ACTIVISM FAILURE OF STATE-OWNED PENSION FUNDS WITH BOARD SEATS IN BRAZIL

Even though controlling shareholders are dominant in Brazil, there is a substantial number of shared control companies, in which a group of large shareholders controls the company by means of a formal agreement, and a few widely held ones (Zeidan & Fontes Filho, 2012; Sternberg, Leal, & Bortolon, 2011). The Brazilian state plays an important role as regulator, and somewhat indirectly, as an occasional activist shareholder through pension funds sponsored by state-owned companies or state-owned financial institutions. Investor protection and participation have improved Abstract


INTRODUCTION
Even though controlling shareholders are dominant in Brazil, there is a substantial number of shared control companies, in which a group of large shareholders controls the company by means of a formal agreement, and a few widely held ones (Zeidan & Fontes Filho, 2012;Sternberg, Leal, & Bortolon, 2011). The Brazilian state plays an important role as regulator, and somewhat indirectly, as an occasional activist shareholder through pension funds sponsored by state-owned companies or state-owned financial institutions. Investor protection and participation have improved through several important new pieces of regulation and self-regulation that have been implemented in the last two decades (Leal & Maranho, 2017;Leal, 2011; Di Miceli da Silveira & Saito, 2009). These changes followed the privatization programs of the 1990s in which pension funds sponsored by stateowned companies and other state-related institutions became large shareholders of newly privatized companies with a greater appetite for engagement with the investees (Crisóstomo & González, 2006).
Pension funds sponsored by state-owned companies are usually not-for-profit foundations that act on behalf of the employees of their sponsors. Some of them are the largest institutional investors in Brazil. Nevertheless, previous Brazilian research conveys no clear evidence that the presence of these institutional shareholders positively influences the performance and corporate governance of their investees (Sonza & Granzotto, 2018; De Oliveira, Leal, & Almeida, 2012; Punsuvo, Kayo, & Barros, 2007;Leal & Carvalhal-da-Silva, 2007). For instance, Sonza and Granzotto (2018) believe that pension funds do not have good monitoring skills and invest only for portfolio diversification, even though they do not produce evidence to support their conjecture. Carvalhal (2012) believes that institutional investors related to the government are more concerned with not being expropriated by asset transfers when they enter into shareholders' agreements than with corporate governance, dividends, financing, and investment, suggesting a more defensive attitude. De Oliveira et al. (2012) allege that pension funds prefer larger companies, regardless of their corporate governance practices because of greater market liquidity and the potential political influence they may exercise. Guimarães, Leal, Wanke, and Morey (2019) find that less efficient companies are more often targets of activists but do not find evidence that activism improves company efficiency in the near future. Maranho et al. (2020) concede that institutional investors may have a positive influence on corporate governance and performance but that their simple presence among the largest shareholders does not guarantee it. They detected weak evidence of a positive influence for foreign investors and those with long-term holdings, yet their results also conveyed a negative impact of institutional investors on corporate governance when institutional investors had representatives on the board of directors or were signatories of shareholder agreements. Moreover, their results are not specific for pension funds. Finally, Collares (2020) found that the presence of independent directors on the board is negatively related to activism, suggesting that poorer CG companies are more frequently targets of activism in Brazil, but this author did not evince and impact of activism on performance. Thus, the evidence about the positive impacts of activism in Brazil is non-existent or, at best, mixed.
The objective of this article, therefore, is to analyze situations in which the two largest public sector related pension funds, jointly with other institutional investors, engaged in non-state-owned companies to shed some light on the potential explanations about the nature of their activism and, as an extension, about the existing Brazilian evidence. Specifically, the article investigates if the actions of these institutional investors are due to lack of skill, lack of interest to engage, alignment with government policy in detriment of the interest of minority shareholders, or something else.
This article addresses two failed activism cases involving the two largest pension funds of Brazil, both sponsored by state-owned companies because their size could have increased their potential to succeed. These funds may, as a matter of alignment with the spirit of regulations, a champion for best corporate governance practices and performance. Yet they may also align with the incumbent government policies, even in detriment of minority shareholder interests, particularly as shareholders of state-owned companies. Thus, these institutional investors may behave differently, depending on the investee company. In order to avoid the second, detrimental, situation, the cases address companies in the health care service and food industries, which are not state-owned or state concessions, such as utilities.
This article contributes to the literature in some ways. It adds qualitative evidence about the reasons that led to the failure of activist pension funds to positively influence corporate governance, as evinced in previous Brazilian research, even when they secured board participation. One of the reasons that the case analyses brought about was the difficulty to assess and predict major corporate changes, even in the short run, or the consequences of these changes on the nature of their investment, suggesting that they may lack the skills to properly assess corporate risks. Another reason was ownership concentration. Thus, this article also contributes to the debate about passive and active investing. The article also offers a Brazilian literature review about activism, a topic without much research in emerging markets. Finally, the qualitative evidence herein may offer new possible avenues for qualitative and quantitative research about activism in emerging markets and help with future hypothesis formulation.
