ownership and control creates a "free-rider problem" and for this reason, distorted incentives and the least competitive outcomes. 177 It is coordination is the result of (partial) coordination among the linked firms only and not industry-wide (in which a case coordinated and non-coordinated effects are mutually exclusive) because the same factors inform the analysis, see Maier-Rigaud, Schwalbe and Forster (n 160) 254-255. 172 Backus, Conlon and Sinkinson, 'Common Ownership in America' (n 16) 4 ("the profit weights approach, which starts with the objective function of the firm, is the only one that offers a fully general path forward for empirical study of the common ownership hypothesis. [...] The theory goes back as far as Rotemberg [1984], is implicit in the MHHI measure of Bresnahan and Salop [1986], has been applied to cross-ownership in O'Brien and Salop [2000], and has seen application in various tests of the common ownership hypothesis."); Vives (n 8) 3 ("It is the weight of the profit of firm k in the objective function of the manager of firm j relative to the own profit of firm j. The relative concentration of ownership and control in firm k versus firm j is what determines the coefficient's value"). 173  O'Brien and Salop analyze total control in case of partial cross-ownership, which implicitly assumes that the acquiring firm, which obtains total control over the rival acquired firm, is fully also notable that while private benefits of control have typically been associated with concentrated ownership (large dominant shareholders), 178 now incentives for tunneling may also be found in widely held, public firm settings due to common ownership even in the absence of an outright controlling interest (minority common owners). 179 The mechanics of this free-rider effect become more clear in case of acquisition of a purely financial interest in a rival by a firm's "controller" 180 as a passive investor. 181 Gilo has emphasized that the competitive effects of such passive investment by a firm's controller (be it a dominant shareholder or a manager 182 ) are more serious and concerning when the and solely owned and controlled (100% ownership and complete control over own firm). As explained in section II.A and in what follows, the analysis of common ownership is different. First, common ownership only exists if the ownership stake held in at least one of the rival firms is partial ("concentrated" common ownership). Yet, in cases of "concentrated" common ownership as defined in this article, the common owner is able to adjust its stake in its own firm it controls (i.e., transform its position of "sole owner"-"sole controller" into one of "partial owner"-"sole controller") rather than in the target firm that it is assumed not to control. This means that concentrated common ownership may in effect entail partial ownership of both commonly held firms, at the option of the single common owner. As such, it is the reverse scenario to O'Brien and Salop's above partial cross-ownership example where the common owner may only adjust its stake in the target firm it controls but partially owns whereas ownership and control over the own firm is complete by assumption (see also n 180 below). Second, in "diffuse" common ownership cases the ownership stakes are by definition partial in both (or all) commonly held rival firms. Hence, the underlying incentives are very different.  , 1932]."). In the case of common institutional ownership, the tunneling effect is driven by the relatively asymmetric partial ownership incentives, relative control is the enabling/ enforcement mechanism due to the absence of any large controlling shareholder(s) and the fragmentation of the retail share of passive shareholders. 180 This scenario corresponds to the "concentrated" variety of common ownership with a single common owner having parallel interests in two rival firms, with no control over the one (passive investment) and total control over the other (either full control due to 100% ownership as in a full merger case, or total control, usually >50% ownership and majority control). As explained in section II.A, common institutional ownership is different because there are several common owners (typically minority investors with <50% ownership), possibly with de facto control, in all rival firms. As such, this latter scenario may fit the paradigm of "diffuse" common ownership. 181 This is because the control effect is isolated and focused on one firm over which the investor has clear control ability while the cash flow rights in the passively invested firm remain constant. 182 Gilo (n 67) 6: "Firms can replicate this anticompetitive effect by including components in their executive compensation packages that are positively linked to industry or competitors' profitability. Such compensation arrangements are analogous to the case in which a controller of a firm holds a stake in a competing firm.". Managerial compensation contracts may be another channel through which common ownership can influence product market competition. On unilateral effects theories and evidence, see Antón and others (n 7) (noting also that it is strategic product market competition and within-industry diversification that drives their model, which does not require any communication or coordination but merely that top managers know and respond to their own incentives); controller's stake in its controlled firm is smaller (e.g., less than full ownership while still remaining controlling) 183 because of a "dilution effect": by diluting its stake in the firm it controls, the controller effectively commits to place relatively less weight on its controlled firm and thus more weight on its passive stake in the rival. 184 Therefore, assuming the controller takes its own interests into account while running the firm, it will induce the firm to maximize its own profits from its (partial) controlling interest in the firm it controls (excluding the interests of non-controlling shareholders) plus its financial interest in the rival. 185 Although this may be seen as an "agency cost" in firm governance, other shareholders may also benefit from the higher supracompetitive profits even in the case without collusion and therefore, they are not expected to oppose such behavior by the firm and its controller. 186 It is important to note that the controller, precisely because of being in a position of sole control, may "selfmanipulate" 187 its ownership stake in the controlled firm to the level of its choice considering the profit maximization calculus that is most beneficial to itself (rather than the company as a whole). Thus, by diluting its stake and altering its own incentives the controller may directly 183 Gilo (n 67) 4 and 23. 184 The key characteristic of this dilution effect is that it can disproportionately affect the degree of profit internalization from the controller's parallel stakes in the firm it controls and its rival (common ownership). In contrast, when it is the firm itself that invests in its rival (cross-ownership) and thus the firm's controller only has an indirect stake in the rival via the own firm, "the controller's stake in the firm it controls will be irrelevant" as the dilution of its indirect stake in the rival will always be proportionate. ibid 22-23. 185 ibid 6 fn 25 and 24. 186 ibid 24-25 (noting that as this "agency cost" tends to benefit minority shareholders, "it would be difficult to claim that [the controller] is in breach of its fiduciary duty toward [the controlled firm]."). If fiduciary duties are seen: i) as negative property claims (a form of residual claim) by non-common owners vis-à-vis the positive property rights of common owners; and ii) as default rules for allocating property rights (and managing conflicts) that shareholders can consent to amend or waive with the aim to increase profitability, see Jonathan Macey, 'Fiduciary Duties as Residual Claims: Obligations to Non-Shareholder Constituencies from a Theory of the Firm Perspective' (1997) 38 Boston College Law Review 595; then there is indeed bargaining space for achieving an amendment of the firm objective in a way that is mutually beneficial for both groups of shareholders and also for the corporation (Pareto outcome). Such agreement is "self-enforcing", see LG Telser, 'A Theory of Self-Enforcing Agreements' (1980) 53 Journal of Business 27 as the division of profits is set (and internalized by shareholders as per the objective function) and thus cannot be undermined by opportunistic behavior of shareholders while managers are expected to be on board (due to compensation schemes or career concerns) since all shareholders are better off. Also, unlike (ongoing) tacit collusion in the market based on partial shareholding and "Coasian joint control" (full joint profit maximization by the managers of the linked rival firms) that could be unsustainable due to conflicting incentives and transaction costs, see O'Brien and Salop (n 65) 582, this is one-off and no direct influence or outside "enforcement" is needed since interests are aligned (so if undiversified shareholders are also better off they will agree to the new objective function and stick to that agreement, otherwise we would expect them to sell out). Further, any coordination problems among shareholders inside the firm are solved due to the presence of a single common owner that is also the dominant shareholder and can credibly implement the "agreed" objective function. Competition effects based on such (internal to the corporation) agreement are still unilateral. 187 Shapiro (n 171) 385.
influence its own profit function (and its controlled firm's objective function 188 ) and indirectly competitive outcomes. 189 What is striking, therefore, in this case is that the "controller effect" is not in fact an effect due to control over the target rival firm in which the controller is passively invested but only due to control over the "own" firm where the controller is the dominant shareholder. 190 In other words, it is not the result of active influence or coercion on the rival firm but a "selfcommitment" by the controller-common investor to compete less aggressively itself made "credible" by the control mechanism (given that the controller has sole and total control over the own firm) which may in turn reassure rivals and induce them to compete less aggressively 191 : a way of giving "hostages" (bonds) to support "exchange" (by acquiring 188 This means a shift from total firm value maximization to maximization of the controlling shareholder's profits (single common owner that is thus transformed from a "sole owner" to a "partial owner" over its own solely controlled firm). Another way to see the standard economic objective function of the firm (that all shareholders unanimously agree upon under the "Fisher Separation Theorem" in perfect competition) is that maximizing total firm value effectively safeguards against the "dilution effect" produced by a firm's controller (proportional distribution of corporate profits among shareholders). As explained, a deviation from this principle need not be constrained by corporate law fiduciary duties on the controller as both the controlled firm and its minority shareholders may come to share in the higher (firm and industry) profits produced in an oligopolistic market setting. Given that the theorem is inapplicable under conditions of imperfect competition and non-separate ownership, the realistic question of an "updated" objective function of the firm arises with common ownership. Azar suggests such a theory of the firm, which also "provides a possible microfoundation for O'Brien and Salop (2000)'s indices as the outcome of a competition for corporate control among potential managers", see Azar, 'Portfolio Diversification, Market Power, and the Theory of the Firm' (n 7) 1-2 (developing "a tractable model of firm behavior in which the objective of the firms is determined endogenously by the outcome of majority voting by their shareholders" and "firms act as if they maximized a weighted average of shareholder utilities."); on the need for revision of the objectives of the firm in light of common ownership and testing alternative theories, see also Antón and others (n 7). 189 Cf n 171 above. 190 cf O'Brien and Salop (n 65) 578-579. Importantly, Gilo's controller does not use control to force a higher price in the rival (as O'Brien and Salop's free riding example; in Gilo the investment in the rival is passive) but rather to commit not to undercut price itself that may in turn induce the rival not to undercut its price also. That is, the control mechanism here works not to make the target less aggressive but oneself as a controller of the firm one controls! The driver is the relative financial interest of the controller-common investor in the two rival firms. Of course, if the controller has control ability over more linked firms, it may choose to adjust its stake in any of the firms in a way that is privately profitable to itself. This latter case of a multi-firm controller is what O'Brien and Salop analyze with the difference that control over the focal (own) firm is full and fixed (assuming that the firm controller is a "sole owner" with 100% ownership and control); in this case, only control over the target is adjustable but this obscures the fact that the controller may manipulate its own profit calculus that is the principal driver of the anticompetitive effects. Essentially, O'Brien and Salop study the reverse scenario from Gilo: in both cases the relative profit ratio creates the distortions, but the identity of the firms is reversed (i.e., for O'Brien and Salop, the stake in the target rival firm is small(er) rather than the controller's stake in the own controlled firm is small). In other words, the type of partial ownership studied by each is different: O'Brien and Salop analyze crossownership whereas Gilo examines (one variety of) common ownership. See n 177 above. This however also leads to reversed policy prescriptions: while O'Brien and Salop suggest that increasing the stake in the rival firm may be competitively beneficial as it reduces the free-rider effect, Gilo cautions that any dilution of the controller's own stake in the controlled firm may strengthen the anticompetitive effect (because it reduces the relative weight in the profits of the controlled firm and simultaneously increases the weight it places on the rival's profits) or reversely that an increase in the controller's stake in the own firm will be pro-competitive. All these policy prescriptions aim at reducing the asymmetry between the financial interests of the partial owner in the rival firms. 191 Gilo (n 67) 5.
parallel investments in rivals). 192 While coordinated effects (tacit collusion) may (or may not) follow, 193 the effect can be purely unilateral in a static oligopoly setting (e.g., refraining from price-cutting) and all actors involved may benefit by the softening of competition (noncontrolling shareholders of own firm and the firm as a whole, rival firm and its shareholders, and potentially other competitors or the firm's managers to the extent they are partially compensated based on industry or rivals' profits as well as own firm profits). As a result, this effect and mechanism may be mutually beneficial for the linked rival firms albeit presumably to differing degrees. The competitive outcome may be more collusive (market power), but the mechanism is unilateral (individual behavior).

