European Company Law Experts: Peter Böckli, Paul Davies, Eilis Ferran, Guido Ferrarini, José Garrido, Klaus J. Hopt, Alain Pietrancosta, Markus Roth, Rolf Skog, Stanislaw Soltysinski, Jaap Winter, Eddy Wymeersch
The European Commission launched, in April 2014, a new initiative to amend the shareholder rights directive as regards to the encouragement of long-term shareholder engagement.[1] Under this heading, the Commission proposal intends to grant rights to shareholders concerning director remuneration (say on pay) and related party transactions. Moreover, it also imposes duties concerning an engagement policy on institutional investors and asset managers and gives rights to the management concerning shareholder identification.
This paper deals with shareholder engagement and identification by referring to the initial Commission proposal.[2] Both instruments are motivated by referring to the support shareholders have allegedly given to managers’ excessive risk taking before the financial crisis. The current level of “monitoring” of investee companies and engagement by institutional investors and asset managers is considered inadequate, leading to suboptimal governance of listed companies (see preamble 2). It is questionable whether the financial crisis revealed weak governance in listed companies and whetherthe rules proposed are likely to meet the objectives as stated in the directive.
1. Engagement policy
The proposed directive contains a double set of requirements addressed on the one hand to the institutional investors – defined as insurance companies and pension funds[3] – and on the other to the asset managers – here the UCITS, AIFS. These investors are expected to develop, as part of their investment strategy, an engagement policy aimed at improving the performance of their equity portfolio and to undertake specific actions listed in the proposal. Institutional investors should disclose how their investment strategy corresponds to their liabilities and contributes to the performance of their assets.[4]
The action to be undertaken as part of the engagement policy includes monitoring investee companies, conducting dialogues with them, exercising voting or other shareholder rights and cooperating with other shareholders. They should annually disclose their policy and how it has been implemented.[5] The beneficiaries of these institutional investors could receive details on the engagement policy and its implementation on their request. The disclosure should further include the way votes have been cast,[6] including an explanation of the voting behaviour. However for institutional investors, the provisions would only apply on a “comply-or-explain” basis, allowing them not to adopt an engagement policy, and hence not disclose details about its implementation. In that case, a “clear and reasoned explanation” was proposed to be given, indicating how this decision meets the best interest of the beneficiaries.[7]
The purpose of this action would be that shareholders would contribute “to a more long-term perspective… which ensures better operating conditions for listed companies” and this in the perspective of their contribution to long-term financing and economic growth. These requirements would represent a considerable extension of the existing national provisions, as present law often only requires “asset managers” to disclose their “voting policy”.
If the potential for engagement by institutional shareholders is to be realized, we suggest that four issues need to be addressed. First, if shareholders supported excessively risky conduct by management before the financial crisis,[8] is increasing the influence of shareholders over management a solution to this problem or its cause? Second, under what circumstances is it in the interests of beneficiaries that institutional shareholders increase the resources devoted to engagement with portfolio companies? Engagement action should be proportionate to the possible influence they can exercise on the investees’ management and the added value that can be expected from specific engagement actions. An overbroad obligation to engage may result in “contentless” dialogues, resulting in boilerplate disclosures; this should be avoided. Third, disclosure may promote engagement but it may also hinder it. Especially but not only in the case of closely-held companies, there have been powerful and effective undisclosed engagement actions, leading to substantial changes in company structures or practice, where the absence of disclosure facilitated the change in managerial policy. Therefore ECLE pleads for an engagement obligation that effectively contributes to wealth maximization and is implemented in a way proportionate to the expected benefits. Institutional investors, due also to their relatively small holdings and their investment incentives that do not favour gaining intimate knowledge of and interest in their investee companies,[9] are usually not the type of “partners” that one would expect as effectively contributing to targeted engagement. Finally, even if institutional investors may become more interested in active engagement pursuant to their investment policy, they may be deterred from actively engaging due to existing legal restrictions.