The structure of this paper is as follows. Section 2 reviews the related Brazilian empirical evidence on activism and also includes a summary of recent regulatory and self-regulatory events in Brazil to set the background. Section 3 includes a brief discussion of the method, particularly case selection, pension fund information, and methodological limitations. Section 4 contains the presentation and analysis of the cases and Section 5 concludes.

BRAZILIAN
LITERATURE REVIEW AND BACKGROUND 2.1. Literature review Crisóstomo and González (2006) report that pension funds have been given a more important role as shareholders during the privatization process that began in the 1990s, particularly those sponsored by large state-owned companies. Their presence in acquisition consortia may have made many of these transactions possible; and their participation in the shareholder agreements of these companies was expressed as a seat on the board of directors. In addition, the authors claim that Brazilian pension funds have adopted practices that encourage activism, such as participation in shareholders' meetings, the exercise of voting rights, and closer monitoring of company management.
High control concentration levels are a deterrent to activism (Gillan & Starks, 2003 Carvalhal (2012) states that shared control companies display a higher relative market value suggesting that investors appreciate the monitoring of the signatories of the control agreement. This author concludes that these agreements mitigate conflicts of interest, thereby protecting minority shareholders and increasing firm value, even after controlling for endogeneity. The major Brazilian pension funds are frequent signatories of shareholders' agreements, and their size makes them potential monitors of controlling shareholders, even when they are not part of the agreement (Sternberg et al., 2011;Crisóstomo & Gonzáles, 2006). However, Sonza and Granzotto (2018) state that although investors seem to value the presence of pension funds as relevant shareholders this is not related to better performance.
Maybe one reason is that the investor protection offered by shareholders' agreements is very low when the signatories are the government and institutional investors (Carvalhal, 2012). These signatories seem to be more concerned with asset transfers rather than with clauses related to corporate governance, dividends, financing, and investment, which may indicate a contradictory role of government-related institutional investors.
It is important to identify the alignment of institutional investors with controlling shareholders, as suggested in Carvalhal (2012). For example, the pension funds of state-owned enterprises may align with the controlling shareholder when they own equity in state-controlled companies. They may vote at meetings of shareholders and boards of directors in accordance with the political vision of the incumbent government, thereby dodging an alignment with other minority shareholders. Therefore, because several of the largest Brazilian pension funds are sponsored by companies controlled by the federal government and hold the largest stock portfolios among pension funds, it is possible that their managers, possibly appointed through the political influence of the ruling coalition, align with the controlling shareholders, to the detriment of other shareholders. Nevertheless, these managers have fiduciary duties with their beneficiaries and regulators may punish them in these cases. Finally, it is important to say that a 2016 law banned political appointees as managers of state-owned enterprises. It is quite possible that the negative relationship between pension funds ownership and the quality of corporate governance practices observed by some Brazilian authors was a reflection of this political facet.
The Brazilian evidence from studies that have attempted to correlate the presence of institutional investors to firm value, performance, and compensation does not find a positive relationship. Leal and Carvalhal-da-Silva (2007) claim there is no impact on firm value when institutional investors are the ultimate controlling shareholders (i.e., after taking into account indirect control structures). Regarding the payment of dividends, the authors found a significant and negative relationship with indirect control by institutional investors. Similar conclusions are presented in Sonza and Granzotto (2018) for a twenty-year period. They speculate that pension funds may invest only for portfolio diversification and have inferior monitoring skills.
De Oliveira et al. (2012) found no relationship between a relevant equity interest (more than 5%) of the three largest Brazilian pension funds (Previ, Petros and Funcef), all sponsored by state-owned companies, and the quality of corporate governance practices of investee companies. According to the authors, these pension funds accounted for 47% of Brazilian pension fund assets in December 2010. After controlling for firm size in multivariate models, no significant relationship was found between the size of the stake of these pension funds and the corporate governance practices of the investee companies. The authors comment that the results may be endogenous since pension funds prefer larger companies regardless of their corporate governance practices. This preference is due to their liquidity or even the political influence they can exert since they represent the beneficiaries of state-controlled companies (Carvalhal- scores. This relationship held both when the dependent variable was the corporate governance practices score as well as when it was the size of the pension fund's stake, with the score as an explanatory variable. Punsuvo et al. (2007), thus, show that there is reverse causality and that possibly these variables are endogenous, which casts doubt on an otherwise significant result. The authors also employed a categorical variable indicating whether institutions unconnected to the company's founders, which might be the institutional investors, exercised control but the result did not display statistical significance. Guimarães et al. (2019) find that less efficient companies are more often targets of activists through a company-level activism index for a sample of Brazilian companies between 2010-2014. However, they do not find evidence that activism improves company efficiency.
