A Comparative Analysis of Dual Class Share Structures

The efficiency of dual class share structures is controversial, and whether to allow them is a difficult choice. Though much has been written about this topic, no comprehensive picture of dual class structures’ governance effects has emerged.

Although dual class structures may take many forms, their key characteristic is that some shares, per unit of their cash flow rights,[1] effectively give their holders more voting rights than other shares do. For purposes of illustration, consider the following example. A company issues two classes of shares: “A” shares, with one vote each, issued mainly to outsiders, and “B” shares, with multiple votes each, generally held by corporate insiders. However, an individual “A” share and “B” share represent an equal piece of economic ownership of the company.

Dual class structures’ governance effects can be sorted into three layers:

Layer 1. Dual class structures have the governance effects of a controlling shareholder (CS) structure. In layer 1, I compare CS structures with those in which stock ownership is dispersed and there is no controlling shareholder (NCS).

Layer 2. Dual class structures also have the governance effects of structures that create voting control without a corresponding right to receive greater financial benefits. In layer 2, I compare such voting-cash flow rights separation with concentrated ownership structures in which one shareholder (or group) owns enough equity to have voting control (under the one share-one vote principle) and the corresponding right to financial benefits.

Layer 3. Finally, dual class structures have governance effects that differ from other control-enhancing mechanisms (CEMs) that also create voting control without correspondingly greater cash flow rights. Such CEMs include pyramids, cross-shareholdings, and securities derivatives in the market for corporate votes. In layer 3, I compare dual class structures with other CEMs.

In companies with public investors, the separation of ownership and control causes agency problems. Corporate controllers[2] may devote insufficient efforts to company affairs or pursue goals other than shareholder value (“shirking”). They may even transfer corporate resources to their own pockets (“tunneling”). NCS structures tend to keep controllers in line through outside shareholder monitoring and the threat of proxy contests and hostile takeovers, whereas CS structures weaken, or even disable, these safeguards. Nevertheless, there are drawbacks to NCS structures.

First, outside shareholders often lack sufficient information about the company. It is sometimes difficult to explain an investment to uninformed shareholders; sometimes the investment requires secrecy for competitive reasons. The risk is that, in order to avoid being undervalued and replaced, corporate controllers may choose projects for their visibility to shareholders rather than for their ability to increase value.

Second, corporate controllers have less incentive to make firm-specific investments when they perceive high risks of being replaced involuntarily. Every company has its own characteristics, and a controller is expected to invest her time and resources to gain knowledge and skills concerning the particular company she serves. However, such knowledge and skills do not necessarily increase her value in the general labor market. More important, corporate controllers usually act as entrepreneurs whose task is to discover new and profitable business opportunities that other market participants do not appreciate and the market cannot price. There is no standardized approach for outside shareholders to measure the value of firm-specific investments. Corporate performance and share prices are noisy signals affected by too many factors beyond corporate controllers’ control. Consequently, it is normally impossible for corporate controllers to formally agree with outside shareholders on remuneration for firm-specific investments. Instead, it is necessary to take control of the company to get such remuneration.[3] Corporate control can bring to controllers many non-pecuniary benefits, like the social status of controlling a successful company, and the opportunity to liquidate their firm-specific investments in a (friendly) takeover.

CS structures can help solve the above problems of NCS structures and can take two forms: concentrated ownership structures or voting-cash flow rights separation.

A concentrated ownership structure ties controlling shareholders’ power closely to their personal wealth within the company, which gives them a strong incentive to run the company well. However, voting-cash flow rights separation decouples controlling shareholders’ power and financial interest in the company, thereby increasing the risk of shirking and tunneling. On the other hand, concentrated ownership structures lead to tension between corporate control and equity financing that voting-cash flow rights separation can solve.

As explained earlier, control power is important to corporate controllers who still await remuneration for their firm-specific investments. Therefore, in many cases, equity financing is conditioned upon controllers’ retaining corporate control. Under the one share-one vote principle, controlling shareholders may cause the company to forgo new, profitable investments if pursuing them requires additional equity and controlling shareholders are unable or unwilling to purchase new shares proportionately. In a company with voting-cash flow rights separation, controlling shareholders have no such reluctance to pursue investments. Moreover, voting-cash flow rights separation enables controlling shareholders to reduce their equity stakes without losing control, which makes them less averse to risk and more willing to let the company pursue risky – but potentially lucrative – investments.

