EN EN

EUROPEAN COMMISSION

Brussels, 25.6.2013 SWD(2013) 239 final

Part 2/3

ANNEX

to the

COMMISSION STAFF WORKING DOCUMENT

Towards more effective EU merger control

2

Annex 1 Economic Literature on Non-Controlling

Minority Shareholdings ("Structural links")

Contents 1. EXECUTIVE SUMMARY......................................................................................... 3

1.1. Horizontal unilateral effects .............................................................................. 3 1.2. Horizontal coordinated effects .......................................................................... 4 1.3. Non-horizontal unilateral effects ....................................................................... 5 1.4. Deterring potential entry.................................................................................... 5 1.5. Efficiencies ........................................................................................................ 6

2. INTRODUCTION....................................................................................................... 6

3. FINANCIAL INTERESTS AND CORPORATE RIGHTS ....................................... 7

3.1. Silent Financial Interest or Passive Structural links .......................................... 8 3.2. Structural links that confer some degree of influence ....................................... 8

4. THEORIES OF HARM CONCERNING STRUCTURAL LINKS ........................... 9

4.1. Horizontal structural links ................................................................................. 9 4.1.1. Unilateral (direct) effects..................................................................... 9 4.1.2. Coordinated effects............................................................................ 11

4.2. Non-horizontal minority shareholdings........................................................... 15 4.2.1. Forward shareholdings ...................................................................... 15 4.2.2. Backward shareholdings.................................................................... 16

4.3. Effects of structural links on entry .................................................................. 18

5. EFFICIENCIES......................................................................................................... 19

6. MEASURING THE UNILATERAL EFFECTS OF HORIZONTAL STRUCTURAL LINKS............................................................................................ 20

6.1. Cournot competition and the Modified Herfindahl-Hirshman Index ("MHHI") ........................................................................................................ 20

6.2. Bertrand competition and the Price Pressure Index ("PPI")............................ 21 6.3. Further issues concerning the application of MHHI and PPI.......................... 22

REFERENCES.................................................................................................................. 24

3

1. EXECUTIVE SUMMARY

(1) The economic effects of minority shareholdings on competition in the market significantly depend on the financial interests flowing from them and the corporate rights conferred by them. Whereas financial interests refer to the acquiring firm's entitlement to a share of the profits of the target firm, corporate rights refer to the acquiring firm's ability to influence the acquired firm's competitive decisions. To be relevant, such influence relates to a firm's choice of prices, output, and product selection, and other competition variables.

(2) Regarding the impact that these two elements have on competition in the market, it must be noted that financial interests in a competing target firm tend to provide incentives for the acquiring firm to increase its own prices (or more generally, compete less aggressively), while the acquisition of significant influence by means of corporate rights gives the acquirer the ability to raise the competing target firm's prices.

(3) The economic literature identifies certain scenarios where the legal definition of control and "decisive influence" under the EU Merger Regulation (hereafter "Merger Regulation") would not be met, but nevertheless the holder of a non-controlling minority shareholding (hereafter "structural links") may still be able to exert material influence over the target firm with potentially significant anti-competitive effects.

1.1. Horizontal unilateral effects

(4) According to the economic literature, structural links among competitors can have anti-competitive effects, since they may increase firms' incentives and ability to unilaterally raise prices or restrict output. Intuitively, if firms have a financial interest in their competitors' profits, they 'internalise' the positive effects on their competitors' profits of a reduction in their own output or an increase in their own price. These anti-competitive effects may materialize even if a minority shareholding is passive, and therefore its holder has no influence in the target firm's decisions.

(5) The unilateral effects of the acquisition of structural links are typically of a significantly lower magnitude than in a full merger for two reasons. First, the holder of a structural link can internalise a smaller portion of the competitors' profits than in case of a full merger. Second, the direct price increasing effect only applies to the acquiring firm and not the acquired firm.

(6) However, severe unilateral effects may still arise if a structural link confers substantive or material influence to the acquirer on the target's strategic decisions even if it does not confer the possibility of exercising "decisive influence" in the sense of the EUMR. In that case, economic theory predicts that the acquiring firm may have the ability and incentives to induce the target firm to raise its prices. In some cases (e.g. cases where material influence is conferred), this price effect may be particularly significant, since the acquiring firm fully benefits from the positive effect of the competitor’s price increase but bears only part of the costs depending on the level of its financial ownership rights.

(7) Similar to full mergers, quantitative indices can be used as a first indication to assess the magnitude of competitive effects. In particular, techniques already

4

applied in merger control such as concepts that measure the upward pricing pressure generated by a transaction can be applied to minority transactions.1 The Herfindahl-Hirshman Index ("HHI")2 can also be easily adapted in order to account for minority shareholdings.

1.2. Horizontal coordinated effects

(8) The acquisition of a structural link may create or enhance competitors' incentives and ability to coordinate their conduct (that is, to engage in tacit collusion). Tacit collusion requires that a common understanding on the terms of coordination can be easily reached. Moreover, according to economic theory3, the market must be sufficiently transparent to allow the participants to monitor these terms and to allow deviations to be easily detectable and deterred against. Finally, reactions from outsiders must not be able to undermine the expected results of coordination.

(9) Structural links can enhance transparency if they offer the acquiring firm a privileged insight into the commercial activities of the firm in which it holds a share. One-sided or, in particular, reciprocal ownership links, i.e. cross-shareholdings between competitors may lead to or strengthen information exchange and thus facilitate monitoring the terms of coordination and detecting deviations. Depending on the circumstances, an increased information flow and higher transparency may therefore facilitate coordination.

(10) Moreover, the economic literature discusses the impact of structural links on firms' incentives to deviate from coordinated outcomes and to deter against such deviations (e.g. through price wars). As in the case of full mergers, structural links generally have an ambiguous impact on firms' incentives to sustain coordination. However, the economic literature also identifies various plausible scenarios where structural links unambiguously facilitate coordination.4 When firms are not expected to be able to adopt aggressive deterrence strategies, coordinated effects of structural links tend to be of a smaller magnitude than those of full mergers. On the other hand, in markets where firms have (or have shown) the required sophistication to credibly sustain collusion on the basis of price wars, structural links may have significant anti-competitive coordinated effects.5

1 A concept extensively discussed in the context of minority shareholdings is the pricing pressure

index ("PPI"). This is similar to the upward pricing pressure ("UPP") methodology that can be used in the context of full merger assessments.

2 See Commission horizontal merger guidelines. 3 These economic findings are also in line with and represented in the relevant case law and

guidance documents. 4 For example, the pro-collusive effects of minority shareholdings tend to dominate in markets

where competition can be expected to be strong in the absence of coordination. 5 Minority shareholdings may facilitate collusion by allowing for aggressive punishment strategies

in order to deter deviations. Aggressive punishment strategies include price wars, during which a potential deviator would be forced to participate in the losses incurred by a firm in which it has a minority stake.

5

1.3. Non-horizontal unilateral effects

(11) According to the literature, anti-competitive effects of structural links in upstream suppliers may primarily arise if firms can discriminate between customers.6 If significant economic influence in an upstream firm is conferred through partial backward integration and discrimination is feasible at the upstream level,7 a downstream firm may have the ability and the incentive to induce the target firm to foreclose inputs to its downstream rivals.8 In addition, a downstream firm may acquire a passive structural link in an upstream firm that supplies its competitor in order to participate in the upstream profits earned with sales of its competitor. This may soften downstream competition as the downstream minority shareholder partially internalizes the positive effects on its competitor of competing less aggressively through the financial interest. In scenarios where upstream firms cannot discriminate between customers, passive minority stakes in suppliers are unlikely to raise non-horizontal competitive concerns, but may rather improve coordination along the vertical chain.9

(12) While the literature has concentrated on input foreclosure so far, it is also conceivable that customer foreclosure may occur, especially if an upstream firm acquires structural links in a downstream firm that confer significant influence.

