Peripheral markets like Brazil are often hit by global trends that are far from the radar screens of local participants. Just like a tsunami, when it is too late to do something when it appears on the horizon, these trends bring consequences for which we may not be prepared.
That is the case of the ongoing discussion in the central markets regarding the exclusion of non-voting shares from the main market indices.
The debate on the “one share, one vote” rule is one of the most recurring in governance circles. On one side there are those who see in this system the one that leads to better corporate governance practices, bearing in mind the perfect alignment of interests between political power and the contribution of capital. After all, the vote is nothing more than the consideration given to the shareholder for the fact that he contributes with the capital without being entitled to any certain return (unlike creditors or employees, for example, whose remuneration is established in a contract). The shareholders are the residual claimants, that is, they only get what is “left over” after the company pays all its bills. Therefore, they have the voting power as an instrument of control to guarantee the return of their capital.
On the other side are those who advocate “market freedom” in the sense that companies should have at their disposal a wide range of securities to be issued according to their needs / possibilities. They also argue that financial investors typically are not interested in political power, remaining satisfied with the economic rights of the assets they buy. Thus, it would be justifiable to meet the demand of each “clientele,” whether it likes votes or economic exposure. Finally, they argue that multi-class stock structures allow the company to expand with the control in the hands of the founders, who would have greater capacity to add value and preserve the culture that leads to success.
In the first group, theorists of governance and institutional investors are usually included. In the second, some companies (usually with defined control), bankers, lawyers and other intermediaries are concentrated.
In Brazil, the experience is eloquent on this issue – although we often ignore it. We developed our capital market within a patrimonial state model, in which the maintenance of the stock control was pursued with fire and sword. Nevertheless, it is important to remember the lesson of the historian Ney Carvalho, in a lecture at the Amec Seminar in 2012 (in this link): the patrimonial state model is an inheritance of Getúlio Vargas, not something intrinsically cultural of Brazil.
Yet, our capital market was developed based on preferred shares, with no voting rights, which could represent 2/3 of the capital stock of the companies. Adding the pyramidal structures to this, often financed by the taxpayer’s money, we have become a worldwide example of such a distortion of incentives. With no surprises, we have lived for a long time with a proto-market that is significantly inferior to what Brazil needs to finance its development. Abuses against minority shareholders were scandalous – but, let’s repeat, not surprising, as controllers often had little or no skin in the game (indeed, we look forward to Nassim Taleb’s homonymous book). Privatization in the 1990s only aggravated this picture.
At the turn of the millennium it seemed we had learned the lesson. Law 10,303 reduced the maximum percentage of preferred shares to half the capital, and the Novo Mercado was established as practically the only viable alternative for the IPO, following the rule of one share, one vote.
Yet, now and again, specialists stand up with the banner of “creative” control structures that sooner or later charge their price for big corporate debacle. They claim that hugely successful companies like Berkshire Hathaway, Google to Facebook have several classes of stocks – omitting that these are exceptions that confirm the rule. In other words, companies that succeeded not because of their control structure, but perhaps despite of it. ISS and IRRC study (available here) makes it clear that on average companies with different classes of stocks perform worse and are worse priced by the market than those that follow one share, one vote. The study commissioned by Amec and CFA Institute to analyze the Brazilian market (available here) came to the same conclusion. For every Google there is an Uber
The exuberance of the American high-tech market, fueled by the close-to-zero interest rate in the last 10 years develops the arguments for the acceptance of creative structures.
But interestingly, such experts do not focus on the reaction of one of the most important players in the capital market: the institutional investors. Those who, in the end, are the final buyers of these assets, have long been complaining about the distortions created by multiclass structures. Simplistic critics said: ah, do not buy those assets! That is cynical. Given the worldwide trend of increasing passive or indexed investments, institutional investors are forced to buy securities they do not like, simply because they would be important parts of the stock indexes they need to follow.
Snapchat’s IPO was the last straw. By diverging from the standard procedure of granting restricted voting rights – but some right – to investors, and selling shares without NO voting rights, Snapchat led to a global reaction of pension funds and large asset managers. Several of them, and the associations that represent them, began to pressure the three major index providers – FT / Russell, S&P / Dow Jones and MSCI – to exclude such exotic and distorting assets from their benchmarks. Immediately, the three providers opened public consultations to consider the request. This is the tsunami that begins to show its face.
