Stakeholder management: evidence on the performance of publicly traded companies

Gestão de stakeholders: evidências no desempenho das empresas de capital aberto

Renato Fabiano Cintra renatocintra@hotmail.com

Universidade Federal da Grande Dourados, Brazil

Ivano Ribeiro ivano.adm@gmail.com

Universidade Estadual do Oeste do Paraná, Brazil

Helder de Lima Fava helderlf@gmail.com

Universidade Estadual do Mato Grosso do Sul, Brazil

Benny Kramer Costa bennycosta@yahoo.com.br

Universidade Nove de Julho, Brazil

Researchers in strategic management seek to explain and predict organizational success ( Rumelt, Schendel, & Teece, 1991). In the development of this field, the concept of stakeholders played an important role in this discussion ( Sarturi, 2016) and, recently, the theory of stakeholders (TS) resurfaced in the debate involving strategy and competitive advantage ( Harrison, Bosse, & Phillips, 2010). Attention to stakeholders is a strategic issue for the company ( Crilly & Sloan, 2012). The task of executives is to manage and shape relations between groups in a way that creates value for all stakeholders and not just shareholders ( Hall, Millo, & Barman, 2015; Stocker, Arruda, Mascena, & Boaventura, 2020). The value is a central issue at TS ( Freeman, Harrison, Wicks, Pamar, & Colle, 2010). Some research has shown that attention has been focused on the theoretical discussion of value in the context of TS ( Cintra, Costa, Amâncio-Vieira, & Ribeiro, 2015; Cintra, Cassol, & Costa, 2017; Cintra, Costa, Oliveira, & Cassol, 2017). Another aspect emphasizes issues related to the value created and distributed to stakeholders ( Barbosa, 2018, 2019; Sarturi, Seravalli, & Boaventura, 2015; Sarturi, 2016). TS has been providing evidence on the relationships between the stakeholder management and the corporate objectives ( Jones, Wicks, & Freeman, 2017), i.e., it seeks to understand the cause and effect relationships between the organization and its stakeholders ( Mascena & Stocker, 2020). Research suggests the need to broaden the scope of the empirical analysis, which seeks to understand the relationship between the stakeholder treatment and the company’s performance ( Bosse, Phillips, & Harrison, 2009; Bosse & Coughlan, 2016; Faleye & Trahan, 2011). Whether focusing on the concept of value creation ( Garriga, 2014; Mitchell, Buren, Greenwood, & Freeman, 2015); value creation dynamics ( Garcia-Castro & Aguilera, 2015); and potential sources of value creation ( Tantalo & Priem, 2016). The adoption of a stakeholder management and the organizational performance is an issue that needs to be furthered ( Harrison & Bosse, 2013; Sarturi & Mascena, 2017). Understanding the factors that determine performance shows a research agenda in this field ( Gomes, Osborne, & Guarnieri, 2020). Studies on value have shown theoretical advances and points of consensus ( Sarturi, 2016). Although the literature advances ( Garcia-Castro & Aguilera, 2015), challenges in future research are noted ( Cintra et al., 2015; Cintra, Cassol, & Costa, 2017). One of these challenges is the expansion of quantitative research ( Cintra et al., 2015), in view of the generalization of results, which are incipient ( Cintra, Cassol, & Costa, 2017; Sarturi, 2016). Challenges that line up the study are: need for empirical evidence of TS as capable of producing value ( Cintra et al., 2017), or superior performance ( Harrison, Bosse, & Phillips, 2010); return to the emphasis of TS on strategic benefits in management ( Cintra et al., 2017); deepening of the empirical evidence on the adoption of stakeholder management and performance ( Harrison & Bosse, 2013; Sarturi & Mascena, 2017); detailed analysis based on the companies’ annual reports ( Dumitru, Guse, Feleaga, Mangiuc, & Feldioreaunu, 2015); and understanding of the empirical behavior of stakeholder service and company performance ( Sarturi, 2016). Based on these points, the research issue consists in analyzing: What is the impact of stakeholder management on the company’s performance? To make the survey operational, it compared companies that manage or not the stakeholders in relation to the result obtained in the performance indicators (net revenue, ROA, EBITDA and net debt). Therefore, it aims to analyze empirically the relationship between the adoption of a stakeholder management and performance. The research seeks to demonstrate evidence of issues addressed so far in the theoretical field, but with no direct relationship with a set of companies. The paper was organized in five parts. Besides the introduction, there is a review of management for stakeholders and organizational performance. In the third part, there are the methodological procedures. In the fourth part, there are the presentation and discussion of the results. Finally, there are the conclusion, limitation, and appointment of future research.

The stakeholder management guides simultaneous attention to stakeholders’ interests ( Donaldson & Preston, 1995). To do so, it is important to understand what their interests and motivations are and how they affect the business ( Maignan & Ferrell, 2004). Freeman ( 1984) highlights that the stakeholder management occurs when you know who your stakeholders are and consider their interests in organizational processes and develop skills in order to balance the interests of stakeholders aiming to achieve organizational objectives.