B. Corporate governance and transmission mechanisms
The distinction between concentrated and diffuse common ownership rationalizes the operation of different "potential mechanisms linking common ownership to anticompetitive effects" for different types of common owners from the perspective of corporate governance. 194 Professors Scott Hemphill and Marcel Kahan propose a useful taxonomy of such mechanisms differentiating between: i) "conflict" and "consensus" mechanisms; ii) "targeted" and "acrossthe-board" mechanisms; iii) "active" and "passive" mechanisms. 195 In this framework, concentrated common owners would be expected to follow "active" mechanisms given their parallel but presumably "asymmetric" ownership structures linking industrial competitors and their likely "targeted" governance and engagement strategies focusing on specific firms that could create conflicts among individual firms and their shareholders. On the other hand, diffuse common owners will typically employ "passive" mechanisms, either targeted ("selective 192 On the analytical differentiation between "credible commitments" and "credible threats", see Oliver E Williamson, 'Credible Commitments: Using Hostages to Support Exchange' (1983) 73 American Economic Review 519. In this setting, control over the rival is a credible threat (conflict) whereas control over oneself is a credible commitment (incentive structure). The shareholding acquisitions by increasing the opportunity cost of competing commit the common investor to reduce the risk of opportunistic behavior. 193  omission") or general ("across the board"), 196 that attenuate conflict and push towards consensus among different classes of shareholders (common versus non-common owners).
This is a natural corollary of the parallel and "symmetric" ownership structures and "portfolio" investment and governance strategies of diffuse common owners such as index funds with "similar" and "stable" holdings across multiple competitors over a long-term time horizon. 197 The distinction put forward in this article also produces novel insights as to the plausibility of potential conflict-or consensus-based mechanisms of anticompetitive common ownership. All suggested that any unilateral anticompetitive effects of common ownership are implausible because the rely on "firm value decreasing" strategies and so they collide with corporate law principles and fiduciary duties. 198 Coordinated effects -in which case PVM and FMV may be both realized without conflict -are plausible but remain empirically untested to date. 199 Indeed, this analysis would be correct when referring to "concentrated" common ownership in light of 196 Hemphill and Kahan (n 12)1400 ("A mechanism that we call 'selective omission' is consistent with both theory and empirical evidence [of common ownership's anticompetitive effects]. A [common owner] engaged in selective omission presses for firm actions that increase both firm value and portfolio value, while remaining silent as to actions where the two conflict. Aside from selective omission, some across-the-board mechanisms may plausibly be employed, but substantial empirical evidence of their use is currently lacking."), 1409 ("The most commonly mentioned across-the-board mechanism is the structure of executive compensation-in particular, whether managers are paid for performance and thereby encouraged to compete aggressively in order to maximize firm value."), 1438-1441 ("across-the-board passive mechanisms and selective omission, which merely involve a failure to take actions that would increase the value of a portfolio company, do not create material fiduciary-duty risks. [They] pose a lower risk of detection-their implementation requires no illicit communications or arrangements with the targeted firm-and a lower risk of sanction."). But see also Elhauge, 'The Causal Mechanisms of Horizontal Shareholding' (n 7) 39-43 (suggesting that "across-the-board" mechanisms have been empirically tested). 197 Hemphill and Kahan (n 12) 1400. 198 If so, the extent to which "common ownership leads to managerial behavior that violates the fiduciary obligation and harms competition" is an empirical question that is not adequately answered yet. O'Brien and Waehrer (n 12) 734, 765-766. 199 Rock and Rubinfeld, 'Common Ownership and Coordinated Effects' (n 20); Hemphill and Kahan (n 12). This claim is based on the premise that existing empirical literature on common ownership employs measurement tools such as the MHHI that are conceptually linked to conflict mechanisms, and therefore coordinated effects are not captured by empirical estimations of anticompetitive effects. But see Elhauge, 'The Causal Mechanisms of Horizontal Shareholding' (n 7) 35-39 (refuting Hemphill and Kahan's theoretical claim and further suggesting that i) consensus effects have been empirically proven and ii) even if more accurate measurement methods were available this would not negate the empirical results of studies based on the MHHI but that those would predict prices even better).
the "asymmetric" internalization mechanisms and incentives of common owners illustrated in the previous section and the "active" transmission mechanisms discussed in this section.
The critical insight arising from this article, however, is that anticompetitive strategies of If, for instance, diffuse common owners have symmetric stakes in (almost) all rivals in oligopoly, then "market-wide" coordination may be possible. If, however, the "industry maverick" is not commonly held by the diffuse common owners, then collusion may not be sustainable, 201 yet unilateral effects may still be plausible. As described in the previous section, in this scenario consensus may be forged by the congruent preferences of PVM and FVM shareholders of commonly owned firms that is built inside the firm: common and non-common shareholders agree on the "altered" objective function of the firm (PVM) as it ultimately operates to the benefit of both. 202 In this setting, the anticompetitive harm may flow from common owners merely failing to adopt "firm value increasing" strategies in particular 200 Elhauge, 'The Causal Mechanisms of Horizontal Shareholding' (n 7) 35-36: "It is always the case that all firms in all markets (and thus all shareholders of those firms) would collectively benefit if the firms could all simultaneously lessen competition among themselves in order to increase prices and profits. But with separate ownership, economic models show that (absent agreement or successful coordination between the firms) each firm has individual incentives to undercut such noncompetitive pricing, and thus they will compete even though they collectively would be better off if they all competed less. The higher the relative influence of the horizontal shareholders, the more those firm incentives to compete are lowered, because competition reduces the horizontal shareholders' profits in rival firms and thus increases the firm's effective marginal cost of taking sales from those rivals. [By contrast, l]ess concentrated non-horizontal shareholdings will thus predictably make consensus effects more likely. [...] Hemphill and Kahan wrongly assume instead that consensus effects must be based on horizontal shareholders' ability to orchestrate coordination across firms. [...] Instead, they depend on the fact that horizontal shareholding increases the costs to each firm's shareholders of competitively gaining sales, which in turn lessens the incentives of each firm's managers to compete aggressively. Because this lessens competition at both firms simultaneously, it increases profits at both firms and benefits non-horizontal shareholders as well.". 201 Gilo (n 145) 1640, 1646 (analyzing passive investment in cases of partial cross-ownership and concentrated common ownership and the likelihood for coordinated and unilateral effects in the presence of a maverick firm in the industry). The same principles apply, even more forcefully, in cases of diffuse common ownership. 202 See n 186 above and surrounding text.
firms. 203 Conflict (and agency costs) is therefore more apparent than real. In any case, individual firms also prefer less competition, but "diffuse" common ownership makes anticompetitive strategies leading to unilateral or coordinated effects plausible given the "symmetric" internalization and "passive" transmission mechanisms associated with it (i.e., the incentives and ability of diffuse common owners to act upon their PVM objectives and alter competitive outcomes on a portfolio-wide basis). 204 In other words, diffuse common ownership may partly enable firms and their shareholders to escape the "prisoner's dilemma" game in their interaction in the market and the "free-rider" problem in corporate governance, which could in turn give rise to coordinated or unilateral anticompetitive effects depending on the surrounding market conditions.
It follows from the above that diffuse common ownership based on passive governance mechanisms operates and may potentially impact product market competition in a completely novel manner. The driver of the anticompetitive effect is strategic competition (oligopolistic market interactions) and the opportunity cost created for oneself by acquiring partial or common shareholding (selective passivity)a form of self-committed profit sharing with rivals and not primarily the quality of governance and the level or cost of active engagement (active influence). 205 Thus rationally and predictably, although this strategy may entail suboptimal 203 Hemphill and Kahan (n 12) 1427: "[common owners] that are engaged in selective omission generate an anticompetitive effect because they selectively fail to push certain firm-value-increasing actions that would be procompetitive, rather than because they actively push the firm to implement firm-value-decreasing measures that are anticompetitive (as in a targeted active mechanism). Only a [common owner]'s failure to push for firm-valueincreasing procompetitive actions is a source of conflict between it and an [non-common owners].". 204 Enriques and Romano (n 89) 17-18: "These [empirical] studies suggest that at least in some instances and some markets, institutional investors might prefer a lower level of competition among firms in their portfolios because aggressive competitive behavior on the part of one of their portfolio firms would negatively affect other firms in their portfolio. [...] On the one hand, each of the firms may independently prefer a lower level of competition, in which case common ownership is merely a way to facilitate coordination. This puts a weak competition strategy in the (privately) "optimal conduct" quadrant of Table 1, that is, the preferred one in terms of both firm value maximization and portfolio value maximization. Importantly, [...] this conduct is optimal merely from the perspective of the firms' shareholders, but it is not socially optimal. [...] On the other hand, suppose that one particularly strong and innovative firm within the relevant industry would be able to maximize its own value by competing aggressively. Its PVM shareholders might still prefer a lower level of competition in order to benefit all of their portfolio companies operating in the market. If they were to prevail, the firm's conduct would be situated in the PVM-only quadrant.". Therefore, consensus and the nature of competitive effects may depend on the potential alignment of FVM and PVM objectives and strategies in light of other surrounding circumstances in each specific case. For instance, in highly dynamic or innovative industries any anticompetitive unilateral or coordinated effects of diffuse common ownership may be less likely to arise or be sustainable. Bebchuk and coauthors suggest that an agency-costs analysis of index funds shows strong incentives to "underinvest" in stewardship and also to "defer" excessively to corporate management. They conclude that institutional investors have insufficient incentives to exert influence over portfolio firms to increase firm-specific value and therefore, anticompetitive effects of common ownership through an "active influence" mechanism are implausible. management performance (some agency cost), this may be tolerable as the overall value to common owners from this ownership and institutional structure is presumably higher: 206 the effect (gain) of less competition (rents from suboptimal industry performance) may be on balance of greater significance, in qualitative and quantitative terms, than any governance and agency frictions (cost). 207 The argument is reinforced considering the portfolio-based business model of large asset managers and in particular index funds. 208 A portfolio perspective of governance together with However, it has been argued in response that: (i) "passivity" does not exclude competitive harm, and that (ii) "index funds incentives" do not prevent anticompetitive effects. See Antón and others (n 7) 27 ("it is precisely the lack of intervention when setting high-powered incentives for top managers [or 'excessively deferential treatment of managers,' as Bebchuk and Hirst (2019) call it] that leads to less competitive product market behavior. In other words, there is no paradox between favoring more effective engagement by institutional investors and being concerned about the anticompetitive effects of common ownership. Weak governance and weak competition are jointly optimal for common owners."); Elhauge, 'The Causal Mechanisms of Horizontal Shareholding' (n 7) 49-72 ("what matters is relative shareholder influence (the incremental effect of common owners relative to other shareholders), not whether shareholder effort is fully optimal" [compared to the incentives of a sole 100% owner]). Indeed, along this line of argument and as this article points out, Bebchuk et al.'s analysis would be the right benchmark for the case of "concentrated" common ownership but not useful or an appropriate benchmark for analyzing "diffuse" common ownership that is not primarily driven by (sole) control. See n 72 and 92 above and surrounding text. 206 Similarly, the "separation of ownership and control" in large corporations, albeit it creates positive monitoring costs and conflicts between principals-owners and agents-managers, is not inefficient if these are offset by other organizational benefits. Thus it is a rational choice of incorporating owners that opt to delegate decision-making power. What matters for the owners-principals and residual claimants of corporate profits is the overall efficiency of this organizational scheme being superior to others (e.g., partnership, sole proprietorship etc. : "AEGS point out that blunter incentives both reduce incentives to cut cost and to compete, and from the common owner's standpoint, the gain from less competition may exceed the cost from less effort-especially as the underlying product market becomes more concentrated. Thus, even a relatively passive common owner [...] would produce the common ownership effect of less competition in product markets." See also Mark J Roe, 'From Antitrust to Corporation Governance? The Corporation and the Law: 1959-1994' in Carl Kaysen (ed), The American Corporation Today (Oxford University Press 1996) 121-122, 125 and passim, who sheds (historical) light on the tradeoff between product market competition and managerial slack -"private profits of oligopoly" versus "private (and public) costs of poorly organized firms". Shareholders (and other corporate actors) were to benefit from the oligopolistic rents that outweighed any increased agency costs due to lesser competition. At the same time, oligopoly was said to be a source of managerial underperformance in a double sense: i) suboptimal competition induced less management effort in concentrated markets, ii) suboptimal operation of the market for corporate control disciplined underperforming management less (or not at all). 208 Elhauge, 'The Causal Mechanisms of Horizontal Shareholding' (n 7) 50-58. Elhauge forcefully argues that for diffuse common owners, namely index funds, (1) the incremental costs of lessening competition are generally zero or negative, and that (2) even when effort costs are positive, they are small relative to the anticompetitive gains. First, as Bebchuk et al. (n 205) note, index funds are typically involved in "standard" (mandatory by law) governance activities such as voting that "do not involve additional cost" or influence effort ("mandatory governance"). Second, index funds may spread any such costs across all their portfolio companies and "across a long time horizon" e.g., by applying any decision on how to vote on executive compensation across all commonly owned corporations ("portfolio-based governance" and "across-the-board mechanism"). Third, although urging firms to increase their individual corporate value and become more competitive compared to other rival firms by engaging in "firm-specific stewardship" is likely to entail positive effort cost and conflicts with the incentives of a focus on lessened oligopolistic competition changes the analysis in two important respects.