Maranho et al. (2020) did not find a relationship between an index disaggregated considering each pair of company (investee) and institutional investor and the quality of corporate governance practices and performance. These authors presented some evidence of a positive influence for foreign institutional investors as well as for those with long-term holdings. Contrastingly, they claim a negative impact of institutional investors on corporate governance when institutional investors had representatives on the board of directors or were signatories of shareholder agreements. They did not single out pension funds in their analysis. Moreover, it is possible that institutional investors prefer to invest in companies that have a priori best corporate governance practices, in line with the international literature (Gillan & Starks, 2007).
Finally, Collares (2020) also built an activism index for the 100 most liquid Brazilian listed companies in 2016. Her firm-level index consists of activist actions such as electing board members, rejecting or voting against management proposals, placing proposals for voting, requesting cumulative voting, and other actions. She found a negative relation between activist activity and corporate governance quality, particularly for the presence of independent directors on the board, suggesting that poorer CG companies are more frequently targets of activism. On the other hand, she finds no impact of activism on performance.  Table 1 summarizes a few selected events as a background for the cases.

Regulation Description
Instruction CVM 324/2000 Regulation issued by the securities commission of Brazil (CVM, Comissão de Valores Mobiliários) that reduced the minimum aggregate participation necessary for a group of shareholders to request the installation of a fiscal board. Fiscal boards are not mandatory for most companies, except state-owned, and may be temporary or permanent. Fiscal board member election procedures may make it easier for minority shareholders to achieve representation than in board of director elections and are a potential activism tool.

Law 10303/2001
Law 10303/2001 introduced several improvements to Law 6404/1976, the law of corporations. The most notable provision was to reduce the maximum proportion of non-voting shares to 50% of the equity capital, down from two-thirds in the previous law, and to concede additional rights to non-voting shareholders. It improved the rights of shareholders to elect and dismiss board members in terms of their aggregate voting rights, through cumulative or separate voting, as well as the introduction of shareholder proposals in assemblies. It also allowed the use of arbitration to solve shareholder conflicts, among other innovations.

Instruction CVM 480/2009
Reformed, structured, and vastly expanded the information that public companies ought to disclose to the market. It created a document, the Reference Form, which organized the way corporate information is presented and discussed.

Instruction CVM 481/2009
This instruction and ensuing updates in 2015 regulated shareholder assemblies. It included many details about the information to be provided in shareholder meeting calls according to several possible corporate events, shareholder participation rights, distance voting, and proxy solicitations. It eased shareholder participation requirements as well. The ensuing CVM instructions that reformed it introduced the distance-voting bulletin, a form with detailed topics shareholders could address. The bulletin allows the inclusion of shareholder proposals and demands for cumulative or separate voting for board members. Instruction CVM 324/2000 in Table 1 came first chronologically among the selected regulatory events. It eased the demands for a group of shareholders to request the installation of a fiscal board. The Brazilian fiscal board is separate from the board of directors and its members are not company insiders, such as senior managers and board directors. They do not partake in the strategic decision-making process but examine company information, particularly financial statements after they become public. They report directly to shareholders. Table 1 shows the main legal reform next, Law 10303/2001, that introduced several improvements to Law 6404/1976, the law of corporations. Table 1 reports the main changes this law introduced. Other significant changes were regulating mandatory bid rights in partial and total takeovers or loss of market liquidity situations, as well as when shareholders do not agree with certain types of proposals approved in a shareholder meeting. This law also defines abusive shareholder voting behavior, offers detailed norms about shareholder agreements, establishes the topics reserved exclusively to shareholder meetings as well as sets periods of time for the securities commission to intervene in order to allow more time for the shareholder assembly to exam proposals about complex transactions or that are potentially questionable or illegal. The law also introduced norms for publicity of information before shareholder meetings, several duties for the fiscal board, additional dividend rights, as well as many aspects regarding corporate crimes, such as insider trading.
Instruction CVM 480/2009 is the third selected piece of regulation in Table 1. It introduced major disclosure demands on internal controls, risk management procedures, compensation disclosure, detailed management, board of directors and fiscal board composition and information, management discussion and analysis, insider shareholding and trading policies, and many other topics. One of the most controversial was compensation disclosure (Barros, Di Miceli da Silveira, Bortolon, & Leal, 2015). It required disclosing the maximum, average, and minimum compensation paid to top management and board members. Even though it did not require individual and identified compensation disclosure, many large corporations resisted disclosing this information and were successful to secure a preliminary federal court injunction that allowed them to avoid disclosing everything about compensation until a final legal determination on the matter is reached. Courts, at the time, agreed with the claim that the maximum compensation disclosure virtually identified key persons and violated their privacy and security rights, turning them into potential crime victims. This injunction was finally overturned in 2018 and all companies now comply with this disclosure.