Because of this, the TS has gained prominence as a perspective for discussion about strategy and creation of competitive advantage ( Sarturi, Barakat, Mascena, & Fischmann, 2017) or performance ( Gomes, Osborne, & Guarnieri, 2020). The survival of the company depends on its ability to create and distribute sufficient value in order to meet the different expectations of stakeholders and ensure that they continue doing business with the company ( Clarkson, 1995; Coff, 1999; Sarturi, Seravalli, & Boaventura, 2015). The creation of corporate value is increasingly influenced by externalities that go beyond market logic ( Mio, 2020).

Research in this area highlights that there is a connection between the capacity of the relationship with stakeholders to bring competitive advantage to the company ( Brito & Berardi, 2010; Jones, Harrison, & Felps, 2018), and better performance ( Harrison, Bosse, & Phillips, 2009, 2010; Tantalo & Priem, 2016). There is evidence of the importance and influence of stakeholders in the survival of the organization ( Schianoni, Moraes, Castro, & Santos, 2013). Stakeholders have distinct relationships with the business and their perceptions regarding the company’s performance need to be considered ( Macêdo & Cândido, 2011).

The stakeholder management establishes that stakeholders, who are well treated, tend to retribute with positive attitudes and behaviors ( Harrison, Freeman, & Abreu, 2015), which makes it a mechanism for achieving superior performance. Motivated by the intense competitiveness in the market, companies start to consider their relationship with their stakeholders in order to leverage relationships to obtain competitive advantage ( Brandão, Diógenes, & Abreu, 2017).

Companies that meet the interests of stakeholders will be able to allocate more value to the organization in the long term ( Harrison & Wicks, 2013) and will therefore perform better. There is a need for external agents (other stakeholders) to have closer and more friendly relationships, in order to make a correct management of stakeholders ( Macêdo & Cândido, 2011). As the organization understands stakeholders as any group or individual that may affect or is affected by the achievement of the organization’s objectives, the need for processes and techniques to enhance the management capacity increases ( Freeman, 1984).

In this sense, it is believed that managers who relate to their stakeholders in a regime of mutual trust and cooperation will certainly achieve competitive advantage and superior performance ( Brandão, Diógenes, & Abreu, 2017; Jones, 1995). It is the nexus of contracts between the stakeholder and the company that sustain the relationship ( Jones, 1995). Therefore, the company ceases to be the unit of analysis, while organizational interactions become generators of value and competitiveness ( Brito & Berardi, 2010).

It is observed that although there is theoretical evidence that the stakeholder management has a relationship with superior organizational performance and connection with organizational practice ( Freeman, Phillips, & Sisodia, 2020), the empirical evidence from quantitative analysis is initial ( Sarturi, 2016). The inclusion of the stakeholder management in the performance measurement model has not been widely tested, which reinforces the need for studies in this direction ( Mascena & Stocker, 2020). The hypothesis is that companies that have a stakeholder management will have a superior performance when compared to companies that do not have a stakeholder management.

Based on the premise that financial measures reflect only part of the performance, that they have limitations, especially related to the time factor and that non-financial events are often the ones that determine changes in financial status ( Vasconcelos, Yoshitake, & Nascimento, 2005), for the research, it uses the quantitative approach oriented to a 16-year time frame (2001-2016) and attributes the materiality matrix proposed by the Global Reporting Initiative (GRI) as the criterion for framing the company that has or not the management for the stakeholders (non-financial event), as well as the proxy evaluation that has the closest proximity to the stakeholder management ( Mascena & Stocker, 2020).

The study was focused on the investigation of publicly traded companies listed at BM&F Bovespa, given that the information is available on the Economaticaâ database, as well as public reports and information on the investor relations portal or on the BM&F Bovespa website itself ( Dutra, Pavinato, Carrer, Camargo, & Olea, 2021; Guimarães, Rover, & Ferreira, 2018; Souza, Brighenti, & Hein, 2016). There are 362 companies of several sectors listed in the class common shares, these with the right to vote in the assemblies. The initial proposal was to analyze companies directly linked to the tourism sector, but only two companies were listed (small sample for statistical analysis). It was opted to amplify for all the companies of the industrial goods sectors, non-cyclic and cyclic consumption, justified having in mind that in its majority, these companies provide activities that make tangent or improve the condition of making the tourist activity. The idea of having three sectors meets the requirement that the sector can influence the behavior of the companies and has an aspect of the control variable in the comparison between the groups. When rescuing the control variables in empirical studies that have investigated performance, the most used are sector and company size ( Boaventura, Silva, & Bandeira-de-Mello, 2012).