If a company prefers voting-cash flow rights separation, it must decide what CEM to employ. In addition to dual class structures, three commonly-used CEMs are pyramids, cross-shareholdings and securities derivatives in the market for corporate votes.

Pyramids and cross-shareholdings separate voting rights from cash flow rights by linking multiple companies, and they may be extremely complex and opaque. Public investors may not even know who the ultimate controllers of their investments are. In contrast, dual class structures are much simpler and more transparent. Investors can easily discern the controllers of a dual class company and calibrate its voting-cash flow rights divergence.

The use of derivative transactions in the market for corporate votes is also not transparent because of its private nature. Moreover, these transactions have a bigger problem: They can be used to establish or reinforce an insider’s control over a company after its equity is issued to the public. Although dual class recapitalizations have the same problem,[4] securities derivatives perform worse. First, dual class recapitalizations at least require shareholder approval, but derivative transactions do not. Second, a dual class recapitalization is a one-time adjustment after which corporate control is clear, while a new corporate controller may emerge through derivative transactions at any time. Third, as explained later, the aforementioned problem associated with dual class recapitalizations can be solved by relatively simple regulations, thanks to the simplicity of dual class structures.

Compared with the other three CEMs, dual class structures have two disadvantages. The first is straightforward: A controller of a dual class company may decouple voting rights from cash flow rights without limit. For example, a company may issue “A” shares without votes to the public, and “B” shares with votes to its controllers.

The second disadvantage is that dual class recapitalizations may prompt shareholders to act against their best interests, depriving them of voting rights without due compensation. To begin with, dual class recapitalization proposals are rarely voted down because of public shareholders’ collective action problems. An individual shareholder has little incentive to do the research necessary for an informed vote, because the cost exceeds the expected return, owing to her small cash flow rights. Moreover, even when a shareholder confirms that a recapitalization proposal will diminish shareholder wealth, she does not have much incentive to organize opposition, because she would bear all the costs but enjoy a small part of the benefits. In addition, corporate insiders may bundle recapitalization proposals with unrelated proposals that public shareholders like and threaten not to pursue valuable investment projects without dual class equity financing.

Then the difference of various recapitalization mechanisms comes into play. Two of them are compared here: exchange offers and pro rata dividends. In an exchange offer, shareholders are given a time limit to choose either to keep their existing shares with inferior voting rights or to exchange them for newly-issued shares with superior voting rights. In most cases, increased dividend rights are granted to existing shares. Due to collective action problems, even a shareholder who knows that the consolidation of corporate control is harmful will choose to keep inferior voting shares. If enough public shareholders opt for superior voting shares and corporate insiders cannot entrench their control, the shareholder gains more by refusing the exchange. She can thereby obtain the increased dividend rights and free-ride on other shareholders’ efforts against managerial opportunism. If too few public shareholders choose superior voting shares, she is still better off refusing the exchange. Despite the increased agency risks, she at least gets a dividend preference. As a result, most shareholders will follow this strategy, and corporate insiders can acquire an immediate voting majority. It should also be noted that the value of increased dividend rights attached to inferior voting shares is unlikely to correspond with the value of superior voting rights,[5] owing to the absence of market pricing in exchange offer recapitalizations.

In a recapitalization by pro rata dividend, the company pays a dividend in the form of superior voting shares to all shareholders. The number of superior voting shares a shareholder receives is proportionate to her shareholding at the time of payment. As long as no restrictions are imposed on transfers of superior voting shares, such a recapitalization does not vary the original distribution of voting power in the company, and shifts in the control distribution can only occur as a result of market transactions conducted by individual shareholders.

Given the benefits and drawbacks of dual class structures, what can law do to minimize the downside?

Layer 1: Absence of shareholder monitoring, proxy contests, and hostile takeovers

In general, policymakers need not deal with the drawbacks of dual class structures compared with NCS structures, because the ability of proxy contests and hostile takeovers to discipline managerial inefficiency are, in practice, largely illusory. Hostile takeovers are generally rare, even in the U.S. and the UK, where widely-held listed companies predominate. Moreover, hostile takeovers are not primarily aimed at replacing inefficient management. Shareholders have even less incentive to discipline management through proxy contests, because most of the relevant benefits will be enjoyed by other shareholders who do nothing. The same economic rationale applies to shareholders’ other monitoring behavior.

