The development of corporate governance has been marked by the quest for minimising conflicts between shareholders and executives. Since the seminal work of Berle and Means (1932), discussing the separation of ownership and control in American companies, the tension between these two groups of stakeholders that form the typical agency problem (type I) has led studies and models on governance to work to reduce the managerial expropriation and creating adequate incentives to achieve the maximisation of business return.

One case representing a landmark in the conflict between shareholders and management interests is Texaco, in 1984. The company’s executives, knowing that a group of investors were willing to purchase stocks to increase their ownership and take over the company’s control, decided to use the company’s resources to buy the investors’ stocks paying US$ 137 million premium over the market price, an option that was not given to the other shareholders. The executives avoided losing control of the company, without involving the other shareholders in the decision making (Monks & Minow, 2011). This case showed that executives were more concerned about holding their positions than with shareholders’ interests.

However, the magnitude of the principal-agent conflicts varies in a cross-sectional way among companies according to the ability of managers and shareholders to exercise their preferences; the complexity of the company; and the various characteristics of the industry. Therefore, agency problems are different in each firm and require different governance structures (Dey, 2008). Especially in contexts where the principal-principal relationship is predominant (because of concentrated ownership with an individual shareholder or group, where conflicts move to the relationship between the controlling and the minority shareholders), the idiosyncrasies of the companies and the different relationships between controlling shareholders, minority shareholders, and management result in distinct governance configuration that cannot be understood by cross-sectional modelling.

Also, the studies between firms reduce or even disregard the singularities of the companies and their external and internal relationships, as well as the tensions between stakeholders and the broader power games that characterise corporate governance (Licht, 2011). The static perspective in the study of governance considered is not always compatible with the transformations and adjustments that companies go through in processes of business rearrangement, changes in the shareholding structure, or to what extent governance evolves along the corporate life-cycle (O’Connor & Byrne, 2015a). This perspective does not allow understanding the effect that endogenous or exogenous mechanisms, or conditions such as propensity to risk of the players involved in the governance and availability of information, can introduce over time on the dynamics of the company’s governance, insufficiency that can aggravate problems of collective decision, free riding, asymmetries of information or moral hazard. In this sense, Filatotchev, Toms and Wright (2006, p. 257) remarks that “Much attention has been focused on the largest mature companies listed on a stock market, concentrating on the static theorising of the principal-agent perspective. Less attention has therefore been paid to the change processes in governance and variations in the principal-agent relationship through the full life cycle of the firm from inception to maturity”.

Over time, a change in the ownership structure, for example, brings about new conflicts, mainly due to changing expectations or interests between the parties, which can affect the quality of the shareholders or board decisions, and aggravate the principal-principal problem. The agency theory does not address the effects of these changes on the nature of the principal over agency problems, particularly for three reasons: a) for not bringing a dynamic view of the principal-agent relationship; b) for considering the principal as a single actor or homogeneous set aligned around the goal of maximization of companies’ value; and c) for putting forward its propositions mainly to avoid the risks associated with the behaviour of the agent.

Even if the ownership structure of a company remains the same, the owners’ interests may change, which is a significant event particularly in contexts of concentrated ownership, due to the greater influence of the controllers. These interests may change in several situations, for example when an investor needs to increase their dividends and imposes a disinvestment strategy; when a parent company is in a controlling position and changes its strategy in the business portfolio where the investee is included; when the state control the company, and a change of government occurs; as well as countless other possibilities.

However, despite this recognition regarding the variation of the agency conflicts over time, the studies on these conflicts and their solutions, internally within the companies, following a temporal, longitudinal perspective, are rare (O’Connor & Byrne, 2015a).

This study aims to contribute to the current literature on the topic seeking, as a research problem, to understand the relationships between changes in the life cycle of companies that influence their corporate configurations and structures, their governance arrangements and the way in which agency problems and principal-principal issues are handled. It aims to describe how the various conflicts inherent to governance change over time, particularly with the change in the group of relevant shareholders, or the change in the shareholders’ interests regarding the objectives and results of the business. Following the premise that the expectation of maximizing company value is not hegemonic to all shareholders (Jensen, 2010), investors can direct resources to a company to access critical resources and carry out synergies, increase market power or seize opportunities to improve the company’s efficiency and results, generating additional revenue (Weitzen & McCarthy, 2011).