First, concentrating only on individual firm profits and governance activities (or effort costs) misses out on the portfolio-wide stewardship initiatives of index funds that may rationalize their likely interest and ability to pursue higher anticompetitive gains in oligopolistic industries (e.g., by means of "across-the board" mechanisms such as executive compensation or by "selective" interventions or engagements with management and due to economies of scale) compared to their aggregate governance and engagement costs across all of their commonly owned rival firms. 209 Second, such narrow analysis deflects attention from the critical point that pushing firms to compete less across the board, rather than more on an individual basis, in an industry may involve less or negative effort costs, rather than positive ones (e.g., by voting for executive compensation contracts tied to rival or industry performance that induce reduced managerial effort across the board, or by selectively failing to promote procompetitive, costreducing yet rivalry inducing strategies in individual firms). 210 In this light and given the portfolio-based model of diffuse common owners, less competition and suboptimal governance are not inconsistent claims conceptually. 211 The "portfolio-wide" perspective of governance highlights the new risk to competition posed by diffuse common ownership, 212 while the "firmspecific" perspective illustrates the traditional principal-agent problem in corporate governance. 213 The upshot of this "dual" governance perspective is that the presence of any index funds that are interested in the profits of all rival firms in their portfolio, by engaging in governance activities on a "portfolio-wide basis", index funds may not only save any costs of effort and influence ("economies of scale") but more importantly, may induce rival firms and their management to increase the total portfolio value of diffuse common owners by lessening competition between them (interest in total portfolio and industry profits). Fourth, the total increase in corporate value that index funds may induce for their entire portfolio of companies, instead of "only by doing a time consuming individuated analysis of each portfolio company", and the portion they may reap through fees they charge and are compounded annually, is likely "massive" compared to any costs. 209 ibid 45-46 ("[t]he fact that the Big Three have powerful incentives to influence corporate conduct does not mean they have any incentive to inefficiently expend unnecessary costs to do so. [...] Such efficiency does not show a lack of influence."); Coates (n 25) 2 ("conventional [agency-cost] analyses [of governance] mistakenly assume that index funds must make significant expenditures to influence companies and neglect economies of scale in exercise of power. They also neglect the power of control threats to discipline [corporate management]."), and 15-17 (discussing the particular channels through which index funds engage in governance activities in a way that minimizes cost and maximizes influence, e.g., by forming general policies or selective engagements on governance issues). 210 Elhauge, 'The Causal Mechanisms of Horizontal Shareholding' (n 7) 54 and 57 ("[the agency-cost analysis] is likely true for efforts to encourage procompetitive cost reductions at a specific firm, but it is not true for figuring out a general strategy for voting or setting executive compensation across all the firms in a way that increases portfolio value by lessening competition."). (residual) managerial agency costs ("managerial entrenchment") in cases of diffuse common ownership does not fundamentally change the above exposition or fully eliminate the anticompetitive risk. 214 The key insight from the preceding analysis is that the appropriate benchmark here is not perfect competition and a "no agency cost" -"sole owner" (100%) paradigm, 215 in which case by definition competitive harm is impossible. Rather, under imperfect competition and partial overlapping ownership, the "profit sharing" force drawing together the linked firms and their shareholders (rivals' profit internalization) may be stronger than and dominant regardless of any "cost sharing" due to ownership dilution (<100%) and partial control (free riding on partial owner's good governance efforts and managerial agency costs). With less unpredictable value changes (profits) 216 given the internalization caused by the common ownership links in oligopoly, the partial common owners may be safe in the knowledge that not only they need not exert the same effort competing (less aggressive competition) but also engage in the governance of particular firms (suboptimal shareholder governance) as what matters most is the supracompetitive industry profits and total portfolio profits than any firm-specific costs or gains.
This model may fit well index investment funds (with a minimum cost governance model) and diversified shareholders across firms who rationally diversify their stock portfolios 217 ("passive" diffuse common ownership). Notably, in this case both the ownership and the control of diffuse common owners in all the linked firms is "partial", albeit (to some degree) common (i.e., minority common ownership and minority common control). 218 The driving anticompetitive motive is external to the firm (partial parallel ownership of multiple firms), hence an analytical focus on entity-centric 219 , action-based and shareholder concentration perspectives ("active" concentrated common ownership) may be distracting, if not misleading. 220 On the other hand, it is quite interesting to reflect upon the (rather idiosyncratic) nature of "common control" in the case of diffuse common ownership based on diversification.
Unlike full mergers, common control in this case is: i) Partial rather than complete control (assuming common owners have some control over the partially held firms absent other more prominent shareholders in their governance structure).
This partial control of common diversified investors and index funds is often modeled based on a "proportional control" baseline assumptioncontrol being equal to the equity share. 221 The criticism against the proportional control assumption underlying (empirical and theoretical) economic research is not wholly justified. This is for a number of reasons. Given the "power vacuum" that institutional investors and large index funds in particular come to fill by replacing atomistic, retail shareholders (meaning that there is no real antagonistic force in firm governance by other shareholders with larger shareholdings and more influence that can press forward and implement their preferences), and given the fact that control has to lie with 218 See section II.A above. 219 A narrow control-oriented competition analysis of common ownership may be misleading in two particular respects. For instance, control is not important when i) the anticompetitive mechanism relies on "pure passivity" arising from the diversification of investment and diffusion of ownership in that the effects transcend firm boundaries and structure; ii) manifestation of the competitive harm (partly) relies on "committed managers" that internalize the common owners' objectives in which case common shareholders' concentration is immaterial. Accordingly, competition policy solutions taking an entity-centric view or focusing on common owners' concentration may not be wholly effective. See Hemphill and Kahan (n 12) 1452 ("fragmentation [break up] could lead to fewer anticompetitive results. However, this benefit does not arise if CCOs employ a passive across-theboard mechanism or if managers, of their own accord, decide to compete less aggressively to further the interests of their shareholders. [...] combining two CCOs into a larger one, or splitting a CCO in two, has no impact on anticompetitive effects achieved through pure passivity."). It is also for this reason that traditional structural indices (HHI) that rely on the nominal number of firms in an industry to measure market concentration (singlefirm control) do not capture well the effects generated by common ownership across firms. some shareholder representative (meaning that not all shareholders can be passive in all firms at the same time) and given that institutional investors are the most likely candidate to exert control or influence in the context of large, public, widely held firms (assuming they possess disproportionate governance power than implied by their seemingly small common financial holdings), then the "proportional control" assumption may be simply understood as a lower bound for potential anticompetitive effects. 222 That is, diffuse common owners may be assumed to have relative control, for instance by being the largest shareholder(s), among other "more passive" shareholders. 223 222 See also n 72 and 242. That said, such "proportional control" assumption is only a starting basis for economic analysis; facts in the specific case may suggest the presence of large, undiversified blockholders in many commonly held firms (in which case common owners may effectively have "zero control") or reversely, the presence of other asymmetric shareholder dynamics (indicating de facto disproportionate control of common owners relative to other shareholders nearing that of "total control" in the limit, as Banzhaf indices suggest when control approaches 50% majority ownership) or to a similar effect, the presence of asymmetric governance structures (e.g., dual-class shares, non-voting stock, contractual arrangements providing disproportionate control or decision-making rights). Accordingly, in such circumstances the "proportional control" assumption could and should be revised downwards or upwards to reflect the reality of the specific case and context. Such updated control assumptions and resulting competitive harm estimations may bring the effects analysis closer to the actual or likely effect. In light of the above however, until we have a better understanding of the (ambiguous) partial common control implications of common ownership and given the "one-share-one-vote" corporate governance principle, proportional control in case of common owners with parallel, symmetric interests and no asymmetric counterweight in governance is a reasonable analytical assumption. On the other hand, assertions that corporate voting relies on majority rule and thus the outcome is not a function of proportional control weights is inapposite and incorrect because it takes an ex post view. See claim by the merging parties in Case M.7932 Dow/DuPont, Commission decision of 27 March 2017, Annex 5, para 78. From an ex ante perspective, however, it may be reasonable that managers care about and seek to maximize the expected vote share or likelihood of gathering majority shareholder support and remaining in office, in which case they will take into account the relative minority power of common owners and other shareholders (assuming none has straight majority control) in proportion to their shares (again assuming no asymmetric governance structures or contracts among shareholders granting disproportional control). Said differently, although the concrete voting outcome may and will usually change, the chances of securing a majority outcome weighted by the relative power of each shareholder is what matters from a manager's point of view and what will shape its incentives and behavior. In an extreme scenario, shareholders may choose not even to exercise their right to vote, but the probability and threat of doing so may in itself discipline firm management and its strategic choices. See Einer Elhauge, 'The Growing Problem of Horizontal Shareholding' (2017) Index Funds -A New Antitrust Frontier? CPI Antitrust Chronicle 4 (noting that "the voting of horizontal shareholders is likely to influence managers" in two fashions: i) "if managers maximize their expected vote share, shareholders will be weighted proportionally to their voting shares"; ii) "if managers maximize their probability of re-election, shareholders will be weighted by the odds that the particular shareholder's vote will be pivotal, which gives extra weight to the largest shareholders, who typically are now horizontal shareholders"); Azar, 'Portfolio Diversification, Market Power, and the Theory of the Firm' (n 7) (developing voting models of firm behavior in oligopoly whereby managers take common shareholding into account So long as each share has an equal chance of participating in a winning coalition, the gains from monitoring will be apportioned so as to preserve appropriate incentives at the margin." The critical point about common ownership is that although "effective control" by common, diversified shareholders may be beneficial for all shareholders assuming they discipline management and minimize agency ownership situations). Accordingly, in attempting to estimate the competitive effects of diffuse common ownership, the analyst must by necessity examine the facts of the case that will also inform the plausibility and reasonableness of the control assumptions.
iii) Shared between the common owners-shareholders and other groups with (partially) heterogeneous goals such as undiversified shareholders or management of the commonly held firms ("joint control"). 229  We have seen how dependent a controlling minority may be upon the cooperation of the management and how a controlling management may have to accede in a measure to the demands of a strong minority in order to maintain its measure of control. It is not unusual for two or more strong minority interests to enter into a working arrangement by which they jointly maintain control; or a minority and a management may combine as 'the' control. In such cases we may say that control is divided and can refer to the situation as 'joint control.' [...] Cases falling between 20 and 5 per cent were usually classed as joint minority-management control." 230 Williamson, The Economic Institutions of Capitalism (n 225) 145: "to observe that ownership and management are separated does not establish that ownership is thereafter wholly lacking in control.
