Edition: November 2018/January 2019
When no means nothing
Two-tier structure defeats one shareholder, one vote. A select few make the rules on executive pay. Intervention needed.
Thanks to Naspers for having so effectively highlighted the contorted face of corporate control. As the giant of JSE listings, where its shares are heavily weighted in investment portfolios, it should be the role model for standards of governance.
That’s some role model when its dual-share structure allows a minority of shareholders to override the majority, most significantly on remuneration. It makes a mockery of stakeholder democracy and the principle of shareholder equality implicit in the Companies Act; for instance, on the election by shareholders of directors. The provision refers to shareholders, not a selective consortium of shareholders with multiple votes.
To its belated credit, Naspers’ latest integrated report is at pains to explain the remuneration policy in intricate and individual detail. To its ongoing discredit, Naspers can proceed to implementation despite a 57% majority of ‘N’ ordinary shareholders having voted against endorsement of the policy. The ‘A’ shareholders, who control the company by having 53% of the votes, predictably carried the day.
Naspers does not reveal the identities of ‘A’ shareholders. It may safely be presumed that they exclude pension funds. By contrast, the JSE-listed ‘N’ shares are substantially held by pension funds; if not directly in their own names then indirectly through asset managers. With index trackers, Naspers can touch almost 30% of a single portfolio.
A consolation for pension funds and other ‘N’ shareholders, aside from the stratospheric effect of Chinese internet company Tencent on the Naspers share price, is that ‘A’ shareholders receive one-fifth of the dividends to which the ‘N’ shareholders are entitled.
The secret ‘A’ shareholders approve the board and the rules it makes. So it’s okay then for chief executive and director Bob van Dijk to receive his R32m in total compensation for the year. There’d be different views on whether this is a little or a lot relative to peers – a lot when compared to some of the most highly paid in SA, a little against say the $12,8m for Apple chief executive Tim Cook -- but the real kicker is in the share options.
In whichever way they’re sliced and diced, allowing for claw-backs and vesting periods – Van Dijk looks in line for around R1,6bn through the gamut of different allocations. The amount could be less or more, depending on the share price when the options are exercised. At face value, as an absolute number and in terms of SA’s social sensibilities, R1,6bn is gobsmacking.
Prominent amongst shareholders at the Naspers agm who voted against the remuneration policy endorsement was Old Mutual Investment Group. It argued that that the total number of ‘N’ shares reserved for equity compensation is excessive. Also, long-term incentives vest (some after only one year) without any conditions for performance.
OMIG additionally voted against the placement of authorised but unissued under the control of the directors: “The dilution exceeds 5% and could involve the issue of new ‘A’ ordinary shares, which have multiple voting rights, and therefore perpetuate the company’s dual-class structure.” For similar reasons, it opposed a board authorisation to issue shares for cash.
Another was Perpetua Investment Managers where chief executive Delphine Govender pointed out that Van Dijk took over as Naspers chief executive in 2014. A theoretical share price of Naspers that excludes its holding in Tencent (assuming a 31% holding) shows that the rump businesses of Naspers have been destroying value especially since 2014.
She explains that effectively Naspers has only performed as a share since 2014 because of its Tencent holding in which current Naspers management has no say: “Current management should therefore be incentivised only on the performance of Naspers excluding Tencent.”
Controversies create opportunities for review of established practices. Naspers represents a prime example, several impacting potentially on the King governance code and the JSE listings requirements as well as the Companies Act:
First is whether there’s justification for dual-class voting structures to be perpetuated. In the US, where these structures are proliferating, two Democrat members of the Securities & Exchange Commission have criticised one class of shareholders getting one vote per share while another class (typically founders in an initial public offering) get multiple votes per share.
Commissioner Kara Stein said that dual classes may “provide a means to evade management and board accountability”. Commissioner Robert Jackson described them as a way of maintaining “corporate royalty”. His suggestion that stock exchanges require “sunsetting” of dual classes was promptly endorsed by the Council of Institutional Investors.
Naspers justifies its unlisted ‘A’ shares -- whereby each shareholder gets 1 000 votes – on grounds that this structure (not unique in other jurisdictions) safeguards the group’s independence. Independence from whom or what? Also, is there good reason why the Companies Act still allows memoranda of incorporation to entrench shareholder inequality?
