Saying farewell to holding companies and hello to low voting shares

Written with the fond memory of the late great Dr. Jos Gerson – a colleague and warm friend who was the complete expert on Corporate Ownership and Control in South Africa. He is missed – especially at a time like this.

For a taste of the earlier work on these issues see on the blog under Research this publication

Shareholders as agents and principals: The case for South Africa’s corporate governance system Journal of Applied Corporate Finance, 1995 8(1)

 

Saying farewell to holding companies and hello to low voting shares – lessons from the Pick n Pay Holding Company unbundling.

One of the last of the once numerous pure holding companies listed on the JSE, Pick n Pay Holdings Limited (PWK), is no longer with us – to the palpable delight of its shareholders. PWK is a pure holding company because its only asset was a 52.69% shareholding in the operating company, supermarket chain Pick n Pay (PIK), from which it received dividends and paid out almost all of them (after limited expenses) to its own set of shareholders. PWK incurred no debt and acquired no other assets.

Its sole purpose was an important one – at least to its majority shareholder, the Ackerman family, who held over 50% of the shares in PIK and therefore continued to control its destiny with a minority stake in the operating company of 26% (51% of 52% = 26% roughly). These arrangements, sometimes unkindly described as pyramid schemes, enabled founding families of successful listed enterprises in SA (and elsewhere) to attract capital from sources outside the family, without giving up a proportionate degree of voting rights. Family control would be loosened should more than 50% of the listed operating company be publicly owned. But this constraint could be overcome by selling up to 50% to outsiders in a listed holding company with at least a 50% controlling stake in the operating company.

This process of divesting ownership rights without surrendering proportionate control was taken to an entirely legitimate extreme by the Rupert and Herzog families who controlled Remgro. Their concern to maintain control of the operating companies of the large Rembrandt Group led to the formation of four JSE listed holding companies. Top of the listed pyramid was Technical and Industrial Investments Limited with a 60.4% stake in also listed Technical Investment Corporation Limited that held 40.56% of listed Rembrandt Controlling Investments that owned 51. 07% of the listed operating company Rembrandt Group Limited that generated all the earnings and dividends.

Just in case you thought that this did not add up to 50% ownership, the top of the listed pyramid Technical and Industrial Investments Ltd held a further 9.6% of Rembrandt Controlling Investments Ltd. In this way, by inviting outsiders to share ownership in a tier of holding companies, the founding families continued to appoint and control the managers of businesses within the large Rembrandt Group with an ownership stake in it of about 5%. It was not democracy but it was a case of capitalist acts between consenting adults.

Clearly all of the other shareholders in Rembrandt and its holding companies, as those in PIK and PWK, understood fully that by buying shares in the operating or holding companies they would be sharing in the fortunes of the operating company without ever being able to force their collective will on the controlling shareholders. That they were willing to do so was to the great credit of the founding and controlling shareholders. They were trusted by those providing a large majority of the risk capital employed to act in the interest of all shareholders in wealth creation. That the family interests in the operating assets were proportionately small but represented a large proportion of the wealth of the controlling families would be a source of comfort to effectively minority shareholders, in votes if not in claims on dividends or assets.

There are of course simpler ways of separating ownership and control than layers of holding companies. Shares in the operating company with differential voting rights can serve the same function more simply and much less expensively. But these arrangements were until recently effectively prohibited by the listing requirements of the JSE, with the exception of a few grandfathered arrangements such as applied to Naspers with its great majority of non-voting shares. With a change in listing requirements and in SA company law, Rembrandt was able to collapse its pyramids while maintaining control with unlisted B shares and Pick and Pay has broadly, with the enthusiastic approval of its shareholders, followed this example. Family control of PIK is being controlled with family ownership B shares with effectively over 50% of the voting rights in PIK.

The shareholders in PWK who are to receive PIK shares in exchange had every reason to welcome the new arrangements. They, on the announcement of the intention to proceed with the collapsing of PWK and the unbundling of its 52.69% holding of shares in PIK to its shareholders, saw the value of a PWK share increase by over 12% on the day of the announcement.

Shares in PWK, the holding company, had until then always traded at a variable discount to the value of the shares of PWK held in PIK. Or, in other words, the market value of PWK was always less than the market value of the shares it held in PIK. In 1999 this discount was as much as 30%. On 13 June 2016, before the unbundling announcement, the discount was 18.8%. By the close of trade on 14 June it had fallen to 3% after the PWK share price had gained 12.6% on the day while PIK shares lost 2% (see below).