[...] The absence of continuous (hands-on) control permits those to whom decision powers are delegated to exercise discretion. But a total absence of control is not thereby implied. To the contrary, if ownership control is reasserted when performance approaches or falls below threshold standards, then the relevant questions are ones of thresholds and competence to intervene." As noted (n 206) above, this "separation of ownership and control" is not inefficient and may be rationalized and thus, there is no absolute loss of control by shareholders-principals. 231 Azar, 'The Common Ownership Trilemma' (n 51) 286-293; Azar and Tzanaki (n 10) 38 (noting factors such as managerial entrenchment and shareholder concentration or dispersion that may determine to what extent the objectives of shareholders and of managers may influence the firm objective function under common ownership); Backus, Conlon and Sinkinson, 'Common Ownership and Competition in the Ready-to-Eat Cereal Industry' (n 9) (suggesting that their empirical tests reject the 'exact' common ownership hypothesis (full internalization of common owners' incentives by managers) but not more moderate versions whereby only up to 30% of common owners' profit weights are transmitted to managers; thus proving some empirical confirmation as to the existence and importance of managerial agency costs (partial internalization of common ownership incentives)).
such diversified and indifferent shareholders 232 (an idiosyncratic kind of agency cost since strictly speaking a subset of owners is favored, yet again this selective attention and preference may be to the corporate entity's and all of its shareholders' benefit) 233 , the anticompetitive effect of common ownership may be robust at any level of within firm shareholder concentration (as long as there is no large, asymmetric and undiversified blockholder) and across firm diversification (parallel ownership in competitors). That is, the size of common shareholding is irrelevant as long as common owners have greater relative influence in the governance of the linked rival firms. In the limit, in fact, as we reach "perfect symmetry" in common shareholdings, the effect of relative investor concentration becomes irrelevant. 234 In such case given that the ratio of common shareholding participations is equal, the weight the manager of a commonly held firm puts on the profit of another linked rival firm is one (full internalization). Analogously to the scenario of a 50/50 joint venture structure, as ownership interests become equal and symmetric (identical financial interests), control although nominally partial (50/50) materially becomes complete (full joint control) and irrelevant (as it is a secondary consideration to the main driver being the symmetric ownership structure). Put 232 Hemphill and Kahan (n 12) fn 168 ("To the extent that managers indeed seek to further the interests of their shareholders of their own accord, as opposed to being induced as a matter of self-interest, it is unclear if anything can be done to reduce the anticompetitive effects of common ownership. As long as managers believe that their ultimate beneficial owners hold broadly diversified portfolios, they will understand that these owners benefit from less aggressive competition and act to confer that benefit. On this view, it does not matter whether common ownership is concentrated. [...] Nor does it matter whether the common owner is a financial intermediary."). Nevertheless, the systemic, long-term governance role of institutional investors as common owners may make the prospect of "committed" managers all the more likely. 233 On controllers' agency costs and fiduciary duties that are unlikely to act as an effective constraint against competition effects (controllers being conceived as either controlling shareholders or corporate managers), see n 186 above and further Elhauge, 'How Horizontal Shareholding Harms Our Economy -And Why Antitrust Law Can Fix It' (n 85) 45: "this argument [that managers' fiduciary duties prevent anticompetitive effects of common ownership] logically conflicts with well-established antitrust law deeming anticompetitive concerns to arise when one firm acquires a controlling interest of less than 100% in a competitor. If this argument were right, such acquisitions would raise no anticompetitive concerns because fiduciary duties to the noncontrolling nonhorizontal shareholders of the competitor would prevent the acquirer from ever using their control to lessen competition. The reality that antitrust law takes the opposite position means that it necessarily rejects the claim that fiduciary duties to the non-horizontal shareholders suffice to prevent anticompetitive effects."; Elhauge, 'The Growing Problem of Horizontal Shareholding' (n 222) 6 (stressing that given that "the net effect of horizontal shareholding is to increase the profits of all the affected firms" it is questionable how non-horizontal shareholders "could show injury from any claimed fiduciary duty violation"; and also that "the operational decisions affected by horizontal shareholding are protected from fiduciary duty claims by the business judgment rule." Since fiduciary duties do not necessarily bar anticompetitive effects in case of "concentrated" common ownership as Elhauge notes above, the argument is even stronger for "diffuse" common ownership. 234 Backus, Conlon and Sinkinson, 'Common Ownership in America' (n 16) 17 ("it is the increase in overlapping ownership, driven by indexing behavior, that explains the lion's share of the rise of common ownership."), 18 ("Holding all else equal, as firm f's own investors become more concentrated we expect them to put less weight on other firms' profits. But a general rise in IHHI [relative investor concentration] will appear in both the numerator and the denominator, so the effect is ambiguous. So, though IHHI has been rising since 1980, relative investor concentration cannot be rising for all pairs of firms simultaneously, and therefore rising investor concentration cannot fully explain the rise over time in κ [profit weights].").
differently, there is a sharp disconnect in the link between corporate control and competition harm as mediated by the level of shareholding (50% equally shared ownership implies complete control and leads to the same effect as a full merger). Similarly, in case of diffuse common ownership, no matter how low the level of the relative ownership interests and of the relative shareholder control in the rival firms so long as they are symmetric (identical interests) and equally shared (proportional, joint control), the anticompetitive mechanism operates just the same and the harmful effect on competition may be functionally equivalent to that of a complete merger (full rivals' profit internalization). 235 It is notable in this regard that the unilateral theory of harm based on a pure change in incentives (without formal control) was introduced by U.S. case law enforcing merger control rules on the basis of a case where the overlapping ownership structure resembled the above assumptions and fact pattern (JV-like structure with 50% parallel, symmetric interests in the two major competitors in the industry, and "committed" managers to the de facto jointly controlling common shareholder-owner). 236 Thus, while diffuse common ownership does not strictly rely on joint control as in a merger or joint venture scenario, a full internalization of rivals' profits will effectively have the same effect (weight equal to one on the linked rival firms' profits). It is for this reason as I have noted elsewhere that we need to shift focus as regards the corporate ownership "atom" and rather than being distracted by visible and "solid particles" (control rights) to also start observing more fluid and "invisible waves" (parallel interests): 237 we may be used to attend the former 235 Assuming always "committed" managers (no managerial agency costs) and all else being equal. See also Schmalz (n 51) 420: "Rotemberg's (1984) benchmark result is that when identical shareholders are fully diversified, that is, when they hold equal fractions of shares in all (symmetric) firms, firms' incentives to compete in the product market are annihilated, with the result of output falling to the monopoly level. Whereas he refers to this outcome as "collusive," he points out that in contrast to the conventional use of the term, diversification takes away incentives to deviate from the monopolistic outcome, and therefore no punishment strategies or communication are necessary to sustain this strategy. Law' (n 56) 28. As the "double slit experiment" in physics shows, waves may be invisible to the observer but their effect (or "presence") -in the form of an "interference pattern"can be. (The experiment demonstrates the wave-particle duality, which states that all matter exhibits both wave and particle properties: the particle is measured as a single pulse at a single position, while the wave describes the probability of "absorbing" the particle at a specific place on the screen. In addition, the very act of "observing" makes the interference pattern disappear given their conspicuous presence (or absence) and our familiarity with property notions of the firm and its shares but reality is forcing us to redirect attention outside the firm to grasp the effect of less familiar and less directly observable ownership phenomena with significant implications for competition outcomes. 238   (by causing waves to behave as particles), which creates a "measurement problem" for quantum mechanics. See: https://en.wikipedia.org/wiki/Double-slit_experiment, and https://plus.maths.org/content/physics-minute-doubleslit-experiment-0.) With this colorful metaphor in mind, empirical economic research should be directed to the estimation of such indirectly apprehensible "probabilistic" outcomes of common ownership incentives by testing alternative theories on the objective function of the firm and control assumptions. In this connection, the famed U.S. Supreme Court phrase "I know it when I see it" (found in Justice Potter Stewart's concurring opinion in Jacobellis v. Ohio, 378 U.S. 184, 197 (1964), the fuller version being: "I shall not attempt further to define [what may be undefinable]. But I know it when I see it") may be of particular relevance. That is, we may be far from a full economic let alone legal definition of common ownership, yet the first step is trying to understand the different dimensions of the problem and develop ways to approach and measure them with the aim to arrive at a more comprehensive theory and definition of the issue in the future on the basis of that knowledge. 238 It is instructive and notable in this regard that in case of diffuse common ownership, as said, the nominal number of firms present in the market (market concentration) is neither the (sole) source nor a (reliable) predictor of the effect (the effect crosses the firms and makes them irrelevant as a unit of analysis). That is, common ownership dilutes the very concept and analytical foundation of the firm as a stand-alone, well-defined entity in economic terms, hence the need to revisit its objective function and inject "realism in motivation" as Williamson has put it. See Williamson, Markets and Hierarchies, Analysis and Antitrust Implications (n 224) 252.