Investors involved in an engagement policy are not protected against the negative consequences of their actions: if they cooperate with other shareholders, they may become exposed to the qualification of acting in concert.[10] If they receive privileged information, even inadvertently, they will be prevented from trading, a position potentially in conflict with their fiduciary duties towards their investors. In some cases they may even have to disclose this information, which may be detrimental to the company in which they have invested. Concerted action may also trigger the obligation to launch a takeover bid, even if the bidder, as an institutional investor or asset manager, really had no intention to acquire control. If an effective engagement policy were to be considered, protection against the application of these provisions should first be organized. According to some, such an investor becomes an insider, a qualification to which certain restrictive consequences are attached, such as a trading prohibition; others have suggested that engagement action can only be undertaken within a separate legal structure that would then be run independently and not subject to the restrictions mentioned above. As long as these impediments to effective engagement have not been lifted, one can reasonably fear that engagement will remain at the level of compliance with formal requirements, to little benefit for either party.
The proposed directive sees the engagement policy as a precondition for the exercise of the voting rights. This may be true for engagement that is based on an active dialogue with the company’s management, although engaged shareholders usually do not express their opinion at the moment of voting, but beforehand in a direct dialogue with management. The proposed directive considers essentially the more formal type of engagement that stands for formal compliance with an announced engagement policy and materialises in formal voting expressed by a proxy agent. Yet it should be avoided that by mandating a formal engagement obligation, the directive contributes to mere formal voting by institutional investors. Commanded, uninformed voting might be worse than no voting at all. Another incentive leading to similar outcomes is the mandatory disclosure of votes cast. On a practical level, the obligation for institutional investors and asset managers to disclose, “for each company in which they hold shares”, “if and how they cast their votes in the general meetings of the companies concerned and provide general information and an explanation for their voting behaviour”, seems excessively burdensome, taking into account the very considerable number of holding lines in their portfolios. Indicating thresholds from which disclosure obligations would be applicable might usefully be considered.
Active engagement takes many forms. Account should be taken of differences in ownership structure, where engagement will have a different meaning in controlled companies, especially leading to specific voting procedures for conflicted transactions (see “related party transactions”). In companies with major shareholders, there is often active and continuous monitoring of the management by these shareholders, which is very different and carries considerably more weight than the type of engagement that the proposed directive considers in its description of the “engagement policy”. It often is beneficial to the management of companies pointing towards issues or shortcomings which management itself has not been able to settle. Activist investors exercise a different type of engagement. Their position is largely unmentioned in the proposed directive and one may wonder whether the directive should not pay additional attention to their activity. In companies with dispersed ownership, engagement can only be effective if several shareholders are able to join forces, which is severely limited by present regulations such as insider trading provisions and the danger of being treated as acting in concert.
ECLE concludes that it is advisable that the provisions on the engagement policy be laid down in a Commission Recommendation. Rather than the proposed mandatory “comply or explain” obligation, a recommendation would create more clarity on the nature of the obligation and allow the taking of better account of the considerable factual and national differences.
The proposal calls attention to the specific conflict-of-interest situations in which institutional investors and asset managers may find themselves, citing those cases when the financial services group to which the institutional investor belongs offer financial products or have commercial relations with investee companies. The proposed directive indicates that the engagement policy includes policies relating to these conflicts of interests “with regard to shareholder engagement” and that it should ensure that engagement is undertaken to the exclusive interest of their investors. In some sectors the conflict-of-interest policies are laid down in the legislation: in this case this regulation would then be part of the conflicts policy. Conflicts of interest as an object of the engagement policy for both institutional investors and asset managers relate, according to the proposed directive, to the offering of financial products, entertaining significant commercial relationships with investees, or conflicts arising from the exercise of voting rights, the last case obviously only applying under the condition that sectoral legislation provides so. The scope of this conflicts policy deserves further clarification: for example, is it applicable to the asset manager or to the entire group to which it belongs? Would it include holdings in unlisted investees? May exercising voting rights otherwise than in the interest of the beneficiaries of the institutional investor be allowed in specific cases and under what conditions, and would the same apply to asset managers?