(13) Vertical structural links may even have more pronounced anti-competitive effects than full vertical mergers. In particular, this applies to foreclosure when the structural links confer significant influence as set out in the previous two paragraphs. However, potential effects should be assessed within the general perspective that vertical mergers are typically less likely to produce anti-competitive effects than horizontal mergers.

1.4. Deterring potential entry

(14) Structural links may have a deterrent effect on future entry. Competition concerns could arise when a structural link either induces potential entrants not to enter or significantly hinders third-party access to the equity of the target.

(15) Acquiring structural links in incumbents may deter potential competitors from entering. Through the structural link, the potential competitor partly internalises the loss that entry inflicts on the incumbent and may therefore credibly commit to not entering. Moreover, a structural link may confer sufficient influence, which allows the minority owner to prevent a potential entrant from entering.10

6 There is no literature on coordinated effects of minority shares in non-horizontal situations.

However, in vertical transactions coordinated effects often only play a minor role. 7 A scenario where the owner of a minority stake obtains extensive control rights may however be

captured under the existing test on "decisive influence". 8 See for instance Bundeskartellamt decision, B8-820000-U-111/94 Stadtwerke Garbsen of

30.9.1994. 9 When upstream products are sold through bidding processes, it is possible that minority shares

held by downstream bidders in the upstream firm may distort competition for these products 10 For example, in case M.4153 Toshiba/Westinghouse the Commission argued that the minority

owner could use veto rights in a joint venture to prevent the latter from entering into the owner's market.

6

(16) In addition, the prospect of foreclosure as a consequence of partial backward- or forward integration along the lines of paragraphs (11) and (12) may deter entry in the first place.11

1.5. Efficiencies

(17) According to the literature, structural links mainly create a financial interest in the performance of other firms in the market, typically without much scope for rationalization or avoiding cost duplication. Therefore, cost-based synergies are likely to be limited in the case of horizontal structural links.

(18) Efficiencies are expected to be more pronounced for vertical relationships. For example, vertical structural link may be useful in aligning the incentives of firms without the need to fully vertically integrate. Among others, this may allow to alleviate double marginalization, mitigate inefficiencies caused by asymmetric information or improve the service which is provided by downstream firms. If structural links entail important corporate rights, this may further facilitate vertical cooperation. However, the scope of these efficiencies is typically more limited than in case of a full merger.

2. INTRODUCTION

(19) The aim of this annex is to examine the theories of harm on unilateral and coordinated effects as a consequence of minority shareholdings.12 Typically, minority shareholdings refer to situations in which a shareholder holds less than 50% of the voting rights attached to the equity of the target firm.13 Although acquisitions of minority shareholdings may confer "decisive influence" as discussed below, this Annex mainly aims at delineating the effects of minority shareholdings which do not confer control in the legal sense and that are therefore not notifiable under the current EUMR. Such minority shareholdings are referred to as "structural links".

(20) This annex concentrates on minority shareholdings as the most prevalent class of minority interests. Other forms of minority interests include interlocking directorships (IDs), loans to competitors and contracts for differences (CfDs).14 Similar economic considerations often also apply to other forms of minority interests.

(21) There is a variety of reasons to invest in minority shareholdings. For example, equity investments may be a means to gain access to specific assets, including new technologies, additional financial resources, and innovative managerial practices amongst others. 15 Equity block purchases are found to lead to increased

11 See for instance Bundeskartellamt decision, B8-820000-U-111/94 Stadtwerke Garbsen of

30.9.1994. 12 As far as possible, this annex also examines the scope for efficiencies as a result of structural

links. 13 There appears not to be a clear cut definition of minority shareholdings. This may be partially due

to antitrust agencies having different approaches for reviewing partial ownership transactions. 14 Interlocking directorates refer to situations in which one or more companies have one or more

members of their respective boards or a top executive in common. Contracts for differences are derivatives on other firms’ equity or debt value. See DotEcon (2010) or OECD (2006).

15 OECD (2006), p. 22.

7

stock values, investment expenditures and operating cash flow of the target firms especially when the participating firms also form alliances or joint ventures.16

(22) However, from a competition policy perspective, the possible anti-competitive effects of structural links have to be balanced with benefits of a partial integration. In particular, structural links may grant the acquiring firm varying degrees of financial interest and corporate rights in the target firm. The economic effects of these financial interests and corporate rights differ considerably, as will be discussed in this annex.17

(23) This annex is mainly based on theoretical considerations as there is currently only a limited empirical literature on the effects of structural links.18

3. FINANCIAL INTERESTS AND CORPORATE RIGHTS

(24) When a firm wholly acquires another target firm, the acquirer typically obtains both control and financial entitlements.

(25) In the context of minority ownerships, it is important to distinguish between the concepts of financial interest and corporate control.

(26) Salop and O’Brien (2000) refer to these concepts as follows: "Financial interest refers to the acquiring firm's entitlement to a share of the profits of the acquired firm. Corporate control refers to the acquiring firm's ability to control or influence the acquired firm's competitive decision making, including pricing and product selection as well as sale of the company's assets."19 Although often greater financial interest is accompanied by greater corporate influence, there may be instances where the levels of financial interest and corporate influence differ considerably.20

(27) Financial interest or cash-flow rights are often relatively simple to quantify as an agent owning X% of a firm is entitled to X% of its profits and earns X% of the proceeds if the firm is entirely sold.

(28) In contrast, the level of influence may be difficult to assess in practice. There are a variety of situations where a relatively small ownership share can provide its owner with significant influence over a firm given the dispersion and the lack of protection of the remaining shareholders.21

(29) This annex primarily concentrates on the economic notion of control, that is, to what extent the objectives of a (minority) stakeholder are taken into account in a firm's decisions. This may differ from legal notions of control. In particular, even in situations where the test of "decisive influence" as used in the context of the

16 See Allen and Phillips (2000). 17 Salop and O’Brien (2000), p. 563. 18 Alley (1997) estimates the degree of anti-competitive effects due to cross-shareholdings but states

that it is impossible to estimate the effect a change in minority shareholdings has on the degree of collusion due to limitations of his data. Reiffen (1998) empirically examines the effect of partial vertical integration on the incentive to foreclose rivals. His findings could be (partly) explained by efficiencies or by the lack of significant influence of the acquiring firm in the target's business decisions after the transaction. Parker and Röller (1997) estimate a structural model of competition for the U.S. cellular telephone industry but use a different notion of "cross-ownership" with different competitive effects than structural links as discussed in this appendix.

19 Salop and O’Brien (2000), p. 568. 20 Interlocking directorates may be viewed as a scenario in which two firms have some degree of

control on each other but typically no financial interest. 21 See e.g. OECD (2008), p.20.

8

EUMR provisions is not met, an external stakeholder may still have substantive influence in a firm.22 The main focus of this annex is on scenarios in which the criterion of "decisive influence" is generally not met, so that the minority shareholders have at most a "material influence" in the target firm.23

3.1. Silent Financial Interest or Passive Structural links

(30) A minority shareholder may not have any influence over the target’s competitive behaviour in situations where the ownership share does not confer special corporate rights. In particular, this may be the case if the acquirer holds only non-voting stock.

(31) Although passive minority shareholdings do not allow influencing the target's decisions, they might nevertheless raise competitive concerns to the extent that they induce the acquirer to increase prices as laid out below.

3.2. Structural links that confer some degree of influence24

(32) Salop and O’Brien (2000) develop a number of scenarios in which the minority shareholder may influence the decision of the target only to a certain degree.25

(33) Even though a minority shareholding may confer corporate rights, legal provisions may protect other minority shareholders and therefore restrain the acquirer's ability to influence the target's decisions.26 The acquiring firm may be induced by corporate law to essentially manage the partially acquired firm as if it were a stand-alone entity. In practice, the partially owned firm may still have some degree of control but may in other respects act similarly to the silent financial interest scenario presented above.