Instruction CVM 481/2009 is displayed next in Table 1 and deals with shareholder meetings. Even though distance voting came into effect in 2017, it is carried out through a bulletin that must be received by the company one week before the meeting. Electronic distance voting is still not allowed but the bulletin may be transmitted through a dedicated company electronic interface. Proxy solicitors may be intermediaries in delivering distance-voting bulletins. Distance participation in the assembly, such as through teleconferencing, is allowed but remote shareholders should have sent a distance-voting bulletin beforehand.
The last line of Table 1 shows Instruction CVM 586/2017 that introduced the comply or explain the practice in Brazil. The Brazilian Corporate Governance Institute introduced the first Brazilian corporate governance practices code in 1999, which is in its fifth version now and was the base for the unified code presented in this Instruction (Manzanares & Leal, 2020).
In closing, one should mention that the São Paulo Stock Exchange, now called B3 after successive mergers, introduced in 2000 premium listing levels that required better corporate governance, disclosure, and shareholder rights practices from companies. Joining these listing levels is voluntary but once listed, companies must abide by these requirements. These listing levels were at the center of a wave of IPOs between 2004 and 2007. The world financial crisis of 2008 and the Brazilian economic crisis from 2010 had an impact on the number of companies listed in these segments, with many going back private and some deciding to move from the premium to the regular listing level, abdicating from compliance with the premium level requirements. The requirements have been reformed twice, in 2010 and 2017, but some key proposals were rejected, such as a mandatory audit committee and an increase in the minimum number of independent directors.

CASE STUDIES METHOD
Two case studies were selected. The selection took into consideration some potential antecedents of activism mentioned in the literature, particularly in Starks (2003, 2007), as well as in the revised Brazilian literature. The cases involve activism in public companies with stocks traded in Novo Mercado, which is the most demanding trading list in terms of corporate governance practices because activists may prefer better-governed firms (Gillan & Starks, 2007;Vargas et al., 2018), even though the Brazilian evidence suggests otherwise (Maranho et al., 2020;Collares, 2020).
The cases also involve large and very visible companies because activist investors target larger companies since their shareholders have more difficulty monitoring management and therefore are more exposed to agency conflicts (Vargas et al., 2018;Goranova & Ryan, 2014). Thus, the two cases also involve companies that did not have a controlling shareholder at the onset of the case, so that there was more room for success if a large institutional investor decided to be more active (Goranova & Ryan, 2014;Chung & Zhang, 2011;Gillan & Starks, 2003). The cases also had to include pension funds sponsored by state-owned companies because these are the largest institutional investors in the country and, potentially, the ones that could be the most successful activist investors in Brazil in order to observe their actions in companies that were neither state-owned nor heavily regulated.
Regarding the two largest pension funds in Brazil, Previ and Petros are shareholders in the two selected cases. Previ was established in 1904 and serves the employees of Banco do Brasil (BB), which is its sponsor. BB is the largest financial institution controlled by the federal government and the second largest in Brazil, according to the Brazilian Central Bank. Previ's board of directors is made up of six board members, three nominated by BB and three elected by Previ's beneficiaries, the employees of BB. The board oversees the executive managers and has an audit committee. Previ also has a fiscal board that reports directly to its stakeholders, and not to the board of directors. Previ boasted an investment portfolio of about R$185 billion (US$35.63 billion) in March of 2020 (ABRAPP, 2020). This portfolio was worth R$155 billion (US$82.6 billion) in December of 2011. The portfolio increased 19.4% in Brazilian currency terms but decreased by 57% in dollar terms in the period due to the 56% depreciation of the Brazilian currency in the period. The accumulated inflation in the same period was 57% according to the official consumer prices index. Petros caters to the employees of the federal government-controlled Petrobras oil and gas group. It had a portfolio of approximately R$77.8 billion (US$15.0 billion) in March of 2020 (ABRAPP, 2020). In December of 2011, the portfolio was worth R$58.0 billion (US$30.9 billion). Even though its portfolio increased by 34% in Brazilian currency, it decreased by 51% in dollar terms due to the depreciation of the Brazilian currency in the period. Inflation in the same period was 57%. The board of directors of Petros is made up of six members, three appointed by Petrobras, and three elected by Petros' beneficiaries. Petros has a fiscal board that reports directly to its stakeholders. The board has an audit committee and other ad-hoc committees.
The two high profile cases selected had sample coverage in the business press and emerged from the examination of the available documentation and discussions with specialists. The sources of information about the cases in the press were, especially, the daily Valor Econômico, and the monthly magazine Capital Aberto, but not exclusively, as well as filings and documentation available through the securities commission, including any complaints and rulings by CVM, and the documents in the companies' websites, such as the calls and minutes of shareholder and board of directors meetings, as well as annual reports and other documents.