Corporate control by shareholders vs managers for varying levels of integration
On the basis of the above exposition, it is important that common ownership works through institutional intermediaries for additional reasons: i) for the dilution effect to manifest it is key that the firm's controller directly invests in the competing firms (rather via the firms) because only then there is the disproportional internalization that can be "manipulated with" (strategic motivation) 239 , ii) institutional investors and in particular the "Big Three" are not just any intermediary but they have the (de facto structural) power, scale and clout to credibly execute 239 Gilo (n 67) 37-38: "Without acknowledging this strategic motivation, one might claim that although passive investment may have an incidental anticompetitive effect, it is motivated solely by investment considerations, and not by anticompetitive ones. It is plausible to claim that the acquisition is thereby deemed 'solely for investment' and is eligible for the exemption. However, once we acknowledge the strategic anticompetitive motivation behind passive investment (i.e., inducing competitors to compete less vigorously themselves), it will be easier for a plaintiff to claim that the acquisition is not solely for investment [...] therefore outside the scope of the exemption." any of their governance threats and discipline management when necessary 240 (selective intervention or omission) 241 ; iii) the relative influence of common institutional owners is in fact disproportionate compared to any other "passive" individual investor with proportional voting rights ("more equal among equals") due to its systemic, institutionalized and informal nature (de facto minority control). 242 240 Bebchuk and Hirst (n 25) (documenting the dramatic growth of the Big Three index funds; that each of them now manages 5% or more of the shares in a vast number of public companies and they collectively cast an average of about 25% of the votes at S&P 500 companies; while as these trends continue, the future "Giant Three" are expected to be casting as much as 40% of the votes within two decades); Fichtner, Heemskerk and Garcia-Bernardo (n 25) (showing the massive shift towards passive index investment funds that are the largest shareholder in 40% of all listed companies and 88% of S&P 500 firms; that they are "permanent owners" that cannot use "exit" strategies but also not "passive owners" as they pursue a centralized voting and governance strategy ["voice"]; they occupy a position of "structural prominence" in the market for corporate control; they have "disciplinary power" over management; and they possess other avenues of "hidden" power such as private engagements with management indirectly inducing firms to internalize the [portfolio or systemic] objectives of the Big Three); Daniel Haberly and Dariusz Wojcik, 'Earth Incorporated: Centralization and Variegation in the Global Company Network' <https://papers.ssrn.com/abstract=2699326> (suggesting that a very small group of passive funds have come to comprise a "de facto permanent governing board" for a growing share of major global companies); Ian R Appel, Todd A Gormley and Donald B Keim, 'Passive Investors, Not Passive Owners' (2016) 121(1) Journal of Financial Economics 111 (finding that passive investment funds are not really "passive owners" as they have an "influential voice" due to their large voting blocs in decisions regarding firms' governance structures, resulting in more independent directors, removal of takeover defenses, more equal voting rights and less dual-class share structures; and also finding that ownership and engagement by passive funds leads to less activism in portfolio companies); Goshen and Levit (n 2) 14 ("Even the most 'passive' of investors-index funds that mimic market portfolios such as the S&P 500-actively agitate for strong governance. "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." Also, the fact that institutional investors advocate for equal voting rights and removal of takeover defenses creates a paradox that can be fully rationalized: on one hand, they are supporters of strong governance aiming to minimize managerial entrenchment (to the benefit of all shareholders), on the other hand, given their large share ownership size and Indeed, index funds with a highly diversified and wide portfolio of companies (number of links), relatively large shareholdings in particular firms compared to small, individual investors (level of links) and relatively symmetric stakes across the leading competing firms in an industry (symmetry of links) are the best fitting candidate for the theory of anticompetitive harm and strategies set out in this and the previous section. In this case, the "network of links" and the "degree of internalization" of rivals' profits is likely to be both wide-spread and significant, indicating sizeable and appreciable "common ownership incentives". Indeed, the "long-term" investment horizon of index funds makes strategies to act on and benefit from these anticompetitive incentives credible. 243

Parameters
In addition, although index funds have a "passive investment" business model and a "low cost, one-size-fits-all approach to governance" 244 compared to other investors, they are not "silent" or completely "passive owners". 245 They do have a duty to vote their shares, 246 they engage with their portfolio companies even if to a potentially lesser (or less informed) degree than structural power, advocating for equal voting rights is predictably and de facto to their benefit, so they are to call the shots in governance decisions (based on a portfolio-wide view). 243 Hemphill and Kahan (n 12) 1445 ("because of their longer investment horizon, [index funds] may be better equipped to execute across-the-board strategies, such as disfavoring relative performance incentives and supporting management against activists who advocate more aggressive competition."); Patrick Jahnke, 'Ownership Concentration and Institutional Investors' Governance through Voice and Exit' (2019) 21 Business and Politics 327, 347 (suggesting that global asset managers have the potential to act as global standard setters (or 'stewards of the commons' as Serafeim has put it) and that "with their long time horizons and common ownership [index funds] are able to provide the 'commitment mechanism' necessary to ensure that companies work together to internalize externalities created within each industry."). 244 Dorothy Shapiro Lund, 'The Case Against Passive Shareholder Voting' (2018) 43 The Journal of Corporation Law 493. 245 Note the important distinction between "added-cost" and "minimum-cost" stewardship activities of institutional investors, drawn by Bebchuk, Cohen and Hirst (n 205) 95-96 ("Stewardship decisions can be split into two parts: 1) spending decisions regarding how much to expend on stewardship; and 2) qualitative decisions regarding which way to vote or which positions to take in communications with corporate managers and other shareholders. [...] In many cases, stewardship decisions may be merely qualitative, and not involve additional cost [e.g. voting]"). Thus, although index funds may not be interested in engaging in costly stewardship activities such as initiating proxy fights (both because of the additional cost they entail in general and because of the private indirect cost that index funds may bear by opposing corporate management given their interest in attracting 401(k) business; see ibid 102), they will regularly vote or undertake the minimum stewardship activities required by law. 246 Coffee (n 89) 32 (noting that investment advisors in their capacity as "fiduciaries [are required] to vote the shares held by their fund, on the theory that voting rights are an asset belonging to the fund and cannot be wasted. U.S. agencies recognize that voting has low costs and that fiduciaries must constantly make these decisions across their portfolios.").
"active" or "activist" institutional investors, 247 and their voting matters. 248 Indeed, their large size (the overall size of their investment portfolios combined with the size of individual shareholdings in portfolio firms) and other characteristics suggest that they have relatively strong incentives to be "engaged" shareholders as they stand to gain considerably from firm value improvements. 249 Also, they are likely to have greater ability to effectively engage and affect firm policy as they have relatively greater influence than other shareholders within large firms with a dispersed shareholder base (absent large blockholders). 250 Indeed, these incentives and ability to effect anticompetitive outcomes will be multiplied and : i) because of the competition faced by mutual fund sponsors, passive fund sponsors need to exercise their governance rights in an informed manner to promote firm value and they must do this by relying on voice, rather than exit; ii) highlighting the structural advantages of passive with respect to certain types of engagement, particularly marketwide initiatives such as improving corporate governancedue to their size, breadth of portfolio and economies of scale; iii) explaining the role that passive investors can play in mediating shareholder activism). 248 Lund (n 244) 493 ("the institutional investors that dominate the passive fund market will increasingly influence and even control the outcome of shareholder interventions -from shareholder votes to those proposed by hedge fund activists") 495 ("the rise of passive investing has the potential to distort hedge fund activism. Hedge fund activists are increasingly moderated by large institutional investors with the power to block campaigns that are not in the interest of their long-term shareholders and catalyze interventions that are deemed beneficial."); Jahnke (n 243) 343 ("[there is] a concern that passive investors may have different objectives to active investors and that these differing objectives could hamper the proxy campaigns of other shareholders, especially activists. What is indisputable is that the sheer size of their combined assets means that in an increasing number of proxy battles they will cast the deciding vote."). 249  ); Jahnke (n 243) (explaining why index funds engage in corporate governance, i.e.: i) "for many institutional shareholders today, voice is more feasible than exit"; ii) "for the largest index investors, the cost of engagement has fallen to a level where it is today negligible" due to economies of scale; iii) "the immense concentration amongst index funds, with the three largest fund managers controlling over 90 percent of assets, ensures sufficient return on their governance investments"). 250 Elhauge, 'The Causal Mechanisms of Horizontal Shareholding' (n 7) 66-67 ("many factors indicate that index funds are likely to exert more effort relative to other shareholders. (a) index funds cannot exit firms, which increases their incentives to exert the effort necessary to exercise voice. [...] (b) The index fund families that vote index fund shares have much larger shareholdings than other investors, which means that the marginal gains from effort are likely to be much larger for index fund families because they have more power to influence the corporation. (c) Unlike individual investors, index funds have fiduciary duties to vote their shares knowledgeably. The law requires them to expend efforts that other shareholders may simply skip. (d) index funds can usually apply any effort to arrive at a position on common governance issues [like executive compensation methods] across many more corporations, which means that index funds will incur less effort cost per stockholding than other investors."); Coates (n 25) 2 ("conventional analyses mistakenly assume that index funds must make significant expenditures to influence companies and neglect economies of scale in exercise of power [and] the power of control threats to discipline. Index funds increasingly possess the 'median vote' in corporate contests. That gives them an ability, even if contingent, to make crucial decisions across most public companies.").
aggregate voting power within firm governance. 251 For instance, index funds usually vote together at the fund family level. 252 The combined effect of multiple "passive" common shareholders having similar stakes that are widespread across the major firms in an oligopolistic industry and voting rights representing those parallel interests, may then imply that the real impact of diffuse common ownership may be "market-wide". Notably, the competitive impact of multiple diffuse common investors with symmetric parallel stakes in rivals may be "market-wide" even in the case of pure unilateral effects without collusion. 253 Therefore, the unique challenge and new antitrust risk of diffuse common ownership is its cumulative (unilateral) anticompetitive effects that may be multiple times that of concentrated common ownership by a single common investor or even index fund family. 254 In this light, index fund common ownership may also represent a new hybrid model of firm ownership and control combining characteristics (widely dispersed ownership and concentrated voting power) from both "outsider" and "insider" systems of corporate governance. 255

IV. Implications for Theory and Competition Policy
There are several implications and conclusions to be drawn from the above analysis regarding In fact, the potential for collusion becomes less acute when unilateral incentives to compete are lessened by [common ownership] and drive markets toward monopolistic outcomes already."). 254 Elhauge, 'Horizontal Shareholding' (n 10) 1283 fn 77: "so far there seem to have been no challenges to stock acquisitions that left multiple investors with substantial horizontal shareholdings that in aggregate lessen competition." 255 See the "taxonomy of ownership and voting power" in Box 1 in Maher and Andersson (n 143) 14.

A. Theoretical implications
As regards theoretical implications, I focus on six of them here. First, it becomes obvious that the legal (absence of action) versus the economic notion of passivity (absence of effects) 256the first focusing on the behavior of the acquired firm (target) given the acquirer's active exercise of control while the latter on the acquirer's incentivesare distinct and not entirely overlapping. Unilateral effects may entirely flow from anticompetitive incentives linked to purely financial interests without any influence or control. Also, small stakes especially if held by rival firms' controllers are not necessarily innocuous; rather, the smaller the controller's stake in the firm it controls the more potentially significant the competitive concerns (concentrated common ownership). As a result, there is no straightforward relationship between the size or type of shareholding (i.e., financial, controlling, mutual) and competitive harm. 257 It is only due to path dependence and the "merger equivalent" approach used to apply to partial acquisitions that we continue to treat leniently equity interests presumably too small to convey control (in case of no anticompetitive intent). 258 Essentially, antitrust analysis of mergers and partial acquisitions is control-centric and thus may capture and address certain problematic cases of concentrated common ownership. However, it largely ignores any harm potential of diffuse common ownership based on diversification. Yet, the above analysis has shown that small or purely financial shareholdings may have significant anticompetitive effects especially when there are many such parallel links among most of the few companies operating in an oligopolistic market. Hence, sole or overreliance on active influence and control in designing merger control thresholds is not justified. 256 What I have earlier called passivity or influence in the corporate versus the antitrust sense. See n 134 above. Another way to put it is influencing corporate conduct of a rival firm is not the same as influencing competition and rivals' conduct indirectly by changing one's own incentives. 257 Gilo (n 67) 40-41: "One could theoretically put forward a technical (but incorrect) legal test that examines the degree of 'linkage' between competing firms after the passive stock acquisition. According to such a test, there would be more linkage and thus, allegedly, more anticompetitive harm, when the controller has a larger stake in the firm it controls while possessing a stake in the competing firm as well. It is clear [...] that such a test is invalid. As we have seen, the smaller the controller's stake in the firm it controls, the larger the anticompetitive harm. Second, control is a useful (and theoretically robust) but imperfect proxy for estimating competitive harm, particularly so in cases of "common minority shareholding" (diffuse common ownership). To begin, its presence signifies the lack of independence between legally separate corporate entities, which come to operate under common management or within the same business group. 259 Control is thus used to formally define the contours of a "single On the other hand, while diffuse common ownership "technically" also entails agency costs in that common shareholders presumably only take their self-interest into account in setting firm strategy (portfolio value maximization objective of the firm), this may ultimately not be to the detriment of undiversified shareholders in case they may share in the higher oligopolistic rents captured by the firms, they have invested in. 267 In such cases, shareholder interests within the firm appear to be aligned in two respects: i) as regards the firm objective function, and ii) in minimizing managerial agency costs generated due to the partial "separation of ownership and control". 268 263 See n 229-231 above and surrounding text. 264 Alchian (n 216) 339 (interpreting "ownership" as the "bearing of value consequences of resources" and "control" as the "authority to control decisions that will affect [that] value"). 265 See n 223 above. 266 See n 179 above and surrounding text. That is, expropriation of minority non-controlling (non-common) shareholders by de facto controlling (common) shareholders. 267 See n 186 and 233 above and surrounding text. 268 See n 227-231 above and surrounding text. Although the incentive to minimize managerial agency costs may be common among diversified and undiversified shareholders in principle, the degree to which this is beneficial to diffuse common owners may differ in that full elimination of such costs may not be optimal given their "passive", "portfolio-wide" competition and governance strategies as noted above.