(34) In other cases, it may be a good approximation to assume that the degree of influence the acquiring firm has in decisions of the target firm may be roughly proportional to its financial interest.27 This scenario may for example occur if the minority shareholder may block certain decisions (e.g. in case of supermajority requirements) so that compromises with other shareholders have to be found.

(35) In special cases, even a relatively small financial interest may yet confer material influence on the minority shareholder. This may for example occur if the minority shareholder has been given special corporate rights or if the minority owner may form a coalition with other shareholders, thereby obtaining more influence than suggested by the joint financial interest.

22 See Article 3 of the EUMR. 23 Salop and O’Brien (2000) also analyse situations where the acquiring firm gains significant

control on the target firm. In particular, "Total Control" is defined as a situation where the acquirer may gain total control by inter alia owning the majority of the voting stock or by having been designated as the operating partner. Similarly, "One-way Control" is described as a situation where the acquiring firm has enough power to force the acquired firm to maximize the sum of both firms' profits. The acquiring firm continues to maximize its individual profit, including the financial interest in the target firm. In the context of joint venture, a similar mode of control is referred to as limited joint control by Bresnahan and Salop (1986).

24 This term has been coined by Salop and O’Brien (2000) and generally refers to scenarios in which the minority shareholder has at least some influence in the decisions of the target firm.

25 Two particularly relevant scenarios discussed by Salop, S and O’Brien, D. (2000), are proportional control and fiduciary obligations.

26 See the fiduciary obligations scenario in Salop, S and O’Brien, D. (2000), p. 580. 27 See also paragraph (84) and Salop and O’Brien (2000), p. 583.

9

(36) Summing up, there is a whole range of corporate scenarios with different economic implications. Acquisitions of minority shareholdings that confer some degree of influence may not satisfy the criterion of "decisive influence" and thus may not be covered under the EUMR provisions.28 In particular, when comparing situations in which the test of "decisive influence" is (not) satisfied, there may nevertheless be considerable differences in terms of influence.

4. THEORIES OF HARM CONCERNING STRUCTURAL LINKS

4.1. Horizontal structural links

4.1.1. Unilateral (direct) effects

(37) As first shown by Reynolds and Snapp (1986) and by Bresnahan and Salop (1986), financial interests among competing firms –even without any control rights– may lead to less vigorous competition. Salop and O’Brien (2000) point out that under suitable assumptions, horizontal minority ownerships have an anti-competitive effect in models where firms compete with differentiated goods in prices ("Bertrand competition") or where firms compete in quantity ("Cournot competition"). Intuitively, when having a financial interest in competitors' profits, firms ‘internalise’ the positive effects from restricting their own output or raising their prices on their competitors' profits. Therefore, in the absence of entry, the equilibrium market output will typically decline as one or more competitors increase the level of financial interests in rival firms. This usually implies that consumer surplus falls as well due to resulting higher prices.29

(38) Whereas financial interest in a competing target firm tends to provide incentives of the acquiring firm to increase its own prices, the acquisition of corporate rights gives the acquirer the ability to raise its competitor's prices.30 If a structural link confers enough influence, the minority owner typically has incentives to induce the target to compete less aggressively as this typically feeds back to higher profits of the acquiring firm. Therefore, to the extent that structural links confer influence over the target company, the anti-competitive effects are likely to be higher than in a scenario where they give only rise to financial interests.

(39) Basically for the same reasons as mentioned in the preceding paragraphs, investors that have (minority) shareholdings in several directly competing firms have an interest that each of their target firms takes into account the externalities that its (pricing) decisions have on rivals of which the investor also holds a share. Depending on how managers in the target firms take the preferences of their shareholders into account, similar effects as described above may thus arise if investors own (minority) shareholdings in direct competitors.31Horizontal unilateral effects are not captured by an antitrust analysis based on conventional

28 Acquisitions of minority shareholdings with total control are typically covered under the European

Merger Regulation provisions whereas passive minority shareholdings typically are not. 29 Farrell and Shapiro (1990) point out that an acquisition of a passive ownership may have positive

welfare effects under special circumstances. Farrell and Shapiro (1990), Flath (1991) and Reitmann (1994) show that in the absence of efficiencies, acquisitions of passive ownership shares tend to be unprofitable in a model of Cournot competition if there are many firms in the market and if firms are not too asymmetric since then non-investing firms benefit from such an investment.

30 See for example Spector (2011). 31 See e.g. Salop, S and O’Brien, D. (2000) or Ezrachi and Gilo (2006).

10

structural indexes, such as concentration ratios. Therefore Reynolds and Snapp (1986) as well as Salop and O’Brien (2000) propose amended concepts to take into account the effects of horizontal structural links. Section 6 first describes quantitative concepts to take into account the unilateral effects of horizontal structural links and then discusses their application in practice.

(40) Similar considerations apply to joint ventures. In particular, the unilateral effects of a transaction involving a horizontal joint venture ("JV") can be analysed once the financial and control rights are known. The pricing interests of parents that compete in the same markets as the JV depend on their financial stakes in the JV. As a JV typically does not have a financial stake in the parents, its pricing incentives depend on its governance and ownership structure.32

(41) Typically the competitive effects of the acquisition of a structural link are of a lower magnitude than those of a full merger. In the framework of Salop and O’Brien (2000), a full merger can be thought of as an acquisition of control via corporate rights and of full financial rights. It is therefore intuitive, that for example an acquisition of a passive structural link creates less upward pricing pressure since the acquiring firm only partially internalizes the externalities of a price increase to the target firm.33 A similar reasoning holds as long as the acquired control is sufficiently narrow compared to the financial interest of structural links.

(42) However, structural links may lead to a relatively large price increase if the acquisition of a minority stake confers significant corporate rights to the acquirer.34 In cases of this kind (which may be already covered by the existing merger regulation to the extent that the legal criterion of "decisive influence" is met), economic theory predicts that the acquiring firm may coerce the target firm to significantly raise its prices, because the acquiring firm fully benefits from the positive externalities of the competitor’s price increase but bears only part of the costs depending on the level of its financial ownership rights. In the context of JVs, Salop and O'Brian (2000) also point out that a parent firm that is also a horizontal competitor and that enjoys significant control rights, may induce a JV to set anti-competitively high prices if it has only a limited financial share in the JV.

(43) On the other hand, as discussed by Dubrow (2001), there might be several countervailing factors that potentially attenuate the effects of structural links. These include, for example, the presence of incomplete information, conflicting

32 Kwoka (1992) investigates firms' incentives to set up joint ventures. Bresnahan and Salop (1986)

provide a more detailed exposition of common control governance structures and their implications.

33 For example, it is straight forward to verify that regarding passive minority shareholdings, the change in the quantitative indices discussed in section 6 increases monotonically in the proportion of the acquired stake in the target firm, and stays always well below the change of these indices in case of a full merger.

34 Brito, Cabral and Vasconcelos (2010) analyse related issues. Amongst others, they show that turning voting shares into non-voting shares may increase consumer surplus.

Foros, Kind and Shaffer (2010) point out that the possibility of increasing prices by help of such a minority stake renders it particularly profitable for both participants to engage in such a transaction (See also Karle, Klein and Stahl (2011)). They argue that in the Norwegian pay-TV market a minority shareholding with ample control rights of Telenor (that fully owns the pay-TV provider Canal Digital) in RTV led to very high prices. According to Ezrachi and Gilo (2006), for the same reason the main shareholder of a firm may also find it very profitable to acquire (minority) shareholdings that confer significant influence in a direct competitor of that firm.