In order to expand the understanding of how the relationships between owners that affect governance are constructed, this study identifies the evolution of these relationships and their effects on governance over time. As a methodology, this study adopts a longitudinal analysis of the transformations in the ownership and governance practices of a closed company. The company studied presents the main characteristics necessary to carry out the analysis, such as changes in ownership, participation of the state, changes in the nature of the business and in the performance expectations, and is less subject to exogenous pressures of the market. The company observed is the Transportadora Brasileira Gasoduto Bolívia-Brasil (TBG), controlled by Petrobras Logística de Gás S.A. TBG was established in 1997 to transport gas from Bolivia to Brazilian consumer markets, which also involved several private and state partners. TBG owns and operates the part located in Brazilian territory of the Bolivia-Brazil Gas Pipeline. The various ownership changes and strategic reconfigurations that TBG went through in the years following its foundation motivated the choice of this company for the study.

In this sense, adapting the line followed in a previous study by Thomsen (2004), we particularly examine changes throughout the company’s life cycle in three governance mechanisms: its ownership structure, the composition of the board and the influence of shareholders.

The agency theory discusses the problems resulting from the separation of ownership and management (agency problems), and it is the dominant theoretical perspective applied to corporate governance studies. The basic premise is that if both parties in a principal-agent relationship seek to maximise their utility function, the agent will not always act in the best interest of the principal. The principal may limit divergences from their interests by establishing incentives for the agent, and by incurring monitoring costs to limit agent’s aberrant activities (Jensen & Meckling, 1976).

Despite the predominance of analysis of static contexts and the use of cross-sectional data, an increasing amount of the literature discusses the limitations of this modelling because they do not reflect time-related interdependencies and present severe limitations on endogeneity and causality problems (Filatotchev & Wright, 2017).

Previous studies on the topic have observed that organizational culture differs according to enterprise life cycle stage (Belak, 2016), that corporate governance parameters are linked to strategic “thresholds” or transitions stages along the firm’s life-cycle (Filatotchev & Wright, 2006). Mature firms tend to practice better overall corporate governance, mainly by strengthening discipline and independence, while young firms tend to be more transparent and responsible (O’Connor & Byrne, 2015a), although governance creates more value for young or “immature” firms than for “mature” ones (O’Connor & Byrne, 2015b).

Other factors, such as changes in the actors’ propensity to risk (Cuevas-Rodríguez, Gomez-Mejia & Wiseman, 2012), the evolution of internal relations between shareholders and the control of managerial opportunism (Shen, 2003), or the changes brought about by deregulation processes, which reduce incentives for shareholders to monitor managers by introducing the disciplinary power of government agencies (Bongjin, 2002; Walsh & Seward, 1990), also impose requirements for reviewing governance structures and practices over time.

Aguilera, Florackis and Kim (2016) propose a new research agenda on the subject, incorporating, as well as the agent-principal, the idea of principal-principal conflicts. The authors recognise that different owners can have different preferences and time horizons, resulting in different conflicts of interest. The research on principal-principal conflicts has recently increased, accumulating knowledge and presenting new proposals.

When focusing on the principal and agency relationship, other issues emerge as relationships and identities are analysed according to their changes over time. Licht (2011) proposes a different perspective on corporate governance problems, in addition to the principal-agent relationship. For the author, corporate governance is, above all, a matter of power and how it is allocated in the company. This perspective refers to the numerous simplifications brought about by agency theory, which considers the principal as a single, homogeneous and continuous entity, and does not observe other influential actors in the configuration of company strategies.

O’Connor and Byrne (2015a) studied companies in 21 emerging economies, observing the relationship between corporate governance practices and the company’s lifecycle. The authors found that mature companies tend to follow the best governance practices, but companies tend to be more accountable when they are younger. Their research indicates that the functions of strategy, monitoring, and control have a distinct relevance over the companies’ lifecycle.

Filatotchev et al. (2006) identified different roles of the corporate governance in different stages of the companies’ lifecycle, pointing out that an appropriate change of stage is accompanied by rebalancing in the governance structure and roles.

Looking at the broader governance structures in franchising models, Mitsuhashi, Shane and Sine (2008) note that static explanations for the organisational form are insufficient to adequately explain the choices for specific types of governance structures. According to the authors, this happens because organisational structure is path dependent, i.e., depends on previous experiences and actions, and, therefore, when first experiences and actions lead the organisations in one direction, it is difficult to change this path. The authors point out that the form of organisational governance is shaped by dynamic historical processes, such as the organisational momentum. In short, “history matters!” (Mitsuhashi et al., 2008, p. 1135).