The reasons for this persistent discount, or more particularly why it varied so much over the years, is not immediately obvious. After all PWK was but a clone of PIK. A discount could be justified by the fact that the holding company incurred listing and other expenses as well as perhaps additional STC. Consequently we calculate that PWK shareholders received less by way of dividends than the 52.69% ownership stake in PIK would ordinarily imply. We calculate from the dividend flows paid by the two companies (share price*dividend yield) that PWK received dividends equivalent to roughly 48% of those paid by PIK (see below).
Consequently the dividend yield on a PWK share consistently exceeded that of a PIK share – a lower entry price making up for the lesser flow of dividends (see below).

 

The value of a PWK share in which control of PIK rested may have been boosted (it was not) by the chance that a takeover bid for control of PIK via PWK might have been offered and accepted. Control of PIK would change with a smaller 50% stake in PWK – a possibility that might have attracted a control premium to a PWK share. I recall Raymond Ackerman announcing that any such change in control premium paid for the controlling stake would be shared by all shareholders, presumably in PIK as well as PWK. If so, there would have been no value to be added holding the effectively high voting rights in PWK rather than in PIK. The premium or possible discount that might be paid for the high voting 26% of PIK held by the family controlling interests in the form of B shares, would presumably not now be subject to formal approval by the full body of shareholders.

For all the variable price discount and the higher dividend yield the total returns holding a PIK share rather than a PWK share were very similar over the years. Though until the unbundling shareholders in PIK reinvesting their dividends in additional PIK shares would have enjoyed marginally higher returns than those in PWK. Though as we show below this total return gap narrowed sharply on the unbundling. A R100 invested in PIK shares in 1990 with dividends reinvested would now be worth R3,463 while the same investment in PWK would have grown to R3,397. Excellent results for long term shareholders have been provided by the managers and controllers of PIK, especially when compared to the returns received from holding the much more diversified shares that make up the JSE All Share Index.

 

The outcomes for PIK and PWK shareholders have not been as favourable since 2010, as we show below. With recent share price gains, PIK and PWK returns have matched those of the JSE All Share Index but fallen below those provided by Shoprite (SHP) a strong competitor and by the General Retail Index which does not include PIK and SHP. This helps make an important point. For any business to succeed over the long run, it demands that the constant threat and challenges from competition that emerges in ever changing forms be successfully withstood. This makes the owners of any business, however well established, at significant risk of underperforming or even failure. Owners sacrificing potential returns for less risk may have appeal at any stage of the development of a business.

Judged by these outstanding returns with hindsight, it could be concluded that the Ackerman family interests might have been better served by keeping the company private and not inviting outsiders to share in the company’s significant successes over the years.

Hindsight however is not an appropriate vantage point to make investment decisions. Start-ups, as Pick n Pay once was, are always highly risky affairs. Most start-ups will not succeed in the sense that the returns realised for their owner-managers exceed those they could have realised, taking much less risk working for somebody else.

But when a start-up is a proven success, the incentive for the successful owner-manager to reduce the risks to their wealth so concentrated, by effectively investing less in the original enterprise and diversifying their wealth, becomes an ever stronger one. Risks can be reduced by withdrawing cash from the original business through selling a stake in the business or equivalently by withdrawing cash more gradually from the business in dividends, cash that is then invested presumably in a more cautious, more diversified way. The Ackerman family appears to have followed this route.

An alternative approach is that taken by the Rupert and Herzog families and that is to use the successful business that is the original foundation of their wealth to fund a programme that diversifies their business risks – by investing in a variety of listed and unlisted enterprises that remain under firm family control. And to invite outside shareholders to share in the risks and rewards the family is taking with its own wealth.

Both approaches to building and diversifying wealth can clearly succeed despite (or is it because?) of the concentration of control and the differential shareholding voting rights, this may call for. It is a wise financial system that does not stand in the way of such potentially highly value adding arrangements shareholders might make with each other – that shareholders be allowed to trade off any possibility of a hostile takeover for the benefits of sharing in the rewards of great family controlled enterprises, as the Pick n Pay shareholders have just agreed to.

Leave a Reply

Your email address will not be published.