The DASA S.A. case deals with a company that had gone public and became somewhat widely held. DASA S.A. is one of the largest diagnostics companies in the world and consolidates a number of clinical and image diagnosis chains in Brazil. BRF S.A. is one of the largest food and protein producers in the world formed from a merger of the two larger companies in Brazil in this industry. It was controlled at the onset of the case through a shareholders' agreement that included both Previ and Petros among its signatories.
All analytical endeavors have their limitations. Information about activism campaigns is not completely public because many times activism is exerted in private between the company and its management. Given the difficulty to have access to the people involved in activist campaigns and to obtain their candid views about it, this article is limited to publicly available information from high profile cases. There were no interviews with the parties involved.
It is also necessary to bear in mind that the goal was not to measure the outcome of the activism of institutional investors in terms of performance impact or improvement in corporate governance metrics but to expose how activism is exercised in Brazil. It is believed that it is difficult to prove that a possible improvement in the company has occurred due to activism since several variables can influence the business outcome. Moreover, analyzing companies' financial information is essential to learn about the possible results of the activist activity, which was not done in this work.
Finally, the case method does not allow, nor does this study intend to produce, statistical generalizations from the analysis of the narratives involving the activism of the aforementioned institutional investors. Yet, it allows analytical generalizations (Yin, 2013). An analytical generalization is a cautious theoretical proposition and may take the form of lessons learned or hypotheses that may be applicable to other cases, which do not need to be necessarily similar. Even though the case studies are not samples of a larger group of cases, which is not the case with analytical generalizations, the lessons learned from a case may bring in implications that extend beyond similar cases, without a sample and population analogy, necessary in statistical generalizations (Yin, 2013).

CASE ANALYSES
This section proceeds with the narrative of the DASA and BRF cases, which are summarily portrayed in Table 2, and with an analysis of the narrative in terms of their relation to the conjectures found in the literature about the reasons for activism failure in Brazil.

Diagnósticos da América S.A. (DASA)
DASA went public in 2004 when the controlling group, composed of the founder and funds managed by Pátria, an asset management firm, sold part of their position, retaining 65% of the equity capital (Adachi, 2004). The controlling shareholders sold a large portion of their equity in the market, retaining 26% of the equity capital, and the company became widely held in 2006 (Viri, 2006).
Tarpon, an independent asset management firm, then became a relevant shareholder with about 6% of the equity capital (Viri, 2009a).
Another fund, managed by the independent asset management firm Skopos, also had a significant stake, and not only did it use its cumulative vote prerogative to elect a board member but also joined forces with other institutional investors to nominate members to the slate of directors to be elected by all shareholders at general meetings in 2009 (DASA, 2009). Various executives were also changed and one of the company's founders stepped down from the board where he had served since the company went public. These changes were deemed to be crucial since the company had various potential corporate governance problems involving related party transactions originating during its previous family-controlled phase (Viri, 2009b).
DASA acquired various companies after going public, becoming one of the leaders in the medical diagnostics industry in Brazil. One of the largest was MD1 Diagnostics (MD1) in 2010 in a transaction that was initially suspended by Brazil's anti-trust authority because the large market share of the new company could threaten competition. Following this acquisition, the MD1 controlling shareholders, Edson de Godoy Bueno, and his wife became the largest individual shareholders of DASA with a stake of around 25% (Santos, 2010). Bueno and his wife were, at the time, also the controlling shareholders of Amil, the leading health insurance company in the country, which caused controversy in markets due to possible conflicts of interest. Despite their sizeable stake, Bueno could only nominate a small number of directors who would have to abstain from voting on transactions with Amil or other firms in which Amil or their controlling shareholders had a stake (Viri, 2010).
Petros, the pension fund of the employees of Petrobras, the largest company in Brazil at the time and controlled by the federal government, acquired around 10% of the equity capital of the company in 2012, which gave it the prerogative of exercising a cumulative voting right to nominate a member of the board of directors. Petros possibly saw DASA as a company that could benefit from the rising income in the country up to that point, in addition to being in an industry that was still undergoing consolidation and had a low market capitalization relative to its book equity. As one of the officers of the pension fund declared, the aim was not only to invest in the company but also to influence its management by taking part in committees and structures and improving corporate governance (Mota & Falcão, 2012). DASA had other large institutional shareholders, such as Credit Suisse and Blackrock, with around 7% and 5% of the equity capital, respectively.
News about the departure of the chairman of the board in 2012 suggested that the Bueno family was more active because the position was filled with the CEO of the company, a trusted ally of Mr. Bueno (Koike, 2012). However, the slate for the board of directors was still nominated jointly by the Bueno family, Petros, and the other asset management firms in 2013.