However, the search for alignment among shareholders of closed companies poses specific challenges, often different from those observed in publicly held corporations, due to the lack of liquidity of the shares, less external control by market agents, and less transparency regarding disclosing information. These issues are manifested especially in situations where the company is owned partially by other companies and is subject to their interests, operating as a joint venture.

Moll (2005) points out that the search for alignment among shareholders in closed companies presents specific challenges since they are characterised by a small number of shareholders, lack of a market to trade shares, and significant shareholder participation in the company’s management. The typical problems in these companies - differently from what is observed in publicly held companies - refer to conditions leading to minority shareholders oppression due to the absence of exit rights; the rule of corporate governance based on the majority’s will; deference to the business judgment role; and lack of adequate planning.

Shareholder’s exit rights, mainly associated with the existence of a market to trade shares, are useful in two main respects. First, they allow shareholders to liquidate their investment and recover the value of the invested capital. Second, the threat of exit in large numbers tends to restrict managers to take actions that detract from the shareholders’ interests. The absence of exit rights may lead to oppressive conduct such as the appropriation of minority shareholders’ investment. This absence in closed companies is perceived in many ways as minorities find themselves unable to sell their shares, to demand that another shareholder purchases the shares, or are unable to cause the company’s dissolution.

Moll (2005) argues that even in the absence of exit rights the oppression would not occur if the minority shareholders could block the actions of the controlling shareholders that may jeopardise their interest, i.e., if the minority has veto power in decisions that affect their interests. Usually, the controlling shareholder has the power to elect a majority of the representatives on the Board of Directors, and there is a tendency for the Directors to respond to the wishes of those who appointed them. If decisions within the board are made by a simple majority, even by appointing representatives, minority shareholders are not able to avoid oppressive conduct. One mechanism minority have, to avoid the principle of “majority rule”, is to require that board decision-making processes consider a voting system such as the adoption of the need for a supermajority.

Finally, the lack of planning before committing capital is one of the main problems in the relationship of shareholders of closed companies, especially affecting minority rights. Safeguard clauses such as supermajority provisions, buy-and-sell agreements, and labour contracts, can be negotiated with the controlling shareholders as preconditions for company formation. However, investors in closed corporations usually fail to demand these types of protection for reasons ranging from the existence of personal relationships or an initial atmosphere of mutual trust, or to the fact that shareholders are often not sophisticated in legal and business terms (Moll, 2005).

In order to understand the evolution of the relationships between owners and their effects on governance practices and structures, the research adopted a single longitudinal analysis (Miller & Friesen, 1984; Thomsen, 2004). It is an exploratory study carried out at the TBG, a closed corporation controlled by Petrobras Logística de Gás S.A. but formed with several partners in its ownership structure. TBG was chosen due to the intense reconfiguration of the shareholding structure throughout its history, as a result of the evolution of the business environment in Brazil and Bolivia, as well as shareholders strategic repositioning.

The study was conducted in three parts. First, a documentary research was carried out collecting and analysing material publicly disclosed by the company, describing its processes and considerations of shareholders and other organisations such as the World Bank and other governments, to understand the changes that occurred in the business environment, shareholders’ strategic objectives, shareholding structure, and the governance mechanisms implemented by TBG. Next, field research was carried out, supported by semi-structured interviews with managers who experienced the company’s governance at different times in its history, based on the issues raised in the theoretical background presented here. Finally, the results were analysed, observed in the light of the theoretical background, presenting a report with analysis and findings.

Fifteen face-to-face interviews were conducted with current and previous TBG managers, and two interviews with managers of the controlling company that performed important functions for TBG’s business. All interviews were conducted during the second semester of 2014. The sample obtained was considered sufficient, since it included almost all the main executives (Superintendent Directors) and a considerable number of Directors from different periods of the company’s history. Also, it was possible to observe saturation in responses (Fusch & Ness, 2015; Thiry-Cherques, 2009).

For TBG, because it was created based on project finance approved with financing agencies such as IADB and IBRD, the transparency on the shareholders’ roles, the limits of the managers’ responsibilities, and the functioning of the governance mechanisms were defined in its origin. This characteristic makes TBG different from most closed companies, which are driven to implement good governance practices as there is an increasing interest in professionalising management and having access to capital market. For TBG, the research showed that there was a high adherence to the recommendations of the Instituto Brasileiro de Governança Corporativa (IBGC) (Brazilian Institute of Corporate Governance), published in 2014 to guide closed companies.