11

management incentives or the inability of the minority shareholder to capture the benefits earned by the target firm due to conduct of the acquiring firm. This points to the need to carefully analyse how decisions are taken in the context of a firm with structural links, and the extent to which investors’ and management’s incentives are aligned.35

(44) Taken together, concerns on the basis of unilateral effects of horizontal structural links are warranted primarily in settings where full mergers would be clearly anti-competitive (e.g. scenarios involving structural links between close competitors with significant market shares and high barriers to entry). Unilateral anti-competitive effects may arise even when a structural link only confers little influence. Moreover, minority stakes with relatively small financial interest may have significant unilateral effects in specific cases where substantial influence is conferred even though the criterion of “decisive influence” is not met.

4.1.2. Coordinated effects

(45) The unilateral effects from the acquisition of structural links discussed so far are based on the assumption that firms in the market set prices independently. Now we turn to the question how structural links affect the firms' incentives to tacitly or expressly collude in order to achieve supra-competitive profits.36 The main focus in the context of coordinated effects is whether structural links among competitors facilitate or hamper collusion.

(46) Economic theory has derived a number of necessary conditions that have to be met in order for the colluding parties to reach terms of coordination and sustain the coordination over time.37 The following conditions may be particularly affected by structural links:

• The market conditions must be sufficiently transparent to allow each cartel member to monitor the other firms' behaviour. • In order to deter deviations from the common strategy, the threat of future retaliation must be must be sufficiently severe and credible.

4.1.2.1. Information and transparency

(47) The acquisition of a structural link may enhance transparency as it typically offers the acquiring firm a privileged view on the commercial activities of the target. According to the OECD (2008), even "passive minority shareholders may have access to information that an independent competitor would not have, such as plans to expand, to merge with or to acquire other firms, plans to enter into major new investments; plans to expand production or to enter or expand into new markets".38 It is possible that a structural link that additionally confers a high degree of corporate rights also provides access to more detailed information.

35 See Salop and O’Brien (2001). 36 Salop and O’Brien (2000) coin the notion of Coasian Joint Control is a case in which "the

managers of the acquiring firm try to maximize the joint profits of both the acquired firm and the acquiring firm" (p. 581). They also point out that unilateral incentives to deviate so as to maximize independent profits may prevent the Coasian outcome from being achieved. This scenario is therefore reminiscent of collusion.

37 See Ivaldi, Jullien, Rey, Seabright and Tirole (2003). 38 OECD (2008), p. 30.

12

Reciprocal ownership links between competitors may therefore permit a higher degree of information exchange.

(48) The extent of increased information flows due to minority ownerships depends on the information that is already publicly available. Information may for example be publicly accessible due to reporting obligations in the case of publicly traded companies.

4.1.2.2. Incentives to coordinate

(49) For a given level of transparency, the scope of collusion depends on the colluding party's profits when adhering to the common strategy as compared to those in case of a deviation. A firm's overall profits from deviation reflect both the profit of the deviating firm until the remaining colluding parties realize that a deviation has occurred and the profit earned by that firm once all market participants have reacted to the deviation.

(50) The literature has so far focused on purely passive structural links.39 Passive structural links generally produce at least two countervailing effects. On the one hand, due to partial ownerships firms internalize part of the losses that they inflict on rivals when they deviate, therefore reducing the incentives to deviate. On the other hand, they can also soften competition following the break-down of the collusive scheme and hence strengthen the incentives for firms to deviate.

(51) Malueg (1992) examines the balance of these two effects under the assumption that once a firm has deviated from a colluding strategy, the market participants revert to quantity-based Cournot competition in the absence of collusion. Malueg shows that under this assumption the impact of an increase in passive investment levels on the feasibility of collusion is ambiguous. That is, partial shareholdings may actually render collusion more difficult for certain levels of passive investments.

(52) In contrast, Gilo, Moshe and Spiegel (2006) argue that passive structural links tend to facilitate collusion on the basis of a framework in which firms compete in prices with homogenous goods.40 In their framework, the collusion-facilitating effect of structural links (due to the internalization of the losses imposed to competitors during a deviation from the collusive scheme) typically dominates the effect of softer competition after a deviation.

(53) These results suggest that the anti-competitive coordinated effects of structural links tend to dominate in markets with intense competition in the absence of collusion. Indeed, Kühn and Rimler (2006) show that as products become more homogeneous in a setup of price competition, the unilateral effects of minority shareholdings on profits after firms have reverted to competitive behaviour are rather small. Therefore, the collusion-facilitating effect of structural links through lower gains from deviation dominates in markets with intense price competition.

(54) Furthermore, Gilo, Moshe and Spiegel (2006) emphasise that structural links facilitate collusion whenever they help the firm with the strongest incentives to

39 There is no literature on coordinated effects of structural links that also confer partial control. Gilo

(2000) argues that an acquisition of competitors' debt may have similar, albeit more diluted effects than in the case of the acquisition of stock if two further conditions are satisfied. First, fierce competition must reasonably induce the insolvency of the debtor and the debt must not be guaranteed by third parties.

40 Concerning the punishment of deviating firms, Gilo, Moshe and Spiegel (2006) make the same assumption as Malueg (1992).

13

deviate to credibly stick to the collusive agreement. This result is a direct implication of the general observation that a collusive scheme cannot be sustained if even a single firm has an incentive to deviate. Collusion may thus be facilitated if the firm with the highest incentives to deviate from a collusive scheme acquires minority shares in its competitors.

(55) Based on this finding, Gilo, Moshe and Spiegel (2006) point out that there is a fundamental difference between horizontal mergers in which firms obtain significant control over their rivals and passive investments in rivals that they study. While the Commission’s Horizontal Merger Guidelines states that the likelihood or significance of coordinated effects may be increased if a merger involves “a ‘maverick’ firm that has a history of preventing or disrupting coordination"41, this is typically not the case if some firm acquires a passive minority share in a firm with strong incentives to deviate.

(56) Kühn and Rimler (2006) show that structural links may be particularly conducive to collusion when firms can credibly recur (or have recurred in the past) to aggressive deterrent strategies in case of deviations from a collusive scheme. Without structural links, the means of punishing deviators are limited because firms can effectively avoid heavy losses by maintaining prices close to marginal costs (and thus reducing their output if market prices fall below their costs). The absence of a severe punishment therefore restricts firms’ possibilities to collude. If instead colluding firms have minority shareholdings in their competitors, then they indirectly suffer also from losses incurred by their competitors. Hence, more aggressive punishment strategies can be implemented and thus collusive schemes may be more easily sustained. This concern is most relevant in industries where there have been past episodes of intense price wars (as a result of which market participants incurred losses), or where aggressive punishment are likely to be implemented in the future (if coordination breaks down). Note that largely passive minority shareholdings and thus the absence of significant control is actually a contributing factor for sustaining collusion in this way.

(57) When several competitors have structural links in a third firm, this may facilitate collusion even in case the third firm is active in another market. In the case a minority shareholder deviates from the collusive behaviour, firms may revert to less collaborative behaviour in the jointly owned firm. The deviating minority shareholders are therefore punished by participating in the losses of the jointly owned firm.42

(58) In order for the passive investment to serve as an effective commitment not to deviate, it is necessary that other colluding parties believe that the transaction is effective. For example, this may be the case if the transaction can be publicly observed.43

(59) Summing up, horizontal structural links may have effects on market participants' ability and incentives to coordinate, especially if firms can credibly recur to aggressive deterrent strategies or if competition can be expected to be intense absent any coordination. Therefore, special attention to this issue would be warranted in industries where there have been past episodes of intense price wars (as a result of which market participants incurred losses), or where aggressive

41 European Commission (2004), paragraph 42. 42 For example, Martin (1995) shows that competitors may set up a R&D joint venture in order to

sustain collusion as long as each of the colluding firms has enough control rights in the JV. 43 See OECD (2010) and Gilo (2000).

14

punishments are likely to be implemented in the future if coordination breaks down.