Second is whether, in SA, shareholder votes on remuneration policy should remain advisory or become binding. In the forefront of arguing for the latter, as part of a process for the redress of inequalities, is Jon Duncan of OMIG.
He’s pointed out that “enhancement” of shareholder say-on-pay, by introduction of a binding vote, will bring SA into line with international practice. In the UK, for example, listed companies must submit their remuneration policy to a binding shareholder vote (simple majority) once every three years.
Whilst he applauds the mandatory disclosure that applies in SA, he considers disclosure alone to be insufficient for holding management to account “for ensuring alignment between executive remuneration practices and the creation of shareholder and broader long-term stakeholder value”. This would be especially so, he might have added, where a minority of shareholders can outvote the majority.
Third is whether the vote against a remuneration policy is sufficient in itself. It’s easier to signal what a shareholder disfavours than favours. Merely to have voted against, especially when it doesn’t follow an engagement with the company on the packages the shareholder will support, carries a purely negative impact.
Fourth is in fiduciary responsibilities. When an asset manager votes at a company meeting, it acts as a shareholder without fiduciary responsibility to the company. But it does have a fiduciary responsibility to the pension-fund clients it represents. In turn, the funds’ trustees have a fiduciary responsibility to the fund members.
Where the fund has not instructed the asset manager to vote in a particular way at a particular meeting, the least to be expected of the trustees is that they insist on explanations from the manager on how it voted on their behalf. And if it didn’t vote, then why.
Fifth relates to research. Larger institutions and funds have the resources for it. Perhaps it can be shared with the smaller, or be publicly pooled, to assist decision-making in advance of company meetings. Many of the sharper asset managers openly discuss their shareholder engagements and proxy votes in stewardship reports, but typically these are after company meetings.
Sixth is in collaboration, “collusion” being a dirty word. An asset manager engaging with a company, speaking for its own smallish stake, will have less clout than a combination of asset managers cumulatively holding a much bigger stake. In a 2003 ruling, still uncontroverted, the old Securities Regulation Panel approved the circumstances of asset managers acting “in concert” to vote at companies.
Seventh is in interest conflicts. These can arise, for instance, where the impartiality of an asset manager is constrained by a controlling shareholder loath to jeopardise other corporate services. They could also arise were the interrogated company to throw back at the manager its own remuneration levels or those of its parent.
Eighth is in disclosure of proxy votes. Some managers do it well, inclusive of reasons. Others do it in ways difficult to access or not at all. The managers of index-tracker funds, especially the one or two most vociferous in their purported support of stakeholder activism, take note. Passive investment is no excuse for passive behaviour, right town to silence on proxies.
All this could be sufficient to keep drafters of the next King iteration busy for a while.
John Plender, an authoritative Financial Times columnist, has warned that remuneration packages risk reaction when they lose their moral and social moorings. He argues that it’s people’s perception of their relative position that matters: “Recent boardroom pay episodes (in the UK) have reinforced the perception that business is run by a greedy, self-serving plutocracy.”
He considers that two-tier voting structures protect founding entrepreneurs from effective oversight, encouraging a hubristic culture: “Amid all this potent governance arcana, a substantial reform agenda clearly needs to be addressed.”
In SA, President Cyril Ramaphosa has told trade unions how he expects them to address it (see article elsewhere in this TT edition). Prepare for action to follow the applause.
To become a pure technology play globally, Naspers is to unbundle video-entertainment operation Multichoice. Between the cracks falls Naspers subsidiary Media24 that houses the group’s newspapers and magazines in print and electronic formats.
Although tiny in the Naspers scheme of things, but not on the local media scene, its retention under Multichoice carries an important principle.
As elsewhere in the world, not least in SA, the viability of iconic titles is threatened. Alone amongst SA media groups, Naspers has cross-subsidised is news business from its TV business.
This is true to the foundational concept that launched M-Net, with rounds of special concessions in the mid-1980s, when the National Party government granted the licence for pay-TV to a Naspers-led consortium of newspaper proprietors. The rationale was to cushion print publications from the advertising revenue they were expected to lose.
After M-Net came MTN and then Multichoice. While other proprietors gradually withdrew, to their detriment as the chickens have come home to roost, Naspers remained consistent.
The history retains relevance (TT Feb-April). It provides the context, notwithstanding the share-price discount, for Naspers to resist pressures for closures in its squeezed publishing operations. A reminder of Naspers’ reciprocation for its government-backed kickstart is in order.