A good influence? Investor liability for antitrust infringements by companies in which they invest

A recent ruling of the EU Court of Justice¹ has important implications for investors – in particular, their potential liability for antitrust infringements committed by companies in which they invest, including where investments are held via fund structures and are not wholly owned investments. Key points to note are:

• It is well-established under EU case-law that there is a rebuttable presumption that a parent company exercises actual decisive influence over a subsidiary if it owns all or almost all of the shares in that subsidiary.  If this is the case, the parent can be held jointly and severally liable if the subsidiary infringes EU competition law.
• The EU Court of Justice has now confirmed in its recent judgment that a rebuttable presumption of actual decisive influence also applies where a parent owns all of the voting rights in a subsidiary (even if it does not hold all the associated shares).  This level of control is akin to the situation where a parent owns all or almost all the shares in a subsidiary.
• These presumptions apply equally to institutional investors as other investors, including where shares are held via funds. And while the presumptions are rebuttable, successfully rebutting them has proven to be very difficult.
• Even if the presumptions do not apply, decisive influence (and therefore joint and several liability) may still be established based on evidence of the economic, legal and organisational links between the parent and the subsidiary. A range of factors may be considered including board representation, informal links between the subsidiary’s directors and the parent, shareholder meetings, access to information etc.
• Joint and several liability for a subsidiary’s infringement not only means the parent is jointly and severally liable for the subsidiary’s fine, but may also increase the level of fine – given fines are based on parties’ turnover, up to a maximum of 10 per cent of their group worldwide turnover. It also means joint and several liability for any damages, which often exceed fines and are now a concrete risk in Europe (especially in the UK, Netherlands and Germany) – and institutional investors (with “deep pockets”) are an attractive target for claims.
• Institutional investors therefore cannot simply assume that competition law infringements by companies in which they invest are nothing to do with them. Careful analysis is recommended to identify investments where the level of interest is likely to amount to decisive influence or the presumption of this. Investors should also ensure that relevant subsidiaries have effective competition law compliance programmes.

We provide further detail about the case below, including the facts and core arguments raised before the Court of Justice that were unsuccessful.

The facts of the case

The case concerns the European Commission’s April 2014 decision that Goldman Sachs (the Bank) was jointly and severally liable for the involvement of the Prysmian group (comprising Prysmian SpA and its wholly owned subsidiary, Prysmian Cavi e Sistemi Srl, formerly Prysmian Cavi e Sistemi Energia) in the power cables cartel.² A previous appeal to the EU General Court was dismissed in July 2018, leading to this further appeal to the EU Court of Justice, itself dismissed on January 27, 2021. Key facts are:

• Between July 29, 2005 and January 28, 2009 (the relevant infringement period), the Bank was the indirect parent, through four funds and other intermediate companies, of Prysmian SpA (and therefore Prysmian Cavi e Sistemi Srl) – one of the leading businesses globally regarding submarine and underground power cables.
• The Bank initially held 100 per cent of Prysmian’s shares but, following divestments, this fell to 91.1 per cent on September 7, 2005 and 84.4 per cent on July 21, 2006, with the Bank continuing to hold 84.4 per cent until an IPO of some of Prysmian’s shares on May 3, 2007.
• The European Commission presumed that: (i) Prysmian exercised decisive influence over Prysmian Cavi e Sistemi Energia during the relevant infringement period; and (ii) the Bank exercised decisive influence over Prysmian, and therefore also Prysmian Cavi e Sistemi Energia, between July 29, 2005 and May 3, 2007 (i.e. prior to the IPO).
• The Commission also concluded that the Bank actually exercised decisive influence over the Prysmian companies during the entire relevant infringement period (i.e. July 29, 2005 to January 28, 2009) – based on its economic, organisational and legal links with those companies.

The Bank appealed on two grounds – one related to the “pre-IPO period” (i.e. July 29, 2005 to May 3, 2007) and the other related to the “post-IPO period” (May 3, 2007 to January 28, 2009).

The pre-IPO period

In relation to the pre-IPO period, the Bank submitted that: (i) its interest in the relevant funds was only around 33 per cent, with the balance owned by unaffiliated third party investors; and (ii) except for the first 41 days, those funds held around 91 per cent and then 84 per cent of the Prysmian companies.  In this context, the Bank argued that the presumption of decisive influence only applies where a parent owns all or virtually all the capital of its subsidiary – so its interest was insufficient for the presumption.

The EU Court of Justice rejected this argument – the Commission’s presumption was not based on the level of the Bank’s indirect holding in Prysmian’s capital. It was instead based on its finding that the Bank controlled all the voting rights associated with Prysmian’s shares throughout the pre-IPO period (which was not disputed).

It is settled EU case-law that: (i) a subsidiary’s conduct may be imputed to its parent, in particular where the subsidiary does not independently decide its conduct on the market and instead in all material respects carries out its parent’s instructions, having regard to the economic, organisational and legal links between them; and (ii) where a parent holds (directly or indirectly) all or almost all the capital in a subsidiary which has infringed EU competition law, the parent is able to exercise decisive influence over the subsidiary, and there is a rebuttable presumption it does in fact exercise such influence (the EU case-law also indicates it is difficult to rebut this presumption in practice).