S&P and FT / Russell announced their decisions in July 2017 and chose very different paths.
S&P decided to no longer include companies that own more than one class of shares. That is, it is an absolute criterion, which does not consider the percentage of votes available to the market – if it has more than one class, it is out. However, there are two important restrictions. First, the current components of the index are grandfathered. And second, the decision applies solely to the US market through the S&P 1500 index and its subcomponents (including the S&P 500). In other words, the initial impact is zero, but new US companies that do not follow one share, one vote will not be included in the index.
It is a pity that S&P’s decision has not been extended to other markets where these problems are even more prevalent. It remains the expectation that, over time, S&P will standardize its approach to different markets.
FT / Russell has created the following criterion: to be included in the index a company must have at least 5% of its voting power among the outstanding shares. Unlike S&P, the decision applies to all market indices, including therefore Europe and some Asian countries, and their global indices (GEIS).
Despite its overall applicability, the parameter chosen by FT / Russell was so low that it affects few companies in practice – and no Brazilian company. The 5% parameter was apparently chosen with the aim of causing the least possible impact on the current constituents of the index, while at the same time responding to the reaction caused by the Snap and its 0% free float voting power. Other companies excluded from Russell’s global indices include Dell, Clear Channel, Federated Investors and Laureate.
But perhaps the most diligent and interesting process is that of MSCI. Unlike its competitors, the company made two rounds of public consultations, and the second is open until May 31, 2018. In the latest version of the consultation, the MSCI included a detailed paper (available here) explaining its approach. It addresses two of its competitors’ top issues. First, the applicability is global, and it does not create different standards for different markets. Second, instead of creating a “line in the sand” (the term used by FT / Russell itself), the MSCI proposes to adjust the weight of the index components according to the voting power of the free float. In this sense, if a company follows the rule of one share, one vote, 100% of its free float is used for its weighting in the indices. If it has different classes of shares, each of them has its voting power added as additional weighting, reducing the percentage of the market value of the free float eligible for the company’s weight in the index. At the limit (for example, in Snap’s case), if the free float’s voting power is zero, the weight in the index is zero.
It sounds complicated – and it is. The existence of special cases, such as golden shares, limits of votes to foreigners, and conditional voting make the analysis even more difficult. But the examination the published paper demonstrates a high degree of diligence and technical rigor in MSCI’s approach. And unlike its counterparts, the proposal brings significant impacts to the components of the main indices.
In the case of the MSCI ACWI (main index of the company), for example, it is estimated that 253 companies and a universe of 2,493 (10%) will be impacted. In the case of emerging markets, this percentage is 13%. Some companies will have small impacts while others will be completely excluded. Facebook shares, for example, which represent 0.95% of the global index, would have their weight reduced by 62%. (1). In the case of Berkshire Hathaway (0.53% of the index), the impact would be 16%.
The impact on Brazilian companies is certainly of interest to local investors. The most significant Brazilian share in the MSCI ACWI is Itaú PN. The drop in its share in the index, according to the simulation performed would be 82%, before the impact suffered by other companies. Pão de Açucar (CBD) would be excluded from the MSCI indices, since only non-voting shares are admitted to trading.
Although the analysis needs to be updated (it was made before the share class unification of Suzano and Vale, for example), MSCI estimates that two Brazilian companies would be excluded from the index (including Suzano, which is not true anymore), and other 16 would have their weight reduced. Brazil would be the country most affected by the changes: its reduced weight in the MSCI EM (emerging markets index) would be from the current 7.3% to 5.7% – the largest drop among all components of the index.
This analysis demonstrates the degree of distortion of our capital market, because of the misguided Vargas view mentioned in the beginning of this article, which will charge its price when the new MSCI methodology is implemented. And this happens even after the success of the Novo Mercado, which has reduced our dependence on preferred shares.
It also makes us reflect on the damage we self-inflict – and here we include companies, investors, banks, regulators and the stock exchange – by allowing a giant step in the wrong direction through “superpreffered” shares.
But on the other hand, the analysis also shows the unequivocal path we need to take if we are to build a healthier and globally recognized capital market: it is not by following sectorial fads or mantras of brokers that we will increase global investors’ interest in Brazilian stocks. On the contrary: increasing adherence to the principle of one share, one vote should be the cornerstone of all market participants mentioned above to recover the relevance of our market in the global environment.