After the anti-trust authority approved the MD1 acquisition, the Bueno family made a public offer to buy the remaining stock of DASA, seeking an additional 26.4% of the equity capital and the majority control (DASA, 2015). The offer price had a 13% premium over the market price. A foreign pension fund managed by Oppenheimer, as well as Petros and Tarpon, who jointly owned about 27% of the equity capital, were unwilling to sell. Even though they conceded that the valuation was within the legal parameters of economic value, they argued that the bid price was too low for investors with a long-term perspective. They also claimed that the various acquisitions involved significant capital spending and operational inefficiencies due to the merger of different companies (Ragazzi & Koike, 2014).
This placed Petros in a delicate situation because it had acquired its stake shortly before by more than the tender offer price. In addition, it had been attracted to the investment mainly due to the absence of a controlling shareholder, which enabled it to exert greater influence in the company and to avoid potential conflicts of interest resulting from the presence of a dominant shareholder. Even though Petros rejected the bid offer, its representative on the board voted in favor of the offer, arguing that the presence of a controlling shareholder could be beneficial to the company as a whole (Ragazzi & Koike, 2014). One of the options available to Petros and other institutional investors would be to undertake a counter tender offer at a higher price, but they did not go ahead with the idea.
In January 2014, the Bueno family carried on with the tender offer but failed to gain control. As the tender offer prospectus stipulated that it was conditional on gaining control, they had to remove this condition and make a new offer. This time around they ended up with 72% of the equity capital (Koike, 2014a). The decision not to sell by the above group of institutional investors was a crucial factor in preventing the Bueno family from obtaining control of the company in the first tender offer. The new ownership structure put Petros in a difficult situation, as it would need 15% of the equity capital in order to continue to nominate board members, thus making it necessary to act in concert with other investors.
Even though the Bueno family managed to conclude the transaction using an investment vehicle called Cromossomo Participações, this tender offer led this new shareholder to acquire more than 15% of the equity capital, thus, theoretically giving rise to the need to carry out a second tender offer in compliance with the poison pill provision in the bylaws of the company. The board of directors decided in favor of the new tender offer based on the legal opinion of a law firm (Viri, 2014). The issue was indeed controversial, since CVM, in a previous case, determined the sovereignty of the bylaws. However, undertaking a new tender offer would not make much sense, given that another one had already taken place. The Bueno family sent a letter to the company clarifying that they would not undertake a new tender offer because they understood that the previous tender offer had already satisfied the liquidity requirements for the departure of shareholders who disagreed with the new control structure (Meibak, 2014). The decision was referred to arbitration at the stock exchange because DASA was listed in the premium listing segment called Novo Mercado that requires the use of arbitration for speedier resolution of shareholder conflicts. Arbitration pitted Petros and the Oppenheimer fund against the Bueno family. The decision favored the family affirming that a new tender offer was not necessary. The board of directors also took part in the arbitration, alleging that a new analysis report would have to be prepared if a new tender offer were undertaken, as there had been a considerable change in business conditions since the previous tender offer (Koike, 2014b).
The large shareholding of the family gave Buenos the necessary power to promote a series of changes and embark on DASA on a new phase. Bueno nominated his son to the position of CEO and chairman of the board of directors and changed members of the board of directors and the executive team. Petros voted against these changes because it represented a threat to best corporate governance practices in their view (Koike, 2014c).
The Bueno family, through Cromossomo Participações, made a proposal to withdraw DASA from Novo Mercado in 2015, the trading list with the highest corporate governance standards in the Brazilian market. DASA argued that the compliance costs of Novo Mercado were not worth the benefits to the company and that it would have greater financing flexibility outside Novo Mercado, such as being able to issue non-voting shares. Another tender offer ensued and they offered R$10.50 per share to withdraw from Novo Mercado, well below the R$15.00 offered in 2014. The corporate restructurings and the bad phase of the Brazilian market were crucial in determining the lower price. Petros, once again, saw itself in an uncomfortable situation, as it had amassed a large position in a company with high standards of corporate governance, which, in their view, were now under threat. In addition, rules issued by private pension and retirement scheme regulators limit exposure (size of position) in accordance with the corporate governance standards of companies, which could require Petros to reduce its stake in DASA after the migration out of Novo Mercado. Petros complained publicly about the situation, criticizing not only the company but also the Novo Mercado listing segment mechanisms, which had allowed this to happen (Ragazzi, 2014).
Powerless in the board of directors, Petros made a final attempt at the general meeting of shareholders to bar the exit from Novo Mercado. Petros, along with the Oppenheimer Fund and the independent asset management firm Leblon Equities, complained with CVM about the lack of reasons for withdrawing the company from the Novo Mercado. They argued that the high corporate governance standards needed to protect minority shareholders could be weakened. The tender offer was suspended pending the decision of the regulatory authority, but the claim of minority shareholders was rejected and the offer finally took place (CVM, 2015). These institutional investors sold their shares at the tender offer held to allow the exit from Novo Mercado.