15

4.2. Non-horizontal minority shareholdings

(60) Non-horizontal transactions may lead to competition concerns, in particular in relation to input or customer foreclosure.44 As in the case of full mergers, structural links can also lead to non-horizontal effects.45 As with horizontal shareholdings, an important distinction can be made between controlling and passive ownership. Specific to vertical shareholdings is their direction: a forward shareholding arises if an upstream firm owns shares of a downstream firm. A backward shareholding arises if a downstream firm owns shares of the upstream firm.

4.2.1. Forward shareholdings

(61) The effects of an upstream firm acquiring a passive structural link in a downstream firm, thereby participating in the profits of the latter, depend on whether upstream price discrimination is possible or not.

(62) In a setting where price discrimination is not possible and in which the upstream firm sets linear tariffs, the upstream firm typically expands the sold quantity and the resulting lower wholesale prices are passed through to consumers.46 Intuitively, once an upstream firm internalizes part of the downstream margin, it is willing to increase quantity at the expense of its upstream margin. This reduces double marginalization and is thus pro-competitive. The effect increases in the ownership share with a full merger as the limit case.

(63) Under the (presumably more relevant) alternative scenario whereby upstream firms may discriminate between buyers, an upstream minority shareholder may have incentives to (partially) foreclose its inputs to other downstream. Spiegel (2011) as well as Gilo, Levy and Spiegel (2012) show that incentives to foreclose are typically weaker compared to fully integrated firms because only a (small) proportion of the downstream profits accrue to the upstream firm through its minority share. Hence the upstream loss of foregone sales is traded off with only a fraction of the downstream gain of a weaker competitor. They also note that input foreclosure is unlikely to be further aggravated if the upstream minority shareholder obtains control rights in the downstream firm through the minority stake.

(64) Passive structural links in a downstream firm may also allow an upstream firm to protect its market power when contracts are not publicly observable. Rey and Tirole (2007) point out that an upstream firm may be unable to sell its products at a profit maximizing price if contracts between upstream and downstream firms are secret.47 By acquiring a passive structural link in a downstream firm, the

44 European Commission (2008), paragraph 18. 45 There are several recent cases of the Commission in which a non-horizontal theory of harm

involving minority shares was pursued. Examples are M.3696 E.ON/MOL, M.4153 Toshiba/Westinghouse or M.5406 IPIC/MAN Ferrostaal. One particularly relevant form of vertical (minority) shareholdings is vertical joint ventures. For example, often joint ventures are set up in order to provide inputs to its parents. See e.g. M.1940 Framatome/Siemens/Cogéma/JV.

46 See Flath (1989). 47 After having concluded a high-price contract with one downstream firm, the upstream firm would

be tempted to offer another downstream firm a cheaper price since the upstream firm normally does not fully bear the negative externalities thereby inflicted on the first downstream firm. Since each firm anticipates the upstream firm's temptation to offer cheaper contracts to competitors, the

16

upstream firm participates in the target's downstream losses when also selling its inputs at more favourable terms to a downstream competitor. This structural link thus allows the upstream firm to alleviate the commitment problem discussed by Rey and Tirole (2007), albeit to a lesser extent than a vertical merger. As a structural link partially restores the upstream firm's power to sell at high prices and may thus lead to higher prices, it reduces the need of the upstream firm to foreclose downstream firms (e.g. by means of exclusive dealing arrangements) in order to circumvent the commitment problem.48

(65) Gilo and Spiegel (2011) also note that controlling partial forward shareholdings may lead to more customer foreclosure than a full merger: A supplier with a controlling partial forward shareholding fully benefits from increased upstream sales when inducing its downstream target not to buy inputs from competing customers, but only bears a share of the foregone profits of the downstream firm.

4.2.2. Backward shareholdings

(66) The immediate effect of holding passive structural links in an upstream supplier is that the downstream firm partly internalizes the upstream margin of its product and thus has an incentive to sell more quantity at the expense of its downstream margin. As a consequence, the supplier faces an increased demand and thus has an incentive to increase its upstream price.

(67) Under the assumption that the upstream firm may not discriminate between buyers, Flath (1989) as well as Greenly and Raskovich (2006) show that the different effects of partial vertical integration tend to neutralize. In Greenly and Raskovich (2006), the sold quantities of the non-owning firms are reduced since these firms end up paying effectively higher input prices, but this reduction tends to be offset by additional sales of the downstream firms with minority shareholdings.

(68) Passive backward structural links may yet have anti-competitive effects if there is some competition in the upstream market and if upstream price discrimination is feasible. In this scenario, passive shareholdings of a downstream firm in a supplier that also serves downstream competitors may soften downstream competition according to Hunold, Röller and Stahl (2012). As the downstream shareholder participates in the upstream firm's profits, it benefits from input sales to its downstream competitors. Hence, a downstream minority shareholder has incentives to increase its sales price.49 Hunold, Röller and Stahl (2012) point out that in order for this effect to materialize, the minority shares must not confer significant control. In particular, in markets where a full merger would decrease downstream prices due to reduced double marginalization, passive backward shareholding increases downstream prices.

(69) A backward structural link in a downstream firm in its supplier may facilitate input foreclosure if the supplier can discriminate among buyers and if significant

upstream firm is essentially forced to sell its goods for a price that is below the price it would charge in case of observable contracts. See also European Commission (2008), paragraph 44 footnote 5.

48 Note that financial interest in the downstream firm is sufficient for this theory of harm – control rights are not necessary.

49 This mechanism is similar to that of passive horizontal structural links, but the magnitude of the anti-competitive effect depends inter alia on the upstream margin rather than on the downstream margin.

17

influence is conferred. As shown by Spiegel (2011) as well as Gilo, Levy and Spiegel (2012), the downstream minority shareholder may induce the target upstream firm to foreclose downstream rivals, even though this typically reduces the upstream profits.50 The controlling downstream firm internalizes the full benefit of a weaker downstream competitor, but only bears its minority share of the upstream losses due to foregone sales. Hence, the specific case where a downstream firm acquires a minority stake in an input provider that confers material influence may be even more conducive to input foreclosure than a full merger. However, it is important to note that the model relies on the assumption that the controlling (downstream) firm is not residual claimant of the partially owned firm's profits.51

(70) Gilo and Spiegel (2011) also note that by the same logic passive partial backward integration is less conducive to customer foreclosure than a full merger as the downstream losses are fully internalized, but the upstream gains only partly accrue to the downstream firm with the structural links.

(71) The following Table summarizes the effect of vertical structural links with material influence on the incentive to foreclose competing firms:52

Partial Forward Integration

Partial Backward Integration

Input foreclosure - +

Customer foreclosure + - "+" means that the effect may be more pronounced than in a full merger. "-" indicates that the effect is typically less pronounced than in a full merger.

(72) If the interaction between two vertically related markets takes the form of tenders, passive minority ownerships of a downstream firm in an upstream firm may lead to another more subtle competitive distortion.53 According to Bulow, Huang and Klemperer (1999), there is both empirical and theoretical evidence that bidders with some toehold in the auctioned object are typically bidding more aggressively than other bidders do. Moreover, when bidders are uncertain about the true value of the auctioned good and when the private information is relevant for all other participants, then bidders without a toehold may bid even less aggressively in order to avoid overpaying for the sold item.54

50 See also Baumol and Ordover (1994). 51 Sometimes, controlling firms conclude profit transfer agreements with the partially owned firm

that guarantees the remaining owners a fixed profit. 52 See Gilo and Spiegel (2011). 53 See Bulow, Huang and Klemperer (1999) and Spector (2011). 54 In this so-called common value environment, each bidder typically incorporates in its bidding

strategy that if winning the auction, the others must have bid less, which in turn means that they had less optimistic private information. Since also the others' private information is relevant for the common value of the sold object, a winning bidder that inferred the object’s value only on the basis of the own private information would be likely to overestimate its value (which is the "winner’s curse"). Therefore, bidders rationally adjust their bids downwards, compared to what they would have quoted based on their "naive" estimate. Since a bidder with a toehold is supposed to bid more aggressively than other bidders, if a bidder without toehold wins, this means that bidders with toehold must have very unfavourable information. Therefore, it is rational for bidders without toehold to bid even more cautious.