Layer 2: Aggravation of tunneling and shirking

Tunneling results in the non-pro rata distribution of corporate resources between corporate controllers and non-controlling shareholders. Law directly controls tunneling, mostly through enforcement of corporate charters and corporate controllers’ fiduciary duties.

On the other hand, shirking impairs the quality of business decisions and that quality is generally not subject to legal scrutiny. Business decisions are inevitably made on the basis of incomplete information and uncertainty. Moreover, companies are unique, and so are the investments and other strategies that work best for each one. That’s why judges are loath to second-guess management and will generally refrain from interfering in business decisions, unless there are conflicts of interest. This judicial attitude is best illustrated in the U.S. by the well-known “business judgment rule:” When directors or officers exercise “process due care” and make an informed decision involving no conflicts of interest, their judgment is accepted as final, unless the complaining party can prove that the decision is irrational and serves no corporate purpose. Although other jurisdictions do not formally have the “business judgement rule,” many find it sensible and follow it to a large extent.

Layer 3: Extreme voting-cash flow rights divergence and coercive dual class recapitalizations

The divergence between voting rights and cash flow rights in a dual class company is generally controlled by limits on the maximum number of votes attributable to a superior voting share. In Sweden, for example, no share may carry voting rights more than 10 times greater than the voting rights of any other share.

It is worth noting that the basis for calculating a superior voting share’s votes does matter. Take the Swedish rule, for instance. Assume that a company’s “A” and “B” shares carry one vote per share, but the par value and dividend rights of “B” shares are less than one-tenth of those of “A” shares. The company thereby decouples voting rights from cash flow rights to an extent exceeding what is actually allowed but is technically in compliance. It is, therefore, better to base the calculation on superior voting shares’ economic elements rather than on inferior voting shares’ votes.

Regarding dual class recapitalizations, the key characteristic of “bad” recapitalization mechanisms is that they coerce existing public shareholders to give up their voting rights, though in a way that seems voluntary. Accordingly, the “New York Stock Exchange Listed Company Manual” permits dual class structures with one qualification: Existing shareholders’ voting rights cannot be disparately reduced or restricted through any corporate action or issuance.

Should dual class recapitalizations be completely prohibited solely because collective action problems prevent shareholders from voting down bad recapitalization proposals? Probably not. Shareholder approval of all sorts of fundamental corporate changes is susceptible to collective action problems, and law generally allows those changes, subject to stringent procedural requirements. Moreover, listed companies may legitimately need the flexibility to adjust their voting structures to accommodate market conditions.

Assuming sufficient legal protections, the real issue with dual class structures is whether their benefits outweigh the risk of shirking. If policymakers, after considering the legal and market conditions of their jurisdiction, prefer constraints on shirking derived from concentrated ownership structures, they should think about controlling not only dual class structures but also other CEMs.


[1] A shareholder is entitled to receive dividends that the board decides to issue and other financial benefits attached to her shares. These rights are referred to as “cash flow rights.”

[2] Depending on the context, corporate controllers may be directors and officers or controlling shareholders. Take a company with one share-one vote, for instance. When stock ownership is dispersed, directors and officers control the company. When a single shareholder or group owns a dominant portion of shares, that shareholder or group generally controls the company.

[3] Another method to get such remuneration is a golden parachute agreement in which management receives special severance pay when a shift of control occurs. However, golden parachutes create significant agency risks. For example, management may pursue a control transaction without considering shareholder wealth.

[4] A listed company may alter its charter to adopt a dual class structure either at the time of its IPO or sometime after its shares are publicly traded. The latter is termed by academia as a “dual class recapitalization”.

[5] It is a general practice that superior voting shares have 10 votes each, and inferior voting shares have an extra 10 percent of dividend rights.

This post comes to us from Junzheng Shen, a Ph.D. candidate at The University of Hong Kong, Faculty of Law. It is based on his recent paper, “The Anatomy of Dual Class Share Structures: A Comparative Perspective,” available here.