From the description above, one may conclude that Petros failed in its initial intentions to influence the management of DASA as a widely held Novo Mercado company after it, once again, became family-owned with a high ownership concentration and later left Novo Mercado. In this case, the good standards of corporate governance and the somewhat widely-held ownership of DASA seemed to motivate Petros to invest, consistently with potential antecedents of activism (Goranova & Ryan, 2014;Gillan & Starks, 2003). Petros also came across as acting in line with the advancements in shareholder rights in Brazil and possibly aimed to take advantage of the improving regulatory environment to use its weight to help further even more the corporate governance practices of a widely held company that was already listed in the segment with the highest required standards in the country. Thus, the case suggests that Petros had the interest to engage and that there was no evidence that it was conflicted when engaging, as it was looking after the best interest of its stakeholders.
However, the events that took place right after Petros invested, which the fund appears to have failed to anticipate or contain, led the company into another direction, resulting in the massive concentration of control rights and exit from the Novo Mercado premium listing segment and, ultimately, in the sale of Petros' stake after some attempts to increase the tender offer prices and prevent the delisting from Novo Mercado. Thus, this case attests to the difficulties of active investing because even large and engaged investors may not succeed to gauge the likelihood and potential consequences of major corporate changes that take place in the short run and that transform the ownership and management of a company. This could be suggestive that Petros lacked the skills to anticipate the major ownership transformation events that took place soon after it invested in DASA, as conjectured in Sonza and Granzotto (2018).

Brazil Foods S.A. (BRF S.A.)
This case involved one of the few Brazilian widely held companies, which facilitates minority shareholders to act together in order to exert influence, as well as Previ and Petros acting in concert with other institutional investors (wolf pack) and disagreement with each other.
Perdigão The two companies supposedly had good corporate governance standards because both had American Depositary Receipts (ADRs) listings and followed various best practices guidelines, such as having independent directors and board committees. Yet, this did not prevent Sadia from recording a large loss in 2009 due to excessive foreign currency hedging operations, which was legally challenged both in Brazil and in the US (CVM, 2008). This placed the company in a delicate financial situation and led to its incorporation by Perdigão, which had been growing since 1997 through various acquisitions. The Brazilian anti-trust authority approved the incorporation in 2011, and BRF, one of the largest food and protein companies in the world, was born.
A shareholder agreement among the pension fund shareholders of Perdigão (Previ, Sistel, Fapes, Real Grandeza, Previ Banerj, Petros, and Valia) was in force until 2011. All of these pension funds, except Valia, were sponsored by state-owned companies. The Vale mining company sponsors Valia, but Vale itself had been a state-owned company and still had large state-related shareholders, such as some of these same pension funds (BRF S.A., 2006). BRF became a widely held company after the expiration and non-renewal of the agreement. The largest stakes belonged to Previ and Petros, with approximately 13% and 10%, respectively, of the equity capital. Tarpon, an independent asset management firm, began to acquire a large position, attaining approximately 8% of the equity capital. These three investors nominated their respective board members on a single slate without any kind of formal agreement until 2012. Increases in the shared ownership of these investors were subject to a bylaws ceiling of 20% of equity before a tender offer to buy all outstanding shares at a premium was triggered.
Tarpon began to adopt an activist stance to increase its influence over the board of directors in 2013 because it wanted to nominate its chair. It approached Previ and Petros with a proposal of friendly cooperation to nominate Abílio Diniz to chair the board . At the time, Diniz was the former controlling shareholder of the giant supermarket and retail group Grupo Pão de Açucar (GPA), who was embroiled in a tumultuous departure process with Casino of France, the new majority shareholder of GPA.
Petros, however, did not accept Diniz, even though he had all the qualifications. It alluded to a serious conflict of interest because Diniz still had a seat on the board of GPA, which was one of the main customers of BRF. Petros, joining forces with Casino, continued to expose the situation and demanded more effective measures. Heated voting sessions ensued at the general shareholders' meetings in 2013, with Previ supporting Tarpon and opposing Petros (Ragazzi, Mattos, & Valenti, 2013).
Diniz becoming the new board chair was the end of an era because the outgoing chair occupied the position since 1994 when the company was solely Perdigão. The outcome of the proposal was uncertain until the day of the voting when the slate with Diniz as chair, nominated by Tarpon, and Sérgio Rosa as vice-chair, former Previ president and nominated by Previ, won by a large margin thanks to the strong support from foreign investors who followed the ISS proxy advisory voting recommendation. Petros, which continued to nominate a director to the slate, abstained, and gradually reduced its equity stake afterward (BRF S.A., 2013).