18

(73) The bidding behaviour increases the likelihood that a partially vertically integrated bidder wins the auction. Moreover, bidders with a minority stake win while paying a relatively low mean price. These distortions are unrelated to the intrinsic merits of the integrated firm and possibly give rise to inefficiencies.55

(74) Summing up, if upstream firms can discriminate among buyers, especially partial backward integration that confers material influence in upstream firms may lead to input foreclosure. Furthermore, vertical minority shares in a downstream firm that confer far-reaching control rights may be conducive to customer foreclosure. In addition, when upstream firms differ in their costs to produce input goods, passive vertical minority shares may be used as a commitment device to soften downstream competition. This is in line with several recent cases of the Commission where a vertical theory of harm was pursued.

(75) Taken together, forward passive shareholdings tend to reduce double marginalization but may to some extent also facilitate (partial) foreclosure and may help upstream firms to commit to higher prices if contracts with downstream firms are not observable to all downstream customers. Backward minority shareholdings that confer control to the downstream firm tend to significantly facilitate input foreclosure, whereas passive backward shareholdings may dampen downstream competition and thereby lead to increased consumer prices. Moreover, passive backward shareholding can distort competition in tender markets. Although the literature has focused on input foreclosure so far, it is conceivable that similar considerations hold for customer foreclosure.

4.3. Effects of structural links on entry

(76) Structural links may in certain circumstances deter future entry. In its contribution to the OECD (2008) report, the Commission has pointed out that competition concerns could arise when a structural link either significantly impedes third-party access to the equity of the target via acquisition, or when it makes it less likely that the acquirer enters itself in the market where the target is active. A formal economic model on these concerns has not been developed yet.

(77) Minority stakes of potential competitors in incumbents may induce them not to enter.56 By help of the structural link, the potential entrant partly internalizes the loss that entry inflicts on the incumbent and may therefore credibly commit to not entering. Conversely, in the case M.4153 Toshiba/Westinghouse the Commission argued that by help of veto rights in a joint venture, the minority owner could prevent it from entering into the owner's market.

(78) In certain circumstances, structural links in a competitor may potentially deter a third firm from taking over the remaining shares of the target firm. If the third firm would be in a position to improve the competitiveness of the competitor, the

55 According to Harbord and Binmore (2000), these effects could have materialized if the acquisition

of the Manchester United football club by the TV channel BSkyB were cleared. Manchester United received roughly 7% of the proceeds from the sales of the Premier League’s broadcasting rights. According to Bulow, Huang and Klemperer (1999), the acquisition was blocked in large part because of concerns that by acquiring Manchester United and therefore participating in the proceeds of the tenders for the broadcasting rights, BSkyB as a bidder in these tenders would be able to shut out other television companies.

56 See Gonzales-Diaz (2012).

19

minority shareholder may thus prevent the competitor from becoming stronger.57 However, this strategy seems only effective if the third firm cannot effectively enter without acquiring the target firm. In addition, the third firm may nevertheless have incentives to acquire the remaining shares of the target firm, because the minority ownership tends to soften competition even after entry of the third firm.

(79) In addition, the prospect of (partial) foreclosure as a consequence of partial backward- or forward integration along the lines of Section 4.2 may deter entry in the first place.58

(80) Apart from these specific cases, horizontal structural links should be conducive to entry since they tend to soften competition. In the absence of entry-barriers, potential anti-competitive effects of structural links may thus be attenuated by subsequent entry.

5. EFFICIENCIES

(81) Structural links mainly create a financial interest in the performance of other firms in the market, typically without much scope for rationalization or avoiding cost duplication.59 Therefore, synergies seem to be limited for horizontal structural links.

(82) Many sources of efficiencies realized by vertical integration may also be relevant for vertical structural links.60 Preferences of upstream and downstream firms may be aligned to some extent by passive minority interests. Among others, this may help alleviating double marginalization (as explained in 3.2), mitigating inefficiencies caused by asymmetric information or improving the service which is provided by downstream firms.61 Minority shareholdings may be also helpful to overcome difficulties arising from incomplete contracts, that is, from not being able to take into account all possible contingencies.62 For instance, issues to do with incomplete contracts may be particularly prevalent in research and development (R&D).63 Control rights conferred by minority shareholdings may further facilitate vertical cooperation. Similarly, cross-shareholdings might be useful in aligning the incentives of firms involved in alliances or joint ventures. This may be relevant when these projects require relationship-specific investments.64

57 Competitiveness could be improved e.g. by providing access to more efficient production

technology or by being able to invest large amounts of capital in R&D. 58 See for instance Bundeskartellamt decision, B8-820000-U-111/94 Stadtwerke Garbsen of

30.9.1994. 59 Gilo (2000) briefly discusses efficiencies. 60 For typical effects of vertical integration, see European Commission (2008), paragraphs 52-57 or

Perry (1989). 61 Güth, Nikiforakis and Normann (2007) show how vertical structural links may facilitate trade if

the seller has private information on the sold good. 62 Dasgupta and Tao (2000) as well as Riordan (1991) analyse to what extent passive structural links

in a vertical setting alleviate inefficiencies due to incomplete contracts. 63 Fee, Hadlock and Thomas (2006) report that indeed equity stakes are much more common when

the supplier is a R&D-intensive firm or when a significant fraction of the sales are made to the customer.

64 Structural links may alleviate hold-up problems. This may lead to more efficient investment choices. For the hold-up problem see e.g. Che and Sákovics (2008).

20

6. MEASURING THE UNILATERAL EFFECTS OF HORIZONTAL STRUCTURAL LINKS

(83) This section first describes quantitative concepts to take into account the unilateral effects of horizontal structural links and then discusses their application in practice.65 If firms are supposed to compete in quantities, then a modified Herfindahl-Hirshman Index ("MHHI") is suggested. In a framework of price competition, a version of the Price Pressure Index ("PPI") is proposed.

(84) Assessing the degree of influence of a minority shareholder in decisions of the target firm may be difficult. Salop and O’Brien (2000) as well as Brito, Cabral and Vasconcelos (2008, 2010) discuss scenarios in which the degree of influence can be related to the equity share of the minority owner. In particular, in the scenario of “Proportional Control” the influence of the acquiring firm over the acquired firm is basically proportional to its financial interest.66 The acquired firm thus behaves as if it had the same financial interest in the acquiring firm as the latter has in the acquired firm. For example, if the acquiring firm has a 25% stake in the target, under “proportional control” the target would act so as to maximize its own and 25% of the acquirer’s profits. “Proportional control” may for example occur if the minority shareholder may block certain decisions (e.g. in case of supermajority requirements) so that compromises with other shareholders have to be found.

(85) Brito, Cabral and Vasconcelos (2008, 2010) develop a more sophisticated model of proportional control that takes into account the dispersion of shareholders and allows for both preferred and voting stock.67

6.1. Cournot competition and the Modified Herfindahl-Hirshman Index ("MHHI")

(86) The conventional HHI is calculated by summing up the squared market shares of the market participants.68 In an oligopoly model of quantity competition among firms that produce homogeneous goods and are protected by entry barriers, the conventional HHI is related to the average margin between the market price and cost.69 The change in HHI caused by a full merger (the delta or Δ) equals twice the product of the merging firms' market shares.70

(87) Similarly, the MHHI is an indicator of the average price-cost margin that additionally takes into account the anti-competitive effects of partial ownerships.