Fourth, control over a firm's strategy may entail that the controller(s) may impose (fully or partially) her personal objectives onto the firm, or more generally what objectives are maximized by the firm and its management. 269 The controller's discretion arising from its decision-making authority may lead to a deviation from the presumably unanimous shareholder group objectives (firm value maximization). 270 Yet, in cases of diffuse common ownership, such self-interested deviation (portfolio value maximization) may be beneficial for other corporate actors and the firm as a whole so long as they may share in the supracompetitive rents, except for consumers that are worse off given the likely higher product market prices. 271 Common (controlling) and non-common (non-controlling) shareholders may thus "agree" on the altered objective function inside the firm as noted above. Remarkably, therefore, the mechanism as well as the competitive effects arising from diffuse common ownership are unilateral to begin with (softening of competition). 272 Perhaps counterintuitively, minority non-controlling shareholders (in this case atomistic, undiversified investors) may not only benefit but may actually facilitate the anticompetitive effect. 273 This is a direct corollary of the "dilution effect" produced when a controller reduces the stake in the firm it controls compared to its parallel stake in a rival, 274 in combination with the "crowding-out effect" produced by the relative growth and power of common owners Price Competition' <https://papers.ssrn.com/abstract=3784072> 10 ("the existence of minority shareholders facilitates the monopoly outcome: investors need to own a lower proportion of rivals' shares to sustain the monopoly price.") 18 ("the higher the proportion of minority shareholders with no control rights (or who are assumed not to exert their voting rights because their control is relatively negligible and face coordination problems), the lower the stakes of other firms that investors must own to maintain the monopoly outcome."). 274 See n 184 above and surrounding text, and Gilo (n 67) 6: "the controller can enhance the anticompetitive effect of such passive investment by diluting its stake in the firm it controls (e.g., by selling part of the firm's stock to public shareholders [or other minority shareholders]). [...] when it is a firm's controller that invested in the firm's competitor, even relatively small levels of passive investment can raise considerable antitrust concern." versus undiversified investors. 275 As diffuse common owners have more dispersed yet more symmetric stakes in competing firms 276 and as they grow in size and influence within corporate governance relative to other non-common investors, 277 then their relative interest (across firms) and relative influence (within firms) may dominate. While non-common shareholders seem to lose influence in the first instance (by having less or no "control weight" in a firm's objective function), their presence enables the implementation of common ownership incentives at low levels of shareholding, from which they may benefit themselves.
That is, diffuse common owners alter their own incentives and profit function by embracing "passive investment" strategies (portfolio diversification) but non-common owners make the "commitment" of diffuse common owners to potential anticompetitive strategies in oligopolistic settings (strategic motivation) "credible" even with small, passive minority shareholdings in competing firms, thus indirectly allowing the internalization of rivals' profits that leads to harmful competitive outcomes (increased market power). 278 In short, while common owners-controllers may dictate the firm objective function, the (partial) control mechanism becomes meaningful only when the relative proportions of common versus noncommon owners suggest higher portfolio gains which can be maximized and shared among the two groups of shareholders, although not necessarily pro rata (as per the typical corporate law convention). 279 Fifth, diffuse common ownership works against traditionally perceived single-firm concentrated control. The apparent division of (partial) ownership within a single firm and its (parallel) diffusion across rival firms among the same owners make the dilution of "sole  to mergers and joint ventures, however, the "integration" effected by diffuse common ownership is informal ("effective" integration). It leads to rivals' profit internalization merely due to common owners' broad investment diversification strategies within oligopolistic industries and not due to traditional "structural integration" of separate businesses into a single economic entity ("integration by hierarchy" à la Williamson Williamson's idea of "selective intervention", thought impossible in the context of traditional mergers, may be attainable by means of de facto "partial mergers" due to diversified portfolio investments by diffuse common owners across firms. 288 Sixth, public policy may face a "dilemma" rather than a "trilemma" as regards diffuse common ownership 289 : portfolio diversification in oligopolistic markets may possibly lead to both suboptimal competition and suboptimal governance outcomes (supracompetitive profits and managerial agency costs). 290 Indeed, partial common ownership and partial common control as depicted above may have implications not only for the interactions of firms and the functioning of product markets (partial internalization of rivals' profits and subcompetitive market outcomes) but also for the operational efficiency of business organizations (organizational slack and suboptimal managerial incentives). 291 In other words, it is possible that both the profit maximization as well as the cost minimization objectives of the firm are affected by the combination of the partial "shareholder overlaps" among competing portfolio firms in oligopoly, the partial "separation of ownership and control" as between common shareholders and corporate managers within firms and the "portfolio-wide" governance model of common owners that characterize diffuse common ownership. In such environment of common ownership, atomistic competition and governance initiatives impose externalities on commonly held product market rivals. It may therefore be rational for diffuse common owners to "intervene selectively" 292 or engage to a lesser degree in the governance of individual firms, which may reshape the behavior and reduce the effort levels of management in minimizing the production costs of individual firms. 293 In this regard, the dimensions and extent of the relative problem posed by common ownership for firm competitiveness and productivity, and their potential interplay, are not well understood yet. 294 288 See n 224-226 above and surrounding text. 289 Azar, 'The Common Ownership Trilemma' (n 51). 290 Antón and others (n 7) 27. See also n 215 above and surrounding text. 291 See n 206-207 and 214 above and surrounding text. 292 See n 225-226 above and surrounding text. 293 Antón and others (n 7) 28: "When firms interact strategically in the product market, from the perspective of portfolio value optimization, it may be optimal for a common owner to act like a 'lazy owner,' a behavior that is often associated with bad corporate governance. In other words, good governance-in the sense of measures that promote efficiency and shareholder returns from the perspective of an individual firm imposes an externality on product market rivals. Therefore, common owners of product market rivals may optimally provide reduced levels of governance interventions, even though they lead to lower productivity, higher costs, and reduced operating performance of any individual firm." 294 For an early attempt to empirically provide some bounds of such dimensions, see Backus, Conlon and Sinkinson, 'Common Ownership and Competition in the Ready-to-Eat Cereal Industry' (n 9).

B. Policy recommendations
In light of the above, competition policy and merger control need to adapt to the new common ownership reality if they wish to remain informed and relevant. 295 In specific terms, that means recognizing the two varieties of common ownership and taking into account their conceptual qualities, distinct supporting mechanisms and underlying assumptions when assessing competition effects and enforcing merger control law.
At the substantive level, a first step is to acknowledge the new competition concerns linked to diffuse common ownership based on an "effects-based" theory of harm (competitive influence) going beyond formalistic legal and economic constructs that lie at the foundations of merger control (e.g., control-and entity-centric models). As shown, diffuse common ownership makes some of the properties of control to lose their analytical vigor (symmetry of parallel holdings) while the mechanism that produces competitive harm, or efficiencies, is unilateral (altered firm objective function). Competition policy is thus called to embrace a novel unilateral effects theory for cases of diffuse common ownership that is flexible on the one hand but also delimited and administrable: 296 any competitive effects flow from the altered incentives to compete, due to diversification and the diffusion and (partial) collectivization of ownership it entails, combined with some informal control, due to the potential de facto aggregation of shareholder power in public, widely held firms, by diffuse common owners. Accordingly, the "passive influence" mechanisms uniquely associated with diffuse common ownership and their harm potential (i.e., strategic influence in the market interactions of oligopolistic rival firms and actual minority control by common shareholders that engage selectively or on a portfoliowide basis in the governance of their commonly held firms) need to be explicitly acknowledged. 297 The analysis in this article has illustrated that the anticompetitive mechanisms of diffuse common ownership are not only theoretically plausible but may also be potentially material under the appropriate circumstances. ownership and governance structures within which the interlinked rival firms operate, a "caseby-case" approach to the analysis of common ownership is advocated. 299 On the one hand, it is clear that the current immunity privilege (per se legality) afforded to diffuse common ownership is not justified in light of its underlying "passive" internalization and transmission mechanisms that may induce harm yet imperceivable by traditional merger control thresholds (individual shareholding and standalone control levels) and measurement tools (standard HHI and market concentration measures). 300 On the other hand, a "rule-based" approach (per se prohibition) to diffuse common ownership (e.g., limiting investor specific diversification to certain ownership levels in rival firms in an industry) is likely ineffective, unable to discriminate between cases and it may also be easily evaded (e.g., by allowing institutional investors to restructure their portfolio of parallel holdings in rivals and further spread them across many formally separate but essentially similar index or other passive funds). 301 Thus any benefits of this approach in terms of simplicity, administrability and legal certainty for business and investors are overshadowed by its likely costs. In addition, the unique potential of diffuse common ownership to generate welfare increasing efficiencies (e.g., in view of positive technological or innovation spillovers) and the potential presence of countervailing factors such as managerial entrenchment or inter-industry common ownership that may 299 Elhauge, 'Horizontal Shareholding' (n 10) 1303; Elhauge, 'How Horizontal Shareholding Harms Our Economy -And Why Antitrust Law Can Fix It' (n 85) 254-255 (providing guidelines and limiting principles on how to pursue "case-by-case" enforcement); Patel (n 12) 282-283 (supporting "a case-by-case approach that evaluates all relevant factors bearing on competitive effects [of common ownership]"); Tzanaki, 'The Common Ownership Boom -Or: How I Learned to Start Worrying and Love Antitrust' (n 38) 10 (proposing "case-by-case analysis" based on detailed enforcement guidance, together with "staggered legal change"). 300 See sections III.A and B above. 301 Posner, Scott Morton and Weyl (n 174) 678-679, 708-710 (proposing a "structural" safe harbor limiting large institutional investors (fund families) to holding no more than 1% of an oligopolistic industry; alternatively, investors are to concentrate and limit their holdings in only 1 firm per industry); 724 (suggesting that their "proposal [may prove] insufficiently aggressive for at least two reasons: 1) there is a strong interactive effect of different mutual funds all having similar holding patterns"; 2) if their policy were to induce fragmentation of the mutual fund industry into hundreds of institutions, all below the 1% threshold while holding a fully diversified portfolio, the harmful patterns could still be replicated even if the effects were somewhat mitigated); Elhauge, 'How Horizontal Shareholding Harms Our Economy -And Why Antitrust Law Can Fix It' (n 85) 257 (noting that "the competitive effects of one shareholder's horizontal stock acquisitions depend on the horizontal stock acquisitions of others", thus rightfully "the Posner-Scott Morton-Weyl proposal, although more rule-like in form, ultimately does make the legality of individual horizontal stock acquisitions turn on the existence of others."; Romano (n 8) 401: (suggesting that Elhauge opposes fixed thresholds as per PSW's proposal because they are "both over-inclusive and under-inclusive"); Hemphill and Kahan (n 12) 1401 (showing that "blunt, wide-ranging reform proposals are likely to be ineffective and counterproductive. The most probable effects of these proposals are greater shareholder passivity and fragmentation of institutional shareholdings in portfolio companies in all industries, not just in concentrated ones. The proposals would thus be ineffective if passive mechanisms are responsible for anticompetitive results, and counterproductive because they reduce shareholder power and incentives to induce portfolio companies to increase their value where doing so is not anticompetitive.") and 1452. mitigate the anticompetitive effects of within-industry common ownership are factors in favor of the "case-by-case" approach. 302 To implement such open-ended approach, competition authorities will need to develop guidelines. 303 Guidance will need to crystallize the new theory of harm relating to diffuse common ownership and elucidate the relevant circumstances under which competition harm may be likely and substantial or any antitrust related efficiencies may be credited, in line with the preceding analysis. Furthermore, antitrust authorities will need to clarify and consolidate the set of critical factors that may affect merger control enforcement and the competitive assessment of cases involving concentrated or diffuse common ownership both as regards unilateral and coordinated effects. Among such relevant factors, consideration shall be given to market or structural factors (degree of product market concentration, nature of competition, number and type of rival industrial firms with or without common ownership links) as well as to transaction specific or behavioral factors (type and characteristics of common shareholdings, number and type of common owners in proportion to other undiversified shareholders within corporate governance, relative power or autonomy of corporate managers vis-à-vis common shareholders). 304 Also, it would be useful to define appropriate theoretical indicators, e.g., reflecting the relative level of ownership and control asymmetry (concentrated common ownership), or symmetry (diffuse common ownership), that may distinguish between the two varieties of common ownership and relate them to different unilateral theories of harm. 305 These and other relevant factors affecting the common owners' measure of control (control weights) and degree of internalization of rivals' profits (profit weights), which go into 302 See n 2, 6, 8 and 168 above and surrounding text. The presence of potential efficiencies also advocates for enforcement under merger control rather than antitrust laws on cartels ( §1 Sherman Act or Article 101 TFEU) that adopt per se illegality rules and automatic nullity remedies. Besides, the type of economic analysis employed for the competition assessment of mergers is more closely fitting to that required in common ownership cases. the unilateral effects analysis of common ownership, are critical to devise and include in the guidelines. 306 Realistically, however, this guidance exercise will need time to mature and come into fruition as scholarship especially on the coordinated effects of common ownership is currently underdeveloped. 307 The substantive assessment will also need to become tailored, according to the foregoing insights about the special nature of control (i.e., majority or minority control) underpinning each variety of common ownership. As noted, concentrated common owners have disproportionate control between their commonly held firms established on a de jure standalone basis (sole control). This type of majority control accompanying partial common ownership is captured by existing merger laws. In contrast, the appropriate benchmark for antitrust analysis of diffuse common ownership is symmetric common control established on a de facto aggregated basis (joint minority control). 308 These factual control situations are not explicitly recognized as problematic or comprehensively addressed under most merger control regimes (most prominently the EU regime being the most conservative one in terms of jurisdiction). 309 Merger policy and practice need to fill this gap.