The longitudinal study showed that the most important contracts for the operation of the company (Shareholders’ Agreements and Contracts on Gas Transportation) were built based on a wide negotiation between the shareholders. The resources used to discuss the clauses and safeguards in these documents aimed to mitigate the risks to the parties involved and were crucial for the preservation of the interests of shareholders and project financiers in the later events.

These prior negotiations proved to be fundamental for the stability of the business, considering the shareholders’ concern with possible risks to their brands in the face of operational failures or other negative events, which would also impact the situation of their parent companies. As Sabatino and Wolf (2010) argue in their analysis on the risks companies run because of their subsidiaries’ performance, it is critical that the subsidiary’s governance system is based on principles that are clear and consistent with the shareholders’ long-term global strategy. As a result, each shareholder brings their particular concern to the agreement, increasing the requirement for proper prior alignment.

For Moll (2005), a large part of the conflicts between shareholders in closed companies are caused by the lack of adequate planning before the creation of the company, lack of legal sophistication, and difficulty in predicting future situations. In the case of TBG, because it involved large multinational companies, with skilled legal departments, and contracts widely discussed, it is possible that larger conflicts were avoided in the events presented below.

An important factor to consider as an enforcement mechanism regarding the agreements was the financing structure. Because it was made via project finance, the guarantees were the project’s receivables, which were established in the contracts of gas transportation (the company’s only source of revenue).

From the minority shareholders’ point of view, the transactions involving TBG demanded strict contractual protection for many reasons. First, due to the uncertainties related to the development of the market and its regulation. Second, because of the presence of a controlling shareholder integrated into the operational chain and with a dominant position. Therefore, the risk of manipulation in the asymmetries of information was present. Third, the specificity of the asset creates a long-term bond between the service provider, TBG and its contractor, Petrobras. In this case, the contractor held 100% of the shares of TBG’s controlling shareholder (Gaspetro). According to the transaction costs theory, the combination of these three factors creates an environment conducive to opportunistic behaviour, something that was not identified in the research. Possibly, the rigour in the elaboration of the contracts has helped to avoid this type of situation. As recommended by Bennedsen and Wolfenzon (2000), the existence of a second shareholder with relevant participation, in line with that advocated by Pagano and Roell (1998), has proven to be an effective practice to preserve the interests of minority shareholders.

According to interviewees, there was no indication of disregard to rules or oppressive practices on the part of the controlling shareholder. Despite several changes in shareholders configuration across time. The conclusion is that the motivation for leaving the business was external to the company, so they are not related to its governance. Reasons such as portfolio positioning to align with global strategies, an evolution of the political environment in Bolivia and regulatory in Brazil, were decisive for the changes in shareholding structure over time.

Hetherington and Dooley (1977) argue that the lack of liquidity of minority shares is a potential condition for an oppressive stance of the controlling shareholder. By the number of transactions involving minority shareholders, however, it is possible to say that the TBG shares were quite liquid since it was not difficult to find investors interested in the business. Even without having access to the TBG Shareholders’ Agreement, it is possible to presume that the terms of the share sale and exit rights were well-defined.

It is interesting to note the importance of a well-defined controller in the case of subsidiaries created to fulfil a specific purpose (Baek, 2003), considering the attempts to diversify the company and the tensions observed within the Board of Directors. In both situations, the controller’s actions to avoid the diversification of activities and the signing of contracts not foreseen in the initial design of the business model, contributed so TBG remained faithful to the purpose for which it was created, honouring its debts and being profitable for the investors.

Box 2 resumes the critical elements listed in the theoretical framework and compares the origins of the main governance problems and the situation observed in the TBG´s longitudinal study.

Box 2 Source: Elaborated by the authors.

Even though the synergies expected by the minority shareholders have not been fully realised for other reasons not investigated in this study, it is possible to argue that TBG represented a gateway for foreign companies in the Brazilian oil and gas sector. The interviews carried out indicate that, from the perspective of Petrobras, the association with large foreign groups in a business partnership sharing decision has provided maturity to the company regarding governance and relationship with shareholders.

The interviews showed TBG gained its own identity, built from the shareholders’ experiences, the qualities they brought from the private businesses (agility, efficiency, cost control, discussion of budgets to exhaustion), and expertise of the majority shareholder, particularly regarding the knowledge of the Brazilian market, relationship with governmental agencies and institutions, understanding of regional cultural diversity, and capacity for mobilisation.