2. Investment strategy and transparency of institutional investors and asset managers
According to the proposal, institutional investors and asset managers have to provide information on their investment activity.[11] The institutional investor has to inform the public about its portfolio investments, how these investments are aligned to the timeline of its liabilities and how they contribute to medium and long-term performance. Further information relates to the incentives or remuneration for achieving these objectives, and how performance is evaluated (including, for long-term absolute performance, the portfolio turnover target). If the information is not disclosed, the institutional investor should explain the reasons. The asset managers acting for institutional investors are held to parallel disclosure duties: they should explain how they comply with the “arrangement” with the investor, provide information on portfolio composition and significant changes in the portfolio, on turnover and costs and on its policy on securities lending and its implementation. Referring to a provision mentioned above, they should also inform of conflicts of interest in their engagement activities and how these have been dealt with. This information should be made available to other institutional investors, professional investors and individual investors, on request. These parties do not have to show a direct financial interest to obtain the information.
One may wonder why this information has been included in this proposed directive dealing with “long-term shareholder engagement”, and not in the directives applicable to the two types of institutional investors viewed. These investors and the asset managers are addressed in specific directives that deal with their disclosure duties. The proposed directive contributes to fragmentation and dispersion of professional duties.
3. Transparency of proxy advisors
The proposed directive deals with proxy advisors (PAs) to the extent that they issue voting recommendations,[12] but does not address the voting instructions. In today’s market structure, proxy advisors are indispensable and their activity is a direct function of the widely spread ownership of equity, mainly in the hands of institutional investors, willing or required to “engage” and or to vote. To the extent that the latter’s voting is de facto or de jure mandatory, recourse to voting agents becomes almost indispensable. For portfolio investors it may constitute a cost-effective way of meeting their mandatory voting obligation, thereby contributing to the company’s governance at a reasonable cost. It is documented that institutional shareholders rely heavily on proxy advisors, especially in relation to cross-border holdings. Proxy advisors usually recommend to vote according to the template that they have adopted on the basis of their internal research for casting votes in most general meetings.
The role of proxy advisors has been the subject of discussion and criticism, some of which is being addressed in the proposed directive. A serious concern stems from the high concentration of this activity – de facto, there are only a handful organisations operating on a worldwide basis and forming a close oligopoly.[13] Moreover, proxy advisors are increasingly setting the standard in corporate governance matters and company decision-making, incurring criticism that they adopt positions without an in-depth knowledge of the different legal regimes or local practices to which companies are subject, or the specific characteristics of the companies concerned. In this way the activities of proxy advisors enhance a formalistic level of shareholder engagement only. In some cases it was reported that a proxy adviser advised against a certain proposal, irrespective of the context in which the proposal was tabled, or sometimes even in disregard of local company law. It was also reported that proxy advisors refuse to change or withdraw recommendations that were based on incorrect factual assumptions after these mistakes were pointed out to them. Furthermore, conflicts of interest were mentioned, with proxy advisers sometimes advising the management of investee companies about subjects submitted to the general meeting where they would cast the votes. Several of these issues have been addressed in the proposal on the basis of a disclosure obligation.
The regime now proposed in the amended directive imposes obligations on proxy advisors. The initial definition[14] as “a legal person that provides, on a professional basis, recommendations to shareholders on the exercise of voting rights” was later extended.[15] One should also consider whether the directive should not provide for a legal framework within which this activity can be exercised and the obligations of the proxy advisers would be verified. Some form of licensing may be discussed, though this may be questionable for proxy advisory firms based outside the EU. In this context special attention will have to be paid to the cross-border recognition of their legal status and activities, particularly in the transatlantic context. Absent a mutual recognition regime, imposing European rules on third country firms that merely issue recommendations and act as proxy for shareholders may prove to be unfeasible and even hinder the development of good corporate governance practices.