In technical terms, this further means that existing analytical tools and structural screens that are applicable during merger control review may need to be finetuned and adapted. For instance, market concentration indices (e.g., MHHI) 310 will need to be further developed to account for richer "relative" (partial) control scenarios in practice (e.g., enhanced or disproportionate control of common owners relative to other shareholders, some managerial control  311 On the "relative" nature of control (i.e., factual, partial control vis-à-vis non-common shareholders, potentially shared with corporate managers) in diffuse common ownership cases, see n 221-231 above and surrounding text. 312 On different corporate governance models of "controllers" (i.e., concentrated or diffuse common owners, or corporate managers), see Figure 5. 313 See n 25-37 above and surrounding text. that could be shown or estimated based on actual evidence (e.g., past voting patterns in annual general shareholders' meetings) and given the specific factual circumstances in each case. 314 In other words, the "control analysis" in diffuse common ownership cases needs to become less fixed and formalistic, and more facts-sensitive while any quantification indices need to be more open to alternative corporate control assumptions. That is, any factual (e.g., de facto minority control by common owners as a group, potentially de facto shared with corporate managers) or aggregated corporate control dynamics (e.g., control of the interlinked rival firms by diffuse common owners established by aggregating shareholdings, either at the fund family level for big asset managers or across all passive, similarly diversified institutional investors in a given industry when individual funds or investors may exercise voting rights in a coordinated manner) would need to be duly taken into account alongside established notions of legal control. Such factual control could be quantified by employing and testing different assumptions going beyond the standard "proportional control" assumption (e.g., assuming disproportionate control by diffuse common owners vis-à-vis undiversified shareholders, that may effectively constitute a "majority of the minority" in firm governance, estimated based on Banzhaf voting indices). 315 On the other hand, if large blockholders are present and active within corporate governance, then as illustrated, diffuse common owners may possess no control at all (although their shareholdings are not formally "silent" and have proportional voting rights). 316 It follows that the control inquiry need not be set aside altogether but rather be made realistic in order to be able to reflect and encompass the actual workings between 314 In a recent market investigation of the Hellenic Competition Commission in the construction sector involving common ownership concerns, evidence on actual voting patterns and participation levels in annual general shareholders' meetings were used to assess the de facto control ability of a common financial investor holding parallel stakes in the two major competitors in the industry. Based on such findings of factual control, MHHIs and MGUPPIs could then be calculated See First Interim Report: <https://www.epant.gr/en/enimerosi/pressreleases/item/1374-press-release-market-investigation-in-the-construction-sector.html>. 315 Azar, Schmalz and Tecu (n 1) 1545 (using Banzhaf voting power indices as an alternative to calculate the MHHI); O'Brien and Salop (n 65) 570 (noting the potential "disproportionate control" of relatively larger minority shareholders that may jointly control the outcome in line with Banzhaf theoretical analyses). Unlike "proportional control" that mechanically equates control rights to equity interests, "Banzhaf control" estimates the actual (disproportional) voting power and likely influence of common shareholders that are considered pivotal in possible voting coalitions. Thus pursuant to the Banzhaf index, as their participation percentage in the voting nears 50% (majority control), their control is thought to actually approximate total control (100%). That is, they are the ones who call the shots as they have the decisive vote despite holding nominally minority positions, given the dispersed structure of the shareholder base (e.g., in large public corporations). 316 O'Brien and Salop (n 162) 622-625 (suggesting "discounting for non-control" when estimating the competitive effects of de facto "non-controlling" minority shareholdings based on the MHHI). A similar rationale could apply to common shareholdings ("discounting" nominal shares and profit weights in firms where common shareholders have no control) that in theory come along with proportional voting rights but in practice those are canceled off by the presence of larger controlling shareholders.
actors within corporate governance, which impact upon the degree and ability of common owners to "internalize rivals' profits".
In terms of process and in line with the above insights, taking into account common ownership as an "element of context" during regular merger review of transactions between portfolio companies and also expanding reporting requirements for common institutional ownership, by requiring filings based on "aggregated shareholdings" of affiliated individual funds within the same fund family are steps into the right direction. 317 Such approaches enable us to ensure that The real challenge for competition policy and merger control enforcement is tackling the cumulative anticompetitive effects produced by diffuse common ownership. That is, regardless of the specific corporate control dynamics, diffuse common ownership raises the real prospect of "hidden" industrial market control by common institutional investors (market power) that is the ultimate concern of antitrust and merger control. 320 The challenge is unprecedented because as shown market control in this case may be achieved by unconventional means (e.g., by altering the firm objective function) meaning on the one hand that the anticompetitive other words, any control or influence over market competition due to widespread common shareholdings between rival firms is informal. This practically means that traditional structural indicators (e.g., number of firms in an industry, market shares, market concentration indices) are not fit to capture actual firm interactions and product market dynamics. 323 Therefore, a direct conclusion from the preceding analysis is that if antitrust authorities wish to incorporate merger control enforcement screens addressing diffuse common ownership cases in their guidelines, those screens or soft safe harbors will need to be aggregate (covering the potential extent and effect of all common investors combined in a given market) or "market-wide" (e.g., considering, by analogy to exclusive dealing arrangements, whether the aggregate market foreclosure effect is substantial). 324 This is in contrast to existing merger control safe harbors that are designed under a paradigm of merging into a single entity ("structural" integration) based on "single firm" dominance and "individual investor" activity. 325 Thus, "market-wide" safe harbors shall aim at capturing the aggregate industry-wide effects of diffuse common ownership ("effective" integration due to diversification 326 ) and counterbalance the (in)ability to effectively address them via screens and thresholds that are firm specific, control-centric or based on individual shareholding size. 327 In practice, this may mean introducing in the medium term residual ex post merger control enforcement in cases of 323 Case M.7932 Dow/DuPont, Commission decision of 27 March 2017, Annex 5, para 4: "concentration measures, such as market shares or the Herfindahl-Hirschman index ("HHI"), are likely to underestimate the level of concentration of the market structure and, thus, the market power of the Parties." 324 Elhauge, 'Horizontal Shareholding' (n 10) 1308-1309 (suggesting that the "collective" anticompetitive effect from all common shareholdings by multiple institutional investors in concentrated markets may and should be liable under the U.S. merger control law, and arguing that the timing of merger control enforcement could shift to ex post review for shareholding acquisitions that may have been initially legal but could be challenged later given the aggregate anticompetitive effect created by these and subsequent common shareholding acquisitionsby analogy to exclusive dealing when a series of agreements make the aggregate foreclosure share substantial);  327 For the problems associated with standalone reporting thresholds for partial ownership acquisitions, see Moss (n 39) 12 (stressing that de minimis thresholds or safe harbors for ex ante merger enforcement may create strategic incentives for investors to circumvent the spirit of the law in two ways, i.e., they induce investors: i) to fragment their partial ownership in smaller individual stakes but held in multiple rivals that in aggregate may have the same market effect; ii) to initially acquire partial ownership stakes in rivals that compete in 'closely' related markets but later encourage their post-acquisition 'repositioning' into the same market. In both cases, partial common investors have strong incentives to maximize profits, potentially leading to monopolistic or collusive outcomes). materially harmful diffuse common ownership, 328 to minimize a "major blind spot" in antitrust enforcement. 329 To the extent possible, this should be done in a predictable way, by providing some guidance to business actors so they may self-assess their own investment strategies and activity in context. In short, aggressive merger control enforcement against market-wide or aggregate effects of diffuse common ownership should be possible based on rigorous economic evidence but competition authorities need to develop clear guidelines for that purpose.
This industry-wide "structural approach" will naturally entail some antitrust risk for financial and institutional investors with parallel investments in multiple competing firms (such as index funds), even if those are "passive" from a financial investment perspective and hold "small" positions in themselves. Yet, this may be an unavoidable side effect linked to the portfoliobased business model of such investors and may be justified from a policy perspective on welfare grounds in cases shown that the anticompetitive effects may be overwhelming compared to any likely efficiencies (balancing the public versus the private interest).