Representatives of Previ, Tarpon, and the Fontana family, as well as Diniz, launched a corporate reorganization program in order to improve operations and adopt a more global stance (Rocha, 2014). The company reduced the number of directors from 11 to 9 in 2015 and raised the mandatory tender offer trigger from 20 to 33% of the equity capital, making it possible that these institutional investors increased their equity participation or a bloc acquisition by another large strategic partner (BRF S.A., 2015). Ultimately, the performance of BRF was far from satisfactory, the strategic re-structuring did not attain the expected results, and it also faced a number of legal issues, related to possible failures in sanitary conditions, which hurt its market value. Its stock price dropped by nearly 50 % between early 2014 and the end of 2018. Diniz departed in 2018 but continued to hold an equity position in BRF.
From the description of the case above, the views of the two largest pension funds sponsored by state-owned companies were in opposition, suggesting that they acted independently from the government, and that, at least Petros, did not seem aligned with the national champions policy of the federal government of the time. This supports the view that these state-owned sponsored investors do not automatically align themselves with the policy views of the government and that they consider their own views of what is ultimately best for their investees and, consequently, their beneficiaries, also taking into consideration their fiduciary duties and regulatory obligations. Petros also saw conflicts of interest in the nomination of Diniz, which is an argument in line with the promotion of good corporate governance practices.
This suggests that both funds had the interest to engage and do not confirm that pension funds, in general, lack this interest. The results also contradict the premise that large pension funds sponsored by federal government-owned companies may be conflicted when engaging because they took different sides to the issue.
Previ, however, failed to assess the devastating consequences of the strategic changes promoted by the new management that they supported, and also to gauge the potential for major legal issues that the company had to face regarding sanitary inspection issues. Once again, this suggests that these pension funds may lack the skills to properly anticipate the consequences of major strategic changes that take place soon after their engagement. This evidence also brings into question the potential benefits of costly activism for asset management companies, such as these pension funds (French, 2008). Finally, the outcome is in line with the conclusions of Maranho et al. (2020) that considering each institutional investor and investee separately is necessary to try to ascertain the failure or success of their activist efforts.

CONCLUSION
Authors have conjectured that the lack of significant positive outcomes about the influence of institutional investors on performance and corporate governance in Brazil may be due to lack of skills (Sonza & Granzotto, 2018), lack of interest to engage (Punsuvo et al., 2007) and conflicts of interest when engaging (Maranho et al., 2020;Leal & Carvalhal-da-Silva, 2007). The ensuing discussion summarizes the findings and associates them with these conjectures.
The cases show that even these large and influential institutional shareholders fail and take on losses when trying to advocate in favor of principles such as less ownership concentration and higher corporate governance standards. One of them also did not do well when possibly trying to align with a "national champions" federal government policy of the time by supporting a notorious business tycoon. In the BRF case, the two largest pension funds in Brazil were in opposition, which suggests that they were not uniformly following a government policy orientation and that they were probably seeking to align with the interests of their beneficiaries, contradicting the conflicts of interest hypothesis when engaging. The analysis of these two cases is also in line with former quantitative studies in Brazil that conclude that there is no clear influence of the presence of these types of institutional shareholders and the improvement of corporate governance and performance ( Pension fund failures in these cases are largely due to failing to assess the consequences of major changes in the investees that were about to happen. They failed to predict a major control re-concentration and a potential drop in corporate governance quality, and the ensuing valuation issues in the consequent tender offer to acquire stocks, in the DASA case, and the failure of more aggressive management and strategic style that one of them supported, coupled with serious legal issues related to sanitary licenses, in the BRF case. Thus, the analysis of the cases is consistent with the conjecture advanced in Sonza and Granzotto (2018) that their assessment and monitoring skills may be lacking.
Yet, these pension funds sponsored by state-owned companies were not alone. Other private-side institutional investors sided with them and suffered similar losses. Thus, active investing, in general, seems to be a very risky game, even for deep pockets investors, as both pension funds and other Brazilian and foreign private-sector investors were together on the losing side. This result relates their activism to the debate about passive and active investing and how active portfolio managers have a hard time to beat passive portfolio strategies, after costs (French, 2008).
The limitations of this study were discussed in greater detail in Section 3. In summary, they relate to sourcing the cases solely from public sources to the difficulty to show cause and effect of institutional investor engagement, and to the fact that the case method does not allow for statistical generalizations even though it does not preclude analytical generalizations that may take the form of lessons learned (Yin, 2013).
Thus, as a final comment, future studies may need to frame their investigation in the active and passive portfolio management framework and question if the costs of activism are in the best interest of the beneficiaries of these institutional investors, given that the likelihood of failure seems high when considering the outcomes of the previous econometric studies as well as the analyses in this article. As a corollary, it is also possible that the assessment and monitoring skills of these large investors are not good enough for activism and, thus, if this is the case, why are not they? Another avenue for future studies is the issue of pension funds sponsored by state-owned companies when they engage in coalition with other state-related institutions to influence the strategic decisions of a state-owned company according to the views of the incumbent government because the results herein or of previous studies did not rule out conflicts of interest as a competing reason for failure in these particular situations.