The Court of Justice held it is apparent from the case-law regarding the rebuttable presumption of decisive influence that applies where a parent holds all or almost all the capital in a subsidiary that it is not merely the holding of this level of capital that gives rise to the presumption but the degree of control that this implies. The General Court was therefore entitled to find that a parent company that owns all the voting rights associated with a subsidiary’s shares is in a similar situation to one owning all or virtually all of a subsidiary’s share capital – i.e. it is able to exercise decisive influence over the subsidiary.

The Court of Justice also dismissed a number of other arguments – including that legal certainty might be undermined by it possibly being more difficult to identify persons holding voting rights associated with shares as opposed to the persons to whom shares belong. The Court of Justice commented that a company which, without holding all or virtually all of a subsidiary’s share capital, reserves for itself or acquires all voting rights cannot be unaware of this. The Commission is also not bound to rely exclusively on the presumption – it could rely on other evidence alone or together with the presumption.

The Bank argued that the General Court had failed to assess correctly the evidence it had adduced to rebut the presumption of decisive influence.³ However, these arguments were inadmissible before the Court of Justice – because it was not alleged the General Court had distorted any relevant facts or evidence. But even to the extent those arguments could be admissible, the Court of Justice noted that where arguments had been dismissed for a lack of supporting evidence, the General Court was not imposing a probatio diabolica (i.e. “devil’s proof” – impossible proof) as had been argued. Instead, this merely reflected that the burden of proof to rebut the presumption lay with the Bank (and arguably demonstrates the hurdle is high).  Similarly, public statements of independence by Prysmian’s board were legitimately rejected because they were not concrete evidence in this regard.

The post-IPO period

The Bank argued that the IPO was a “watershed moment” – Prysmian had a duty of transparency vis-à-vis the market from May 3, 2007 and its funds held only around 46 per cent of Prysmian’s capital, then 26 per cent on November 12, 2007. The finding that the Bank was jointly and severally liable for Prysmian’s infringement during the post-IPO period was based solely on evidence of the links between them indicating that the Bank actually exercised decisive influence (i.e. not a rebuttable presumption that it did so). The Bank sought to challenge this evidence – essentially relating to board appointments or changes, and control over voting rights or a majority shareholding at shareholder meetings.

The Court of Justice held it is settled case-law that in examining whether a parent can exercise decisive influence over the market conduct of its subsidiary, account must be taken of all relevant factors relating to the economic, organisational and legal links between them and the economic reality. The actual exercise of decisive influence in this regard may be inferred from a body of consistent evidence – even if some of that evidence in isolation is insufficient to establish such influence. This requires an assessment of facts contemporaneous with the infringement period, but can also include elements relating to a prior period if relevant and provided conclusions relating to such elements are not applied automatically. Certain evidence relied on by the European Commission related to both the pre-IPO and post-IPO periods, but the Court of Justice rejected arguments that the General Court’s analysis of that evidence had failed to distinguish between the two periods or had reversed the burden of proof.

The Bank also argued the General Court was wrong to conclude that it had sufficient representation on Prysmian’s board to influence its market conduct – including that the General Court was misconceived in relying on the fact that the board’s composition remained the same after the IPO, the Bank appointed only a minority of the board (three “PIA Employee Directors” out of ten directors in total) whereas board resolutions required a simple majority, and after the IPO those directors were prevented from acting solely or primarily to the benefit of other parties (including the Bank). The General Court had also concluded that two of Prysmian’s non-executive directors had links to the Bank (through previous advisory or consultancy services) and even if they were taken into account (which the Bank disputed) the Bank directors made up only half the board.

The Bank again also challenged the General Court’s failure to take account of declarations by Prysmian’s board confirming its independence, arguing this reversed the burden of proof and the declarations risked civil, administrative and potentially criminal penalties under Italian law if inaccurate. Other arguments by the Bank included that powers delegated to the PIA Employee Directors before the IPO, their appointment to Prysmian’s Strategic Committee after the IPO, receipt of regular updates and monthly reports and a number of other measures the General Court considered in finding actual decisive influence were not in fact evidence of this.

The Court of Justice also dismissed these arguments – including again on the basis of case-law that actual exercise of decisive influence may be inferred from a body of consistent evidence, even if some of the evidence in isolation is insufficient.  There is also existing case-law that the existence of an economic entity formed by a parent and its subsidiary may be based not only on their formal relationship but also on informal relationships, including personal links (and previous advisory or consultancy services can be relevant). The General Court therefore did not make a mistake in holding that such personal links could in principle be relevant, and the Bank had not succeeded in undermining the General Court’s conclusion that they were a relevant factor. Also, the mere fact that Prysmian’s board had evaluated that certain directors were independent or published this in corporate governance reports did not preclude the General Court concluding that those directors retained links with the Bank.

Prysmian’s role in the appeal

Another interesting aspect regarding the appeal is that Prysmian intervened in support of the European Commission. This might seem unexpected, especially given some of the evidence relied on to establish that the Bank exercised decisive influence over Prysmian cast doubt on previous statements by Prysmian about its independence. However, this can be explained by the risks related to follow-on damages actions – it is in Prysmian’s interests to ensure that the Bank is jointly and severally liable for its infringement, including any damages regarding the same.