The regime proposed in the amended proposal merely refers to the proxy advisors’ obligation to make reference to a code of conduct to which they are subject, or to which they voluntarily adhere, and to identify this or these codes.[16] The codes would be applicable on a “comply or explain” basis, including those cases in which no code has been adhered to. It should further be indicated where and why the proxy adviser has derogated from the code, or what alternatives have been followed. The proxy adviser should, on an annual basis, disclose the methodology that lies at the basis of his positions, mention his sources of information, and whether he had any dialogue with the companies. The annual disclosure should also mention how local market conditions have been dealt with, as well as legal and regulatory requirements (this in response to the criticism voiced above). Furthermore the proposal contains a provision dealing with conflicts of interest, and calls for a policy for the prevention and management of these. Although the text does not make it clear, this provision relates to the conflicts of interest between the proxy advisor and the companies in which they cast votes. Specific disclosure is provided if the conflict would have influenced the voting recommendation.[17]
At least at a first stage it would be preferable to further continue to address this matter in accordance merely with industry codes of conduct, as has been proposed by ESMA after extensive consultation.[18] Over the longer term a more stable, institutional approach should be developed, allowing proxy advisors to exercise their function in a regulatory context containing clear professional requirements, under an appropriate supervisory mechanism and with provision for recognition of non-EU firms. A comparative and empirical study should be undertaken beforehand to identify the actual practices and needs of this important activity.
4. The shareholder identification rules
The rationale put forward for the shareholder identification rules is the facilitation of shareholder voting and engagement more generally. ECLE believes this is a suboptimal mechanism to facilitate shareholder voting, as is set out below. In addition,the rights to be established under the Directive are vested in companies, not their shareholders. This approach creates two risks: first, companies may use the shareholder identification information to make engagement more, not less, difficult. Second, shareholders may not have access to the benefit of these rights when they need them.
Whether companies have the possibility of identifying their shareholders is both a legal and factual question. In several EU jurisdictions, legally enforceable identification rules have been adopted. In some Member States, identification takes place at the initial depository bank, is centralised at the Central Securities Depository and is used for different purposes, such as establishing a dialogue with shareholders, informing them about important ongoing developments and disclosing the reasons for the management’s decisions. Both the management and supervisors are interested in being informed about toeholds being built up. The identification of shareholders may become crucial when the board or the management are being contested by some shareholders, activists or not, whether within the functioning of the general meeting, as part of a proxy contest, or in the context of takeovers when control is being contested.
According to the proposal, Member States should introduce a provision giving companies the right to request that financial intermediaries disclose the identity of their shareholders and the number of shares on the accounts of which the shares are deposited. The decision to require the financial intermediaries to identify the shareholders lies in the hands of the company, and may be varied depending on its ownership structure. The proposal contains some deficiencies: there is no reference to sanctions for failing to inform, no reference to supervisory measures and no reference to current EU work on shareholders’ information and voting platforms.[19]
It has been argued that since identification is seen as an instrument for allowing shareholders to exercise their voting rights, imposing a general shareholder identification obligation for all listed companies may be needed in some, but not all, Member States. A full identification regime would not present much added value, particularly in Member States with concentrated share ownership, as companies are if course aware of their major shareholders. Even in companies with dispersed ownership, the 20 to 25 largest shareholders with whom they regularly interact will be known. These often account for 70 – 85 per cent of the outstanding shares and they frequently take part in the general meeting and exercise their voting rights.
A full identification of all shareholders as is proposed would result in considerable costs, as the identification process (including the identification of the next intermediary in the chain,[20] the information related to the shares[21] and the confirmation of the votes cast[22]), would have to be taken up throughout the entire holding chain and be adapted the same way after each of the numerous transactions in the shares. The administrative burden for all parties involved will be voluminous, expensive, and unlikely to result in more participation from shareholders, in particular (the cost of this machinery may be imposed, inter alia, on the shareholders[23] irrespective whether they are interested in casting their votes or not).
As identification would be mandatorily organised top-down, it is unclear whether individual shareholders could opt out: the proposal provides that the communication will pass through the chain, although the “company (may) directly communicate with its shareholder” (the proposal fails to provide any indication as to how this alternative route would work).[24] The normal method of communication is essentially through the holding chain, possibly locating the final point of data collection at the Central Securities Depository.[25] The CSD will collect the information from the intermediaries in the holding chain, as it has no direct access to the data relating to the individual investors. It would only add an additional layer to the data transmission process.