65 Unilateral effects of horizontal structural links can be also assessed based on a merger simulation,

as set out by Brito, Ribeiro, and Vasconcelos (2013). 66 See Salop and O’Brien (2000), p. 583. 67 The European Commission also acknowledges the importance of shareholder fragmentation on

effective control. See e.g. Case IV/M.0025 Arjomari/Wiggins, Commission decision of December 12, 1990 (1990 O.J. C321) where the Commission found that Arjomari was able to exercise decisive influence over Wiggins with 39% of the shares since the rest was spread among 107,000 other shareholders, none of whom holding more than 4%.

68 Formally, when denotes the market share of firm , . 69 More precisely, Cowling and Waterson (1976) have shown that if firms compete with

denoting marginal cost of firm , referring to the market share of firm , being the equilibrium

price and the demand elasticity, then .

70 Formally, .

21

Salop and O’Brien (2000) derive a general formula for the MHHI that allows for partial ownerships entailing different degrees of financial interest and control rights. The MHHI is defined such that it collapses to the conventional HHI in the absence of any partial ownership. The delta between the MHHI pre- and post-acquisition provides an estimate of the competitive effects caused by the transaction. Similarly to the HHI, computations based on the MHHI are only rough indicators in that "they assume a relevant market that entails no substitution to products outside the market, prohibitive entry barriers, no other competitive effects factors, and no efficiency benefits".71

(88) As an example, consider a market where firms A, B and C with market shares 40%, 30% and 30%, respectively, are active. Before any transaction takes place, the HHI and the MHHI are 3,400. If firm A takes over firm B, the HHI increases by 2,400, thereby indicating a merger with very significant effects on concentration.72 When firm A acquires a passive 10% minority stake, the MHHI increases only by 120.73 According to the Horizontal Merger Guidelines of the European Commission (2004), a delta of less than 150 would be unlikely to raise competition concerns in case of a full merger as long as certain special factors are absent. As the MHHI and the HHI are similar concepts, this threshold could also be used to conclude as a first rough indication that concerns are unlikely to arise in this case. In contrast, if firm A would acquire a stake of 25% in firm B and obtain also “proportional control”, then the MHHI would increase by 780, indicating a stronger effect on concentration and the possibility that competition concerns may arise.74

6.2. Bertrand competition and the Price Pressure Index ("PPI")

(89) The concept of PPI, provides a rough estimate of competitive effects of acquisitions in a Bertrand oligopoly framework with differentiated products.75 Salop and O’Brien (2000) refer to PPIs as indicators that measure the economic pressure to change prices in response to a change in ownership structures. Unlike the MHHI analysis, however, a separate PPI indicates the pricing pressure of each firm in the market. Therefore, a separate delta (Δ) for each firm that is affected by the acquisition has to be computed. The pressure to increase price after an acquisition depends on the reduction in the opportunity cost of raising prices. An advantage of the PPI approach over the MHHI concept is its ability to incorporate efficiency benefits into the analysis in a practical way.

(90) If firm A acquires a passive structural link in target firm B, the pricing pressure of firm A increases in the level of A's financial interest in firm B, in B's margin

71 Salop and O’Brien (2000), p. 597. 72 For simplicity, the merged entity is assumed to obtain the joint market share of both merging

firms. This could e.g. be the case in a capacity driven market. 73 If firm A acquires a passive minority ownership share in firm B, then .

See Salop and O’Brien (2000), p. 595. 74 Formally, if firm A acquires the minority ownership share in firm B and also obtains

"proportional control" in firm B, then . See Salop and

O’Brien (2000), p. 595. 75 The PPI is defined slightly differently compared to the related concept of upward pricing pressure

(UPP) as discussed by Farrell and Shapiro (2010). The latter concept does not express the pricing pressure relatively to the own marginal costs.

22

relative to A's marginal cost and in the proportion of sales diverted from A to B after a price increase of A.76 The diversion ratio, which refers to the proportion of sales diverted from A to firm B after a price increase of A, depends on how "close" the products of the two firms are. Intuitively, B's margin relative to A's marginal cost and the diversion ratio jointly indicate, how much of the profit lost by firm A after a price increase will be recouped by firm B. Acquiring a financial interest in firm B allows firm A to partially internalize the positive effect of its price increase on firm B's profits, thereby leading to an increased pricing pressure of firm A.

(91) Note that the PPI reflects the pricing pressure of a firm when taking its rivals' prices as given. In particular, as in case of a passive minority ownership the target firm's incentives to charge higher prices are unchanged when holding the other firms' prices constant, its PPI remains unaffected as well.77

(92) It is straightforward to apply the PPI when analyzing more complicated acquisitions. For example, when the acquisition of a minority share also confers a degree of influence in the target firm, then the acquiring firm will typically also induce the target firm to raise its prices, which typically feeds back to higher profits of the acquiring firm. Therefore, in this scenario, the PPIs of both the acquiring and the target firm increase.78 The can also be computed for more complicated ownership changes such as joint-ventures.

6.3. Further issues concerning the application of MHHI and PPI

(93) As the OECD (2008) points out, the MHHI methodology is relatively easy to implement, as it is based on information which is normally available to antitrust enforcers (such as the parties’ market shares and the percentage of the ownership acquired), whereas the PPI methodology requires a set of information (such as the profit margins and the diversion ratios for differentiated products) that may be more difficult to obtain.79 Importantly, these quantitative tools do not replace a comprehensive analysis of the factors that may be relevant in a competitive assessment, and as such they only provide an indication of the possible competitive effects of a minority shareholding in a competitor. When assessing the impact that a network of widespread structural links might have in a particular market, indirect effects have to be taken into account when applying the MHHI or PPI concepts.80 For example, when firm A has a structural

76 Formally, if firm A acquires the passive minority ownership share in firm B, then

, where is the diversion from firm A to firm B. See Salop and O’Brien (2000), p. 599.

77 It can be formally shown that the target's equilibrium prices are likely to increase as a reaction to the increased prices of the acquiring firm. However, the PPI is a simpler concept that only captures the direct effect of partial acquisitions.

78 Formally, if firm A acquires the minority ownership share in firm B and also obtains proportional control in firm B, then of firm A is the same as in Footnote 76 while for firm

B, , where is the diversion ratio from firm B to firm A. See

Salop and O’Brien (2000), p. 599. 79 This issue is discussed more extensively in the context of the concept of upwards pricing pressure.

For a summary of the literature on this respect, see Werden and Froeb (2011). 80 More precisely, indirect minority shareholdings have to be taken into account whenever target

firms hold themselves minority shareholdings in rivals - see Flath (1992).

23

link of 20% in firm B which has a structural link of 20% in C, then 4% of firm C's profits indirectly accrue to firm A via its structural link stake in B. When ignoring indirect structural links, the unilateral effects are typically underestimated. Therefore, before the above mentioned concepts can be applied, the effective (i.e. direct and indirect) ownership stake of all firms has to be computed.81 Indirect shareholdings are typically not relevant if investors have minority shareholders in several competing firms.

81 Flath (1992) and Dietzenbacher, Smid, Volkerink (2000) discuss a computational method of

deriving effective ownership stakes in the case of passive minority shares.