Alternatively, market-wide safe harbors may be combined with or complemented by behavioral filters that may single out truly passive investors with parallel investments in multiple rivals, which are not voting their shares ("silent"). 330 The premise is that such passive investors presumably do not have any corporate (voting) control or influence over their commonly held firms to implement any anticompetitive incentives that may flow from their common ownership stakes. 331 The drawback of such behavioral safe harbors (e.g., voting prohibition for diffuse common owners) is that they may not address other channels or mechanisms of anticompetitive influence that do not operate through voting (e.g., contractual mechanisms such as executive compensation packages that align the incentives of corporate managers with those of common shareholders). Similarly, investors could commit to "purely passive" ownership by "putting their shares in the drawer" 332 , i.e., only mechanically voting their shares and not engaging in corporate governance in any other way. 333 Yet, such behavioral commitments do not necessarily or fully eliminate any antitrust risk. 334 Importantly, they may also have adverse unintended consequences on corporate governance (e.g., de facto empowering activists and other non-common investors with disproportionate voice and influence ability, increasing potential managerial agency costs). 335 In any event, behavioral screens would also need to apply across the board to all rather than a single common owner-individual fund or fund family. 336 As other diffuse common owners have similar and parallel interests, their voting and behavior is likely to be aligned according to their portfolio interests. 337 In this light, structural market-wide thresholds or solutions (e.g., limiting within-industry diversification to certain levels for all investors) are more likely to effectively address the root causes of the problem diffuse common ownership creates for competition and any side effects of behavioral solutions that fully undercut governance activity by institutional investors. 338 Relatedly, a further behavioral solution is disaggregating any 332 Romano (n 8) 381-382, 397, 402. 333 Posner, Scott Morton and Weyl (n 174) 722 (putting forward, besides their "structural" safe harbor, a "purely passive" option for common investors (index funds) to avoid antitrust liability, pursuant to which they shall commit to engage in no communication with top managers or directors, to vote their shares in proportion to existing votes so that they have no influence in any corporate governance decision, and to own and trade stocks only in accordance with clear and non-discretionary public rules (matching an index). The rationale for their proposal is that "institutional investors can affect competitive outcomes only by exercising control over operational firms, and control requires communication and voting or the ability to sell the shares of the firm", therefore, any weaker form of "passivity" (such as the one currently used under U.S. merger control) will not effectively restrict common owners from affecting corporate governance.) 334 Elhauge, 'Horizontal Shareholding' (n 10) 1305-1309, 1315 (suggesting that the antitrust risk is not eliminated "because nonvoting stock might still influence management in anticompetitive ways": e.g., if managers take into account the interests of non-voting common shareholders out of fear that otherwise they will sell to others in a corporate takeover attempt or will "not support the managers' hire at subsequent corporations"); Hemphill and Kahan (n 12) 1452 fn 168 (noting that if "committed" managers attend to the interests of their shareholders out of their own accord, and not due to self-interest, "it is unclear if anything can be done to reduce the anticompetitive effects of common ownership"); Tzanaki, 'Common Ownership and Minority Shareholding at the Intersection of Competition and Corporate Law' (n 56) (showing that "voting" and "tradable" stock even if relating to minority shareholdings is not really passive in the antitrust sense). 335 Romano (n 8) 402 (highlighting this problem which in turn entails that "the relationship between control rights and cash flows would completely break down"); Elhauge, 'Horizontal Shareholding' (n 10) 1315 (noting that "having institutional investors refrain from voting increases the separation of ownership and control in a way that harms corporate governance and efficiency on a host of issues that do not raise anticompetitive concerns"); Coates (n 25) 21 (analyzing perverse corporate governance consequences of restricting voting by index funds). 336 Posner, Scott Morton and Weyl (n 174) 724 (underscoring "the strong interactive [anticompetitive] effect of investment funds with similar holdings patterns" that their investor-specific [structural] proposal may not effectively address); Lund (n 244) 528 (noting the "first-mover disadvantage to abstaining from voting" [the least costly governance activity] and the public disrepute involved if other passive funds continue to vote, which means that "unless all passive funds collectively gave up their voting rights, it is unlikely that any one institution would voluntarily choose to do so"). 337 See n 44 and 86 above and surrounding text. 338 To be sure, diversification limits would also undermine any stewardship incentives by passive funds but to a lesser extent than a complete ban on voting or other governance practices. See Coates (n 25) 21 (suggesting that centralized voting practices by large asset managers (e.g., requiring that each fund within a fund family votes its own shares). 339 While in principle this is set to undercut the joint minority control mechanism underpinning diffuse common ownership , in practice this solution too may be circumvented by institutional investors informally or tacitly coordinating their voting and governance activities. 340 Another possible "structural" solution in diffuse common ownership cases is re-concentrating ownership for each individual investor in a single firm operating in each concentrated product market. 341 This option would allow investors to tap on the benefits of diversification by diversifying their portfolio of investments across industries while eliminating competition concerns and also restoring canonical "sole owner", "firm specific" corporate governance. 342 This solution is meant to essentially transform common ownership that is "diffuse" in nature into the "concentrated" variety. 343 It follows that the most promising and likely effective means to eradicate the competition risk of diffuse common ownership is to reshape or transform it, either by limiting within-industry diversification across the board or by re-concentrating ownership in individual firms (i.e., no diffusion of ownership and no dilution of control). 344 In view of the above insights, sensible competition policy should not lose focus from the overall market effect by overly zooming into the (single firm) governance mechanics, important as they may be. Once one decides to be a "partial" common owner in an oligopolistic market, it structural limits on index funds would negatively affect their incentives to monitor or act to improve portfolio company value and the challenge of balancing this effect against "the corporate governance benefits of increased monitoring that flow from less dispersed [common] ownership"); Romano (n 8) 402 (noting that such solutions would in practice undo decades of corporate governance reforms aiming to induce institutional investors to become marginally more engaged owners). 339 Elhauge, 'The Causal Mechanisms of Horizontal Shareholding' (n 7) 74 (noting that the diversification benefits could remain without harm for individuals investing in various index funds even at the same fund family if: i) the funds' managers were incentivized to maximize only the value of their fund; and ii) the fund family allowed each fund manager to vote separately, rather than (as presently) voting all their shares at the fund family level). 340  is also to bear the risk (rather than consumers and society) of potential anticompetitive effects and hence enforcement action arising from later strong interlocking shareholding links that collectively undermine competition in the pursuit of private self-interest (of common owners and potentially other corporate actors). 345 Such a public policy stance not only enables competitive harm to be remedied ex post in the specific case (enforcement) but most importantly, it induces private economic agents to internalize the law prohibition effect ex ante (deterrence). 346 It is therefore more likely that their investment and governance behavior will be shaped accordingly. The task of guidelines is precisely to flag theoretical and factual conditions under which enforcement action is likely to be expected (e.g., highly concentrated markets, high levels of common ownership, substantial extent and proof of aggregate anticompetitive effect). 347 Investors are likely to know for example the extent to which a given market is oligopolistic or whether there are other similarly diversified investors present in the firms in which they invest. Furthermore, lack of any antitrust enforcement against diffuse common shareholdings (e.g., by preserving their present de facto immunity status) is only to make the antitrust problem bigger, meaning that even more drastic solutions may be required at a later time to ensure operation of competitive markets. 348 It is interesting at this point to compare and contrast the new role antitrust is propelled to embrace given the distinct remedial solutions it may offer against the concerns associated with each variety of common ownership. On the one hand, diffuse common ownership and its potential aggregate anticompetitive effects call for structural solutions at the market (rather than the firm) level by competition policy and enforcement, in direct proportion and as a counterbalance to the portfolio-wide corporate governance and financial investment model of passive institutional investors. This state of affairs brought about by financial innovation and rampant market developments may mark the end of "atomistic antitrust" as we know it. On the 345 That is, the risk of changed circumstances subsequent to any individual "partial" stock acquisition is born by investors in cases of substantial cumulative impact on competition. 346 For the continuous possibility of ex post merger enforcement against partial acquisitions in the U.S., see Scott Morton and Hovenkamp (n 78) 2045, 2047 ("the competitive effects of partial stock acquisitions (in contrast to complete acquisitions that create a single firm, and the antitrust laws apply only to the 'acquisition'), including horizontal shareholding, can generally be appraised as of the time of the lawsuit. This entails that the challenge is not merely to the 'acquisition,' but also to post-acquisition performance or behavior. Section 7 enables the antitrust enforcement agencies to reach back in time and aggregate small purchases, which is critical in enforcement against institutional investors that slowly accumulate large positions over time."); Daniel A Crane, 'Antitrust Antifederalism' (2008) 96 California Law Review 1, 53 (noting that prior to the pre-merger notification system introduced by the HSR Act, the government often relied on post-merger evidence of supracompetitive pricing by the merging firms to prove a merger's adverse market effects, and that such suits still occur today, but are rare). 347  other hand, potential concerns in concentrated common ownership cases pertain not only to adverse effects for product markets but also for undiversified investors of individual commonly held firms. To the extent non-common shareholders may be harmed by the asymmetric ownership and control structures created by concentrated common ownership, and corporate law rules and principles such as fiduciary duties cannot provide adequate relief, antitrust may offer an alternative intra-firm remedy against potential conflicts of interests. As such, antitrust emerges as an instrument of "investor protection" (ancillary and parallel to consumer protection) that is called to address and rebalance the allocation of property rights among different groups of shareholders inside the firm. 349 These insights reveal that both varieties of common ownership demand a "structural approach" for the effective resolution of any concerns (i.e., internalization of common ownership externalities) precisely due to their primarily "structural" nature (altered incentives to compete due to changed ownership structures). 350 The difference is that such structural approach operates at the level of the market in case of diffuse common ownership (industry ownership and structure) whereas inside the firm in case of concentrated common ownership (corporate ownership and structure).
A broader and more challenging question remains as to what extent the design of merger control regimes and their different applicable jurisdictional thresholds should be amended to reflect common ownership concerns. Going past substantive and remedial issues, that appear uniform in principle although potentially variable in application depending on the empirical evidence in the specific case, it is a more demanding task to offer country specific recommendations on the appropriate jurisdictional scope of merger control. This requires not only a better economic understanding of the issue at hand, that ongoing empirical research and this article aim to advance, but also accounting for further complex legal and institutional considerations, all of which may differ considerably from jurisdiction to jurisdiction. For that purpose, an in-depth and tailored cost-benefit analysis will need to be conducted in order to also been reviewed some twenty years ago. 352 While these policy initiatives have not been translated (yet) into concrete action, it is critical to note that the analysis merely focused on standalone shareholdings. As shown throughout this article, the competition assessment and any cost-benefit analysis regarding the size of the problem versus solutions may be considerably different in case of diffuse common shareholdings (i.e., multiple, parallel, symmetric holdings in rivals within a given product market). Thus, findings in those previous consultations are without prejudice to any policy conclusions to be drawn as to the proper merger control design in light of the contemporary debate on the common ownership issue.
Besides, the analysis in this article has crystallized some useful points as regards jurisdiction.
To begin, it has become clear that U.S. merger control is flexible to capture any anticompetitive partial acquisitions, even if based on pure incentives and actual unilateral effects. 353 However, other merger control regimes that may flexibly and pragmatically rely on joint de facto (Germany, UK) rather than de jure control (EU) as a jurisdictional criterion could also capture diffuse common ownership under a teleological interpretation of the law. 354 On one level this means that merger control reform and any legislative change may be more minimal than often portrayed: so long as de facto control is assessed on an informal and cumulative basis, 355 merger control jurisdiction may be apt to address the changed incentives to compete associated with diffuse common ownership that may in turn lead to durable market power and competitive harm. At the same time, the current EU interpretation of a "stable coalition" or "strong common (strategic) interests" to establish jurisdiction based on de facto joint control appears to be mechanical, formalistic, and obsolete. Given the systemic and de facto identical character of diffuse institutional common owners (even if their specific identity is different), their symmetric holdings (even if individually small and dispersed across firms) and their similarly exercised voting rights (potentially forming a de facto minority voting bloc with effective firm 352 Commission Staff Working Document (n 111); White Paper (n 102); and Green Paper (n 101). 353 Elhauge, 'Horizontal Shareholding' (n 10) 1305-1309. 354 See section II.B above, and Note by Germany (n 116) 7: "Since investors could be presumed as having equal interests up to a certain degree and since there could be ties between diversified investors, [antitrust agencies could] aggregate the shares of all institutional investors involved in a company that are equally diversified within the industry, if such equal interest can be assumed. This would be similar to an actual form of minority control, in which several minority shareholders effectively control a company through joint action. From a merger control perspective, common ownership by institutional investors could be regarded as a passive form of strategic influence, where even small shares could allow investors to have a decisive impact on their portfolio companies' decisions." 355 See n 296-297 above and surrounding text.
To recap, a narrower delimitation of the "passive investment" safe harbor as well as a more flexible, fact-based interpretation of the "joint control" criterion are advocated with reference to ex ante merger control jurisdiction. At the same time, no immunity privilege to any future post-acquisition anticompetitive "conduct" is to be afforded in view of pervasive diffuse common ownership patterns. Ex post enforcement may thus be justified in this case given the breadth and magnitude of the competition effects within a given oligopolistic industry. 363 Although ex post merger scrutiny is uncommon today given the ex ante merger notification and approval system adopted by most merger control regimes, it historically preceded such a system. 364 In fact, ex post review remains an available enforcement option against partial acquisitions in certain jurisdictions (U.S.), whereby not only the "acquisition" but also "postacquisition performance" is liable under antitrust law for any harmful effects on competition. 365 The EU doctrine of collective dominance offers a potential alternative route for ex post enforcement against any aggregate effects of diffuse common ownership. 366 But merger control seems a preferable option given the "structural" nature of the common ownership problem, the potential countervailing efficiencies associated with it, and the fact that the type and tools of economic analysis employed to assess common ownership are similar to those applied to mergers. 367 Besides, it appears that the public enforcement will is missing to employ less flexible antitrust rules and procedures to tackle minority or common shareholding concerns.
In any event, starting with and tapping on their experience with ex post review, competition