Furthermore, the transmission of shareholder information – now including data about the amount and value of the assets – throughout the often numerous levels in the holding chain may raise considerable questions of data protection and privacy. This issue has not adequately been considered and no sufficient guarantees have been suggested in the proposal. The general declaration that the information may not be “used in a way incompatible with the purpose of the facilitation of the shareholder rights” will not protect the shareholder against a third party having access to this information and using it for its benefit. The same is true for the provision that there will be no breach of restrictions on disclosure.[26] Moreover, some shareholders may have a genuine interest in keeping their position secret, e.g. in private litigation or arbitration. In addition, some investors may prefer to keep their position confidential when engaging with the company. This information would be accessible to contesting shareholders, as this would relate to the exercise of voting rights.
Moreover, the identification provisions will not prevent the most critical group of shareholders from hiding under different financial constructions, such as derivatives, contracts for differences and other means that have proved effective in the run-up to a takeover. In dispersed ownership companies, identification has not resulted in higher voting participation.
Some consider that companies need to be able to identify their shareholders in order to allow/invite them to exercise their voting rights, and to send them information on the basis of which shareholders can exercise their voting rights. This so-called push approach is also adopted in the explanatory memorandum to the original proposal of the directive (in the paragraph on impact assessment) without, however, much explanation and analysis of alternative approaches. No mention is made of the fact that most company information relevant to voting in the general meeting is generally made available to all investors on the company’s website. They can pull the information from the website without any cost and with minimal efforts. This is the so-called pull approach. There is no need to push information to shareholders to provide them with information when they do not intend to take part in the general meeting, while those that intend to vote can more easily retrieve the information they want to know from the website. It is submitted that active investors already proceed in that way and can contact the company to be kept informed. It might be useful to verify the extent to which companies already make use of this type of facility.
A more efficient way for shareholders to be able to vote may be the following: requiring the financial intermediary to provide to all account holders with an electronic certificate of entitlement by which they can identify themselves as the holders of the shares, deposited in the account at the date specified in the certificate (usually the reference date). Shareholders who wish to take part in the meeting can identify themselves for the meeting or can take part in the voting through the company’s website facility. In order to avoid double voting for the same shares, only shareholders who are not intermediaries could obtain such a certificate of entitlement. Intermediaries such as asset managers should be entitled to vote in their own name, whether as formulated in their general asset management mandate or on the basis of an individual instruction.
There are various internet-based voting systems or platforms available where shareholders can identify themselves as holders of shares and exercise their voting rights; the information is then processed and provided to the company for casting during the general meeting. As all votes are registered electronically, there would be no need for the company to confirm the votes cast (cf. Article 3 c, 2). The company could keep this information available for institutional investors for the benefit of their own disclosures. If this system is in place, it is very doubtful that companies would still need the cumbersome facility of shareholder identification as provided for in art. 3a. Such a solution would also be in line with the green paper issued on building a Capital Market Umion.[27]
To conclude, ECLE sees no need to have a Europe-wide directive with a cumbersome requirement for companies to identify their shareholders, but rather pleads for improving the efficacy of the exercise of voting rights by investors who are at the end of the intermediation chain. This can best be done by a Europe-wide rule requiring the intermediary who has a direct contractual relation with the investor to issue an electronic or written certificate of entitlement that allows the investor to vote at the company’s voting platform on the basis of the shares he owns at the record date. Information on the certificates issued might also be made available to the company itself, giving it a more precise view on its interested shareholders. The company should provide information directly to interested shareholders, in an electronic way.
Finally, the transmission of information by an intermediary[28] to the beneficiary for which he holds shares – in which case he is legally recognised as the shareholder – should take place primarily upon the shareholder’s request. As the notion of intermediary includes EU and non-EU intermediaries, this obligation also applies to the latter. Yet it is unclear how this rule will be made effective, especially for intermediaries that have no establishment within the EU.