24

REFERENCES

Allen, J.W. and G.M. Phillips (2000), "Corporate equity ownership, strategic alliances, and product market relationships", The Journal of Finance, Vol. 55, No. 6, pp. 2791–2815. Alley W. A. (1997), "Partial Ownership Arrangements and Collusion in the Automobile Industry", The Journal of Industrial Economics, Vol. 45, No. 2 (Jun., 1997), pp. 191–205. Amundsen E. S. and L. Bergman (2002), "Will Cross-Ownership Re-Establish Market Power in the Nordic Power Market?", The Energy Journal, Vol. 23, No. 2, pp. 73–96. Baumol W. and J. Ordover (1994), "On the Perils of Vertical Control by a Partial Owner of a Downstream Enterprise", Revue d’économie Industrielle, Vol. 69, pp. 7–20. Bresnahan T. and S. Salop (1986), "Quantifying the Competitive Effects of Joint Ventures", Int. Journal of Industrial Organization, Vol. 4, pp. 155–175. Brito D., Cabral, L. and H. Vasconcelos (2008), "Duopoly Competition with Common Shareholders", Mimeo. Brito D., Cabral, L. and H. Vasconcelos (2010), "Duopoly Competition with Competitor Partial Ownership", Mimeo. Brito, D., Ribeiro, R. and H. Vasconcelos (2013). "Measuring Unilateral Effects in Partial Acquisitions", CEPR Discussion Paper no. 9354, London, Centre for Economic Policy Research, http://www.cepr.org/pubs/dps/DP9354.asp. Bulow J., Huang M. and P. Klemperer (1999), "Toeholds and Takeovers", Journal of Political Economy, Vol. 107, No. 3, pp. 427–54. Che, Y.-K. and J. Sákovics (2008), "Hold-up Problem.", The New Palgrave Dictionary of Economics, Second Edition, Palgrave Macmillan. Cowling, K. and M. Waterson (1976), "Price cost margins and market structure", Economica, Vol. 43, pp. 267–274. Dasgupta, S. and Z. Tao (2000), "Bargaining, bonding, and partial ownership", International Economic Review, Vol. 41, No. 3, pp. 609–635. Dietzenbacher, E., Smid B., and B. Volkerink (2000), "Horizontal integration in the Dutch financial sector", International Journal of Industrial Organization, Vol. 18, pp. 1223–1242. DotEcon (2010), "Minority Interests in Competitors", available at http://www.oft.gov.uk/shared_oft/economic_research/oft1218.pdf. Dubrow J. B. (2001), "Challenging The Economic Incentives Analysis of Competitive Effects in Acquisitions of Passive Minority Equity Interests", Antitrust Law Journal, Vol. 69, pp. 113–145. European Commission (2004), "Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings", OJ C 31 of 5 February 2004, pp. 5–16. European Commission (2008), "Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentrations between undertakings", OJ C 265 of 18 October 2008, pp. 6–25.

25

Ezrachi A. and D. Gilo, "EC Competition Law and the Regulation of Passive Investments Among Competitors", Oxford Journal of Legal Studies, Vol. 26, No. 2, pp. 327–349. Farrell, J. and C. Shapiro (1990), "Asset ownership and market structure in oligopoly", RAND Journal of Economics, Vol. 21, pp. 275–292. Farrell J. and C. Shapiro (2010), "Antitrust Evaluation of Horizontal Mergers: An Economic Alternative to Market Definition", The B.E. Journal of Theoretical Economics, Vol. 10, No. 1. Fee, C.E., Hadlock, C.J. and S. Thomas (2006), "Corporate equity ownership and the governance of product market relationships", The Journal of Finance, Vol. 61, No. 3, pp. 1217–1251. Flath, D. (1991), "When is it Rational for Firms to Acquire Silent Interests in Rivals?", International Journal of Industrial Organization, Vol. 9, pp. 573–584. Flath, D. (1989), "Vertical Integration by Means of Shareholding Interlocks", International Journal of Industrial Organization, Vol. 7, pp. 369–380. Flath, D. (1992), "Horizontal shareholding interlocks", Managerial and Decision Economics, Vol. 13, No. 1, pp. 75–77. Foros, Ø., Kind, H. J. and G. Shaffer, (2010), "Mergers and Partial Ownership", CESifo Working Paper Series No. 2912; Simon School Working Paper No. FR 10-11, available at SSRN: http://ssrn.com/abstract=1539366. Gilo D. (2000), "The Anticompetitive Effect of Passive Investment", Michigan Law Review, Vol. 99, pp. 1–47. Gilo D., Levy N. and Y. Spiegel (2012), "Partial vertical integration", Mimeo. Gilo D., Moshe Y., and Y. Spiegel, "Partial Cross Ownership and Tacit Collusion", RAND Journal of Economics, Vol. 37, pp. 81–99. Gilo D. and Y. Spiegel (2011), "Partial vertical integration in telecommunication and media markets in Israel", Israel Economic Review, Vol. 9, No. 1. Gonzalez-Diaz, F. E. (2012), "Minority Shareholdings And Interlocking Directorships: The European Union Approach", Antitrust Chronicle, Vol 1, No. 1. Güth W., Nikiforakis N. and H.-T. Normann (2007), "Vertical cross-shareholding: Theory and experimental evidence", International Journal of Industrial Organization, Vol. 25, No. 1, pp. 69–89. Harbord D. and K. Binmore (2000),"Toeholds, Takeovers and Football", European Competition Law Review, Vol. 21, No. 2. Hunold, M., Röller, L.-H. and K. Stahl (2012), "Backwards Integration and Strategic Delegation", ZEW Discussion Paper No. 12-022. Ivaldi M., Jullien B., Rey P., Seabright P. and J. Tirole (2003), "The Economics of Tacit Collusion", Final Report For DG Competition, IDEI Toulouse Karle, H., Klein, T. J. and K. O. Stahl, "Ownership and Control in a Competitive Industry", TILEC Discussion Paper No. 2011-013. Kühn, K. and M. Rimler (2006), "The Comparative Statics of Collusion Models", CEPR Discussion Paper no. 5742. Kwoka J. E. (1992), "The Output and Profit Effects of Horizontal Joint Ventures", Journal of Industrial Economics, Vol. 40, pp. 325–338. Martin (1995), "R&D Joint Venture and Tacit Product Market Collusion", European Journal of Political Economy, 1995, vol. 11, pp. 733–741.

26

Salop S. and D. O’Brien (2000), "Competitive Effects on Partial Ownership: Financial Interest and Corporate Control", Antitrust Law Journal, Vol. 67, pp. 559–614. Salop S. and D. O’Brien (2001), "The Competitive Effects of Passive Minority Equity Interests: Reply", Antitrust Law Journal, Vol. 69, pp. 611–625. OECD (2008), "Policy Roundtable concerning Minority shareholdings", Paper DAF/COMP(2008)30, available at www.oecd.org/dataoecd/40/38/41774055.pdf. RBB (2011), "Conjectural Variations and Competition Policy: Theory and Empirical Techniques", available at http://www.oft.gov.uk/shared_oft/research/CV_Competition_Policy.pdf. Reiffen D. (1998), "Partial Ownership and Foreclosure: An Empirical Analysis", Journal of Regulatory Economics, Vol. 13, No. 3, pp. 227–244. Reynolds, R. J. and B. R. Snapp (1986),"The Competitive Effects of Partial Equity Interests and Joint Ventures", International Journal of Industrial Organization, Vol. 4, pp. 141–153. Parker, P. M. and L.-H. Röller (1997), "Collusive Conduct in Duopolies: Multimarket Contact and Cross-Ownership in the Mobile Telephone Industry", RAND Journal of Economics, Vol. 28, No. 2, 1997, pp. 304–322. Perry, M. K. (1989) "Vertical Integration: Determinants and Effects", Handbook of Industrial Organization, Vol. 1, North Holland, Amsterdam. Reitman, D. (1994), "Partial Ownership Arrangements and the Potential for Collusion", The Journal of Industrial Economics, Vol. 42, No. 3, pp. 313–322. Rey, P. and J. Tirole (2007),"A Primer on Foreclosure," Handbook of Industrial Organization, Elsevier. Riordan, M. H., "Ownership without control: Toward a theory of backward integration," Journal of the Japanese and International Economies, 1991, Vol. 5, No. 2, pp. 101–119. Spector, D. (2011), "Merger control and minority shareholdings: Time for a change?", Concurrences, No 3, pp. 14–41. Spiegel, Y. (2011), "Backward integration, forward integration, and vertical foreclosure", Mimeo. Werden, G. and L. Froeb (2011), "Choosing Among Tools for Assessing Unilateral Merger Effects", Vanderbilt Law and Economics Research Paper No. 11-19.