Considering the universe of Brazilian closed companies and the characteristics of TBG such as clarity of the shareholders’ and management’s roles; maturity of control mechanisms; and degree of transparency, it is possible to say that TBG is at a higher level organisation in terms of governance, which is also acknowledged by the Associação Brasileira das Companhias Abertas (Abrasca) (Brazilian Association of Public Companies). For seven years, ABRASCA considered TBG’s Annual Report the best of Brazil in the category closed company. In 2010, for the fourth time, TBG was awarded with the “Transparency Trophy” for the best publication of accounting demonstration in the group of closed companies granted by Associação Nacional dos Executivos de Finanças, Administração e Contabilidade (ANEFAC) (Brazilian Association of Finance, Administration and Accounting Executives).

This study aimed to contribute to a new perspective of the analysis of corporate governance practices, based on longitudinal studies, having as reference the role played by the partners in the changes of ownership and its reflexes in the attributions of the board of directors. This analytical perspective allowed to observe that not only governance solutions and configurations to deal with agency problems must be re-evaluated, but also the nature of the principal’s interests and their relationship in the ownership structure.

As in the study by Filatotchev et al. (2006), it is possible to observe different stages in the company’s governance history and identify the main problems in each stage.

In line with previous studies that similarly followed a longitudinal approach to analyze changes in the company’s governance in its life cycle (Esqueda & O’Connor, 2020; Filatotchev et al., 2006; O’Connor & Byrne, 2015a) we observed different stages in the company’s governance history, associated with different problems at each stage.If initially, the controller is supported by deep discussions to align interests, over time, with the arrival and exit of other shareholders, the main nature of the ownership (strategic or financial) becomes more evident. At this point, the tacit trust building proved to be a differential in the joint venture’s sustainability.

The results of the study point to a temporal variation in the agency problem, be it a type I agency problem - between the shareholders and managers - or type II - among the shareholders. This is the predominant situation in the situation studied, and it is diffused in emerging countries. Also, as a closed company, TBG was less influenced by external control mechanisms, especially the pressure of stock markets. Thus, the internal conflicts and the interests of shareholders are stressed over time.

As shown in TBG´s study, the variable ‘power’ can also be highlighted, as discussed by Licht (2011) to influence the business configuration. The various shareholders had distinct interests, mostly unrelated to maximising business value, which brings complexity to agreements aimed at defining the company’s strategies and priorities and exacerbating the potential for principal-principal conflicts. The methodology, based on a longitudinal study, made it possible to highlight how shareholders’ interests and the governance arrangements change, indicating the need for further studies that consider these variations over time. The results also suggest the need for new proposals for corporate governance guidelines that consider that agency problem, the nature and goals of shareholders and managers, risk propensity, and asymmetries of information between the parties, are subject to variations over time.

It should be considered, as a possible limitation of the study, that the choice of TBG for analysis brought the implicit bias of the participation of a state-owned majority shareholder with the capacity to risk the initial risks of the enterprise and to offer demand. In this sense, the presence of Petrobras as the controlling shareholder of TBG may have changed the attractiveness of the business and changed the parameters of participation in the shareholding structure of minority shareholders, although the presence of a controller may bring additional risk of expropriation and tunneling of minority shareholders. However, as the case of a similar investment in other companies shows, this guarantee does not mean the absence of challenges or security in the results. As the case of Sete Brasil shows, a company created in 2010 by Petrobras and public and private partners to facilitate the production of oil rigs (Gradilone & Galembeck, 2016), state control and commitment to demand does not detract from the importance of quality governance practices. As argued by Young, Peng, Ahlstrom and Bruton (2002, p. E-5), “In emerging economies, state ownership and the control of corporate assets are positively associated with an increase in the propensity and magnitude of the principal-principal agency”.

Other limitations that may have influenced the study are the adoption of a qualitative research based on bibliographical review and field research may not have exhausted the possible types of issues that result in governance problems in closed companies, and the sample of respondents by former managers appointed by the controlling shareholder. Of the 17 interviewees, only three were appointed by the minority. This restriction is because the controlling shareholder is responsible for appointing the three members of the Board of Executive Officers and a majority of the representatives of the Board of Directors. Although the analysis was conducted with as much impartiality as possible, this numerical superiority may have influenced the analysis. Despite these limitations, in advancing a path that has been little explored in the studies on governance with an emphasis on the history of the configurations and agreements, it is expected that this research contributes and stimulates a new dimension of analysis on governance issues.