If in some Member States companies wish to have a more precise view of who owns their shares (and this is usually for the purposes of identifying actively engaged shareholders), they might be able to do so by reducing the ownership disclosure threshold – at present at 5 per cent according to the Directive 2004/109 of 15 December 2004[29] – in their own their national law insofar as they are allowed to do so under European law. There is no need for introducing Europe-wide legislation and reporting mechanisms to that effect.
[1] European Commission, Proposal for a Directive of the European Parliament and of the Council amending Directive 2007/36/EC as regards the encouragement of long-term shareholder engagement and Directive 2013/34/EU as regards certain elements of the corporate governance statement, Brussels 9.4.2014 COM(2014) 213 final.
[2] For later developments see Interinstitutional File 2014/0121 (COD), inter alia the latest Presidency compromise text, Council of the European Union, 14 January 2015, 5215/15 [Presidential Compromise Proposal] and the Presidency compromise text Brussels, 5 December 2014, 15647/14, the Presidency compromise text Brussels, 10 November 2014, 13758/14.
[3] Article 1, amending the shareholder directive, Article 2 (f).
[4] Amended shareholder directive, proposed Article 3 g (1).
[5] Amended shareholder directive, proposed Article 3 f (3) (1) [Article 3 f (1) Presidential Compromise Proposal].
[6] Amended shareholder directive, proposed Article 3 f (1) (b)[Presidential Compromise Proposal, for each company in which they hold 0.3% of the voting rights].The initial proposal does not contain a threshold such as a block of 0.1 %, 0.25 %, 0.5 % or 1.0 % of the voting shares.
[7] Article 3 (f)(4) added to the initial proposal and skipped in the Presidential Compromise Proposal of 10 November 2014, 13758/14.
[8] Recital 2 of the proposed directive.
[9] Mats Isaksson and Serdar Çelik, Who Cares? Corporate Governance in Today’s Equity Markets”, OECD Corporate Governance Working Papers, No. 8, 2013; OECD Publishing ; Serdar Çelik and Mats Isaksson, Institutional Investors as Owners: Who are they and what do they do?, Corporate Governance Working Papers, no. 11, 2013.
[10] Directive 2004/25/EC on takeover bids but also the Transparency and Acquisition Directive, see ECLE The Application of the Takeover Bids Directive (Response to the Commission’s Report), November 2013.
[11] Articles 3 g and 3 h, the latter reintroduced in the Presidential Compromise Proposal (January 2015).
[12] Defined in Article 1(2)(i) of the initial proposal as “a person that provides on a professional basis, recommendations to shareholders on the exercise of their voting rights”.
[13] Referring to the lack of competition Action Plan 2012, 3.3.
[14] Art 1 (i) amended shareholder directive.
[15] According to the Presidential Compromise Proposal, proxy advisor means a legal person that analyses, on a professional basis, the corporate disclosures of listed companies with a view to informing investors’ voting decisions by providing research, advice or voting recommendations that relate specifically to the exercise of voting rights.
[16] See Best Practice Principles for Providers of Shareholder Voting Research & Analysis, Best Practice Principles Group, March 2014.
[17] See Article 3 i (3).
[18] http://www.esma.europa.eu/system/files/2013-240.pdf
[19] Report of the Reflection Group on the Future of EU Company Law, Brussels 5 April 2011, 3.1.5.
[20] Article 3 a (2) (b) [Presidential Compromise Proposal].
[21] Article 3 b (1) and (2) and (3).
[22] Article 3 c (2).
[23] Article 3 d (2).
[24] Article 3 b (1).
[25] Referring to the intermediaries in the holding chain Article 3 a (2)(a) [Presidential Compromise Proposal].
[26] Article 3 a (4).
[27] European Commission Green Paper, Building a Capital Market Union, Brussels, 18.2.2015, COM(2015) 63, 4.3: technology.
[28] The intermediary is defined in Article 2 (d) amended Directive.
[29] This has not been changed by Directive 2013/50. Under the proposal, the reporting threshold for institutional investors and asset managers about how voting rights have been exercised, with explanation as to their voting behaviour, will be 0.3%: see Article 3 (f)(1) (b) [Presidential Compromise Proposal, initially 0.1 %].
European Company Law Experts (ECLE) 