Summary of MSCI Consultation Paper on Voting Rights and Index Inclusion

Dimitris Melas is Managing Director and Global Head of Core Equity Research at MSCI, Inc. This post is based on his MSCI publication.

Related research from the Program on Corporate Governance includes The Untenable Case for Perpetual Dual-Class Stock (discussed on the Forum here) and The Perils of Small-Minority Controllers (discussed on the Forum here), both by Lucian Bebchuk and Kobi Kastiel.

Equity indexes have evolved to fulfil multiple roles in the investment process and meet the needs of various types of investors. All institutional investors use indexes as market indicators and research tools. Asset owners employ them as policy benchmarks in their asset allocation. Active managers use them as performance benchmarks while passive investors use indexes as the basis for investment vehicles. To fulfil these multiple roles successfully, equity indexes aim to achieve comprehensive coverage of the underlying opportunity set by including all investable equity securities listed in the markets they seek to represent. In the MSCI consultation discussion paper, [1] we address the question of whether stocks of companies with multiple share classes having unequal voting rights (“unequal voting shares or stocks”) should be eligible for inclusion in equity indexes. We approach the question in two steps. First, we assess if unequal voting shares meet the definition of equity. Then, we examine the impact of unequal voting stocks from different institutional investor perspectives.

Are share classes with unequal voting rights equity securities? Are economic rights alone sufficient for a security to be deemed to be equity? Or are control rights also an inherent characteristic of equity? Separation between ownership and control is a common premise underpinning many corporate and investment structures. Examples include general partners and limited partners in private equity funds, investment managers and unit holders in mutual funds, and plan sponsors and beneficiaries in defined benefit pension funds. In the corporate world, companies obtain funding through different types of debt and equity capital. Investors hold particular debt and equity instruments to meet their objectives and constraints. Unequal voting rights stocks may be attractive to issuers who wish to separate ownership from control and to investors who require economic exposure without the need or desire to exercise control.

In addition to meeting the definition of equity, a security must be deemed to be investable in order to qualify for benchmark inclusion. In particular, the security must be deemed investable from the perspective of different types of international institutional investors, including asset owners, active managers and passive managers. Investability for the purpose of index inclusion is typically assessed in terms of company size, trading liquidity and security free float. As more unequal voting power issues come to the market and as indexes are increasingly used as portfolio implementation tools in passive investing, it may be appropriate to add voting to the existing criteria for index construction.

The need to add a new category of inclusion eligibility criteria to the set of investability criteria used in index construction results from the fact that unequal voting rights affect different investors in different ways. Active investors may judge for themselves if the growth prospects of a particular company or the superior skills of a visionary founder justify relinquishing voting rights. After assessing the potential benefits and drawbacks, an active investor may decide not to buy or to sell at a later stage. Passive investors have no such choice, their process mandates holding all index constituents. Large institutions with a long-term investment horizon (universal owners) also have little choice but to hold all index constituents. These investors may influence corporate policy only through voting, making voting rights an important dimension of investability for this group.

In this post, we argue that unequal voting shares represent equity securities and therefore should be eligible for index inclusion. However, we note that these securities have different impacts from certain investor perspectives. As a result, we propose to adjust the index weights of unequal voting stocks to reflect both their free float as well as their company-level listed voting power. Before discussing the rationale and implications of this proposal, we review the theoretical and empirical evidence around unequal voting structures.

Why we propose adjusting index weights

Equity indexes adjust constituent weights for free float, defined as the proportion of listed equity that is available to purchase in the market. Shares deemed strategic holdings are excluded from the calculation of free float because these shares are not investable. Our proposal to adjust constituent weights according to their issuer-listed voting power follows the same logic. Only the portion of total share capital that offers voting rights is deemed to be eligible for inclusion.

The ability to influence corporate policy through voting may be less important for certain types of investors. Active investors, such as actively managed mutual funds or hedge funds, are able to act on their assessment of the growth prospects and management quality of companies that offer reduced or zero voting rights equity. Following an informed assessment of potential risks and rewards, they may decide to invest in these companies. Even without equal voting rights, active investors can subsequently sell or short the stock of companies when growth prospects deteriorate or when insiders mismanage the company or fail to use their voting power to maximize shareholder value.

On the other hand, the ability to vote is particularly important for passive investors and large asset owners that have very long investment horizons and hold the “entire” market (universal owners). These types of institutions, either because of their process or because of their size and investment horizon, have committed to not sell the stock of public companies when insiders misuse their superior control rights. In these cases, engagement through voting or public agitation is the only way to affect changes in corporate policy for the benefit of outside shareholders, making voting power a more important criterion for passive investors.

