Lessons in control for the tech IPOs

By John Plender

As Google vies with Microsoft to secure second place after Apple in the market capitalisation stakes, few investors give a thought to Google’s two-tier voting structure. Maybe they should. Recent initial public offerings of technology companies with capital structures that leave founders with a majority of the voting stock, but a much smaller portion of the economic value, have been decidedly mixed.

LinkedIn admittedly has done well. By the end of August its shares were up 138 per cent from its IPO last year. Yet Facebook was down 52.5 per cent since its IPO, while the comparable falls at Zynga and Groupon were 72.0 per cent and 79.3 per cent. Could such volatility have anything to do with the capital structures of these companies? If the arguments for multi-tier voting and enhanced director election rights have any merit, the answer ought to be no. Reinforcing control in this way is said to permit companies to avoid short-term pressure to forego investment in research and development, new products and so forth. In the media, where multi-tier voting rights are common, the structure is also supposed to protect editorial values from vulgar capital market pressure.

The counter argument is that restrictive control rights of this kind encourage lax corporate governance and poor management accountability – failures that have been all too visible recently in the hacking scandal at Rupert Murdoch’s News International. Who is right?

Fresh light is cast on the issue by a new study conducted by Institutional Shareholder Services for the Investor Responsibility Research Center Institute in the US, which looks at 114 closely controlled companies in the S&P 1500 Composite Index*. These are defined as companies in which one holder has at least 30 per cent of the voting shares. In terms of total shareholder return the report shows controlled companies with multi-class voting outperform dispersed ownership companies over one year, but underperform over three and significantly underperform over five and 10 years. This is the opposite of what multi-class advocates would have us believe. Controlled companies with one share, one vote structures significantly outperform those with dispersed ownership or multi-class voting except over one year.

On risk, the report finds that controlled companies with single class voting show less volatility than both multi-class voting and dispersed ownership over one, three, five and 10 years, while dispersed ownership is less volatile than multi-class over all periods.

The report also throws up evidence that material weaknesses in internal control and related party transactions are more likely to arise at controlled companies than with dispersed ownership. Perhaps predictably, a sample of institutional investors felt that controlled owners were less responsive and less open to shareholder engagement than others.

The results markedly underline the point that control rights matter. There is a clear hierarchy from controlled companies with single voting, down to dispersed ownership companies and on to controlled companies with multi-class voting at the bottom of the performance league table. This makes sense. The best performers are those with skin in the game, operating with a fair amount of accountability to outside shareholders. They deliver higher returns for less risk. The worst are those with total protection from shareholder pressure, who deliver the lowest returns for the highest risk. Dispersed ownership delivers something in between.

This also fits with experience in continental Europe where the controlling minority shareholder model delivers patchy results. Sometimes it works well; at other times badly. The point is that outsiders are dependent on the goodwill of the dominant shareholder to allow them to share in the higher returns that stem from the tighter ownership control exercised by management. It would be good to see a comparable piece of research in Europe to confirm that picture.

What are the lessons from all this? One is for the regulators. The ISS researchers found that the average level of board independence was much higher at non-controlled companies, partly because exchange listing standards allow controlled firms to have a majority of non-independent directors. There is an obvious case for the listing authorities to insist on more independent representation at multi-class companies to provide a greater check on controlling directors.

For institutional investors the question is whether to have a policy against backing multi-tier companies. I would argue for a presumption against them, rather than a rigorous rule. In technology the stakes, in terms of market value, are very high. And there are often exceptions that prove the rule. The vital thing is to recognise that control matters.