The shareholder empowerment movement (the “movement”), driven primarily by public pension funds and union-related funds with over $3 billion in assets, has renewed its effort to eliminate, restrict, or at least discourage companies from creating dual class share structures in initial public offerings (IPOs). The impetus was the issuance of non-voting stock in the recent Snap Inc. IPO. Such advocacy, if successful, would not be trivial, as many of our most valuable and dynamic companies, including Alphabet (Google) and Facebook, have gone public by offering shares with unequal voting rights.
The movement’s vigorous response to Snap Inc.’s hugely successful IPO was unsurprising. The Council of Institutional Investors (CII), the trade organization that has represented the movement since its founding in 1985, has promoted as a bedrock principle a “one share, one vote” policy. Dual class share structures clearly violate this policy and are an obvious threat to the power of the movement. That is, a public company that provides control to insiders through a dual class share structure can more easily resist the demands of the movement.
In my paper, A Private Ordering Defense of a Company’s Right to Use Dual Class Share Structures in IPOs, I rely on Zohar Goshen and Richard Squire’s newly proposed “principal-cost theory” to argue that dual class shares in IPOs is a value enhancing result of private ordering, making the movement’s renewed advocacy unwarranted.
An Overview of the Argument
Typically, a company that creates a dual class share structure — multiple classes of stock with unequal voting rights — in an IPO will issue a class of common stock to the public that carries one vote per share (ordinary shares), while reserving a separate class, a super-voting class, that provide insiders with at least 10 votes per share. However, both types of shares will have equal rights to the cash flow of the company. Snap’s IPO was an exception to the rule, because it created three classes of stock, reserving super-voting shares for the founders and ordinary shares for early investors and only offering non-voting stock to the public.
In the U.S., only a relatively small number of IPOs have used dual class shares: 27 of 174 in 2015 and 36 of 292 in 2014. Over half of those IPOs were in the technology sector, and they tended to have the largest market values, as was the case with the Snap IPO.
Such IPOs can create a widening gap between voting and cash flow rights. This gap is called the “wedge.” For example, in 1978, almost 40 years ago, the Roberts family held 42 percent of all equity shares of Comcast Corp., but only 0.4 percent in 2015. Yet, through a family trust holding super-voting shares, Brian Roberts, the current and very successful Comcast Chairman and CEO and son of the co-founder, effectively controls the company with 33.33 percent voting power.
Lucian Bebchuk and Kobe Kastiel find this aspect of dual class share structures extremely troubling, arguing that even if the structure is efficient at the time of the IPO, “the potential advantages of dual-class structures (such as those resulting from founders’ superior leadership skills) tend to recede, and the potential costs tend to rise, as time passes from the IPO.”
To remedy the problem of the growing wedge and potentially receding advantages of superior leadership skills, they recommend that all dual class share structures be required to include sunset provisions that obligate the structure to expire “after a fixed period of time (such as ten or fifteen years) unless their extension is approved by shareholders unaffiliated with the controller.”
Yet investors have seemed very willing to ignore the potential issues that are associated with the dual class share structure, investing in ordinary shares with no voting rights and barely any sunset provisions, such as those issued in the Snap IPO, which was priced at $17 per share, giving it a market valuation of roughly $24 billion. The book was more than 10 times oversubscribed and Snap could have priced the IPO at up to $19 per share.
Why would investors be so willing to tolerate a corporate governance arrangement that so obviously increases the potential for agency costs? There are two reasons: the wealth maximizing efficiency that results from the private ordering of corporate governance arrangements, and the understanding that agency costs are not the only costs of governance that need to be minimized.
Private ordering is considered efficient and desirable because it allows for the implementation of market-driven corporate governance arrangements. That is, it “allows the internal affairs of each corporation to be tailored to its own attributes and qualities, including its personnel, culture, maturity as a business, and governance practices.” In effect, “observed governance choices are the result of value-maximizing contracts between shareholders and management.”
Through the bargaining process, the dual class share structure arises. This process requires informed investors to consider the value of having insider control and the costs of shareholder participation in corporate decision making, not just agency costs. It is best understood in the context of Zohar Goshen and Richard Squire’s “principal-cost theory,” which posits that a firm’s optimal corporate governance arrangements result from the minimization of total control costs, as defined below:
[E]ach firm’s optimal governance structure minimizes the sum of principal costs, produced when investors exercise control, and agent costs, produced when managers exercise control. Both principal costs and agent costs can arise from honest mistakes (which generate competence costs) and from disloyal conduct (which generate conflict costs)….
A firm that seeks to maximize total returns will weigh principal costs against agent costs when deciding how much control to allocate to managers and how much to restrict the power of investors to hold the managers accountable.
The need to protect the value of authority is greatest, and agent competence costs are lowest, when insiders, such as founders, possess an “idiosyncratic vision” of the company’s future at the time of the IPO. Dual class share structures, by giving insiders voting control of the company even when they do not own a majority of the common stock, “provides the entrepreneur with maximum ability to realize her idiosyncratic vision” and appears to be an explanation for why this structure has been bargained for at Snap (Evan Spiegel and Robert Murphy) and also at firms such as Alphabet (Larry Page and Sergey Brin) and Facebook (Mark Zuckerberg). The protection of this idiosyncratic vision more than compensates for an expected increase in agent conflict costs.
From the perspective of market participants, Bebchuk and Kastiel’s expectation of an erosion of leadership skills and a significant wedge, respectively, may not occur for many years or decades, if ever. Examples are the long-term success of the Roberts family at Comcast and Warren Buffett at Berkshire Hathaway, the buyout of LinkedIn by Microsoft soon after its IPO, and the voluntary elimination of super-voting stock by the Ratner family at Forest City Realty Trust, Inc., referred to as a mid-stream conversion, when the family decided that such a conversion would be wealth enhancing for them and the other shareholders. Moreover, the odds of an “elderly leader problem” — where the elderly leader is no longer competent to hold voting control — developing at a company like Snap look extremely remote. For better or worse, unexpected events unrelated to having a dual class share structure can affect Snap and its young leadership in the many years between now and when this issue may develop, making the risk of having such a result very low.
Of course, the pricing of an IPO is central to the private ordering process and optimal corporate governance arrangements. Creating a dual class share structure when it does not minimize total control costs harms both insiders and the company, creating an incentive to implement such a structure only when it is truly value enhancing for all parties. That is, the offering price and future trading price of the stock will need to be lowered if the optimal corporate governance arrangements are not implemented up-front. Conversely, the pricing of an IPO will be highest when total control costs are minimized. Perhaps this is why IPOs with dual class share structures are still the exception to the rule, as the market provides sufficient incentives and penalties to make sure they are used only when minimizing total control costs.
Conclusion
While there are excellent arguments for incorporating sunset provisions and similar terms in dual class share structures created through private ordering, academics and shareholder activists overreach when they try to interfere with sophisticated investors — those that take the financial risk of buying stock in IPOs — in the structuring of corporate governance arrangements. Obviously, sophisticated players in an IPO with dual class shares can read the latest articles on their strengths and weaknesses, including the excellent new article by Bebchuk and Kastiel, and incorporate that information into the bargaining process without being dictated to by outside parties.
This post comes to us from Bernard S. Sharfman, who is an associate fellow at the R Street Institute, a member of the Journal of Corporation Law’s editorial advisory board, a visiting professor at the University of Maryland School of Law (Spring 2018), and a former visiting assistant professor at Case Western Reserve University School of Law (Spring 2013 and 2014).