How can equity indexes continue to offer comprehensive coverage while recognizing the importance of voting power for certain types of investors? One solution that would satisfy the need for complete coverage would be to continue to include in indexes all companies at their full weight irrespective of voting power. This approach may have been satisfactory in the past when relatively few companies listed unequal voting shares and when the primary purpose of benchmarks was to act as market indicators and research tools.

However, two recent trends in corporate governance and institutional investing may render the legacy approach obsolete or inadequate. First, we have seen a rising number of companies offer reduced or even zero voting power shares to outside investors. In addition, as passive investing grows in popularity, indexes become the building blocks for asset allocation and portfolio construction. In this new paradigm, the legacy approach of ignoring voting rights in index construction may no longer be adequate for passive investors and universal owners as it would create misalignment between economic exposure and voting power. In other words, it would force them to have excessive capital allocation to public companies where they have relatively low protection against insider misuse of control and limited or no possibility to influence corporate policy through engagement and voting.

At the other extreme, a solution that would recognize the importance of voting power would be to exclude all companies that offer differential voting rights from equity indexes. This approach would also be problematic for all index-linked investors as it would reduce their opportunity set and violate the basic index principle of offering comprehensive coverage of the listed investable equity market.

Other possible solutions that would avoid the two extremes of full inclusion irrespective of voting power or complete exclusion of all unequal voting structures would require arbitrary thresholds to determine at what level of listed voting power securities should be included in benchmarks. The challenge with these approaches is that there is no clear theoretical or empirical basis on which to select an appropriate exclusion level. Fifty percent appears to be an intuitively appealing middle point but screening out companies with listed voting power below 50% may lead to the exclusion from indexes of several successful high-profile companies. This would deprive investors of the opportunity to benefit from the growth prospects and diversification potential offered by these companies. [2]

In the MSCI paper, we propose to continue to include unequal voting stocks in equity indexes but to adjust their weights to reflect both free float and listed voting power. This proposal recognizes the additional constraints of passive investors and universal owners by aligning economic exposure and listed voting power, while continuing to offer all investors access to the entire universe of listed and investable equity securities.


A quote from Financial Times journalist Andrew Hill aptly summarizes the ambiguous nature of the one share, one vote debate:

“The advantage of a dual class share structure is that it protects entrepreneurial management from the demands of shareholders. The disadvantage of a dual class share structure is that it protects entrepreneurial management from the demands of shareholders.”

We examined the literature surrounding unequal voting structures. We found credible theoretical arguments both for and against the existence of these structures. The empirical evidence regarding the impact of unequal voting rights on company performance and portfolio returns was also mixed, with roughly similar number of studies reporting positive and negative effects. Our own analysis of index constituents with unequal voting power revealed substantial differences in the exposures of these securities to risk and return drivers, compared to the overall equity market. These differences are likely to have an impact on the performance of portfolios and indexes with varying allocations to unequal voting stocks.

Equity indexes aim to include all investable equity securities. Therefore, unequal voting shares must pass two tests to qualify for index inclusion. First, they must be listed equity. This is indeed often the case, as they offer economic rights and proportional ownership and many exchanges will agree to list such securities. Second, they must be deemed to be investable. Many unequal voting stocks may be highly investable in terms of size, liquidity and free float. However, we argued that the rising number of new issues with unequal voting rights and the increasing popularity of passive investing through indexes call for adding another eligibility requirement. The challenge for index construction is to include differential voting rights shares in the index to maintain comprehensive coverage of the equity universe, while appropriately reflecting the reduced voting power characteristics of these securities in index weights.

Our proposed solution: Continue to include in equity indexes securities with differential voting rights but with adjusted weights that reflect both free float and company-level listed voting power. This approach offers several potential benefits. It avoids artificial cut-off points that create binary outcomes and may be subject to gaming. It provides an incentive to companies to reduce the gap between free float and voting power. It penalizes companies that offer low voting power but excludes only a very small number of companies, such as Snap Inc., that provide zero voting power to outside shareholders. It enables indexes to continue to capture all investable equity securities and offer all investors the possibility of benefitting from the growth prospects of new companies that may have elected to offer differential voting rights. At the same time, the weight of constituents with reduced voting rights is adjusted according to company-level listed voting power, recognizing the impact of unequal voting rights, particularly for passive investors.

Equity markets and investment processes evolve through time. Index methodologies should follow this evolution to ensure they continue to serve the needs of market participants. We believe our proposal achieves the right balance between comprehensive coverage and index investability, and appropriately reflects the evolving needs and priorities of different types of investors by keeping unequal voting shares in indexes with weights that align economic exposure and listed voting power.


1See “Should equity indexes include stocks of companies with share classes having unequal voting rights?(go back)

2See “Putting the Spotlight on Spotify: Why Have Stocks with Unequal Voting Rights Outperformed?(go back)

Both comments and trackbacks are currently closed.