Materiality is among the three most innovative items within the reports ( Mio, 2020) and as a proxy for the stakeholder management is aligned with recent research in the field and brings to light what really matters to stakeholders in resource allocation ( Barbosa, 2018, 2019), stakeholder engagement ( Stocker et al., 2020), dialogue with stakeholders ( Campra, Esposito, & Lombardi, 2020; Hsu, Lee, & Chao, 2013; Torelli, Balluchi, & Furlotti, 2019). GRI is a multi-stakeholder organization that proposes a worldwide standard for producing management reports. By using these guidelines, organizations have the possibility to evaluate and compare their operations and practices through internationally accepted criteria. The form is formed by seven dimensions: general; nature of the product; corporate governance; economic-financial; social; environmental; and climate change ( Sousa & Zucco, 2016).

The final sample was composed of 152 companies divided into three sectors: 51 of industrial goods; 81 of cyclic consumption; and 20 of non-cyclic consumption ( Appendix A). Once the companies were defined, it was proceeded to the identification of which had the GRI materiality matrix, as well as which years they were carried out. For that, documentary research was used, considering that it investigated which ones had the matrix and the documents that prove the facts. In an initial analysis, the group that had (49) and did not have (103) the GRI materiality matrix was compared. In the sequence, the same comparison was carried out for the three sectors (industrial goods, cyclic consumption, and non-cyclic consumption) in an isolated manner.

In order to make the comparison between the groups (with or without the matrix and sectors) the available values from 2001-2016 were used: net revenue; ROA; EBITDA; and net indebtedness. The use of these indicators is in the argument that: net revenue is important to start the managerial analysis of the company’s result, as well as reflection on profit and performance ( Sousa, Albuquerque, Rêgo, & Rodrigues, 2011); ROA measures the company’s profit generation potential ( Matarazzo, 2007) and it is used in empirical research to measure performance ( Boaventura, Silva, & Bandeira-de-Mello, 2012; Sarturi, 2016); Among the accounting indicators used to measure the effectiveness of the organizational performance, the EBITDA stands out, because it shows the generation of resources considering only the operational activities, it eliminates the effects of non-disbursable expenses such as depreciation, amortization and exhaustion, besides showing the capacity of investments, payments to creditors and distribution of dividends to shareholders ( Ritta, Jacomossi, Fabris, & Klann, 2017); and indebtedness may affect profit and restrict the behavior of managers ( Barnett & Salomon, 2012; Drigo & Mendes Neto, 2017) and is relevant for long-term analysis ( Souza, Brighenti, & Hein, 2016).

The information net revenue, ROA, EBITDA and net debt were collected from the Economatica® system. The system offers information about all the companies listed in the United States, Brazil, Argentina, Chile, Mexico, Peru, Colombia, and Venezuela. The database consists of several years’ history of financial statements; daily share prices; earnings (dividends, splits, etc.); and name and participation of the main shareholders.

The following parameters were used to make the database: (a) t0 as the year of entry into the GRI and thus the first year of preparation of the materiality matrix; t_before composed of t-1 (1 year before the entry for those who have the matrix or the most recent year for those who do not have the matrix), t-2 (2 years before the entry for those who have the matrix or the second most recent year for those who do not have the matrix), t-3 (3 years before the entry for those who have the matrix or the third most recent year for those who do not have the matrix) and so on until t-16; t_after composed of t1 (1st year after the entry for those who have the matrix), t2 (2nd year after the entry for those who have the matrix) and so on until t15. For the companies that did not have the matrix, the average of the whole period was used to compose the t_after, in order to proceed with the comparison of performances. In total there were 1,886 observations of net revenue, 1,908 of ROA, 1,781 of EBITDA and 1,915 of net debt.

In order to verify if the two groups (with and without the matrix) had significant differences in relation to the indicators, the Mann-Whitney U test was used. The value of U (statistic used in the test) is obtained by the number of times that a score in the group with n2 cases precedes a score in the group with n1 cases in the group ordered incrementally. The Mann-Whitney U test is a non-parametric alternative to the t-Student test, from which it is possible to compare the distribution functions of a variable in two samples, being indicated where there are heterogeneous variances, and reduced sample ( Field, 2013; Marôco, 2011). Unlike the t-test, which tests the equality of means, the Mann-Whitney U test tests the distribution parameters. The U values calculated by the test evaluate the degree of data interlacing of the two groups after sorting. The greater separation of the data together indicates that the samples are distinct, rejecting the hypothesis of equality of the groups. For the tests, a significance level of 0.05 (p. <0.05) was adopted.

In order to demonstrate the quantitative evidence, it was decided to make comparisons between groups and sectors. For the comparison, the information was used to construct three periods t_before (average from t-1 to t-16), t0 (year of entry in GRI) and t_after (average from t1 to t15). The comparison is based on the premise that the companies that carried out the matrix demonstrated guidance to the stakeholders, attending the management of stakeholders, even if in a partial way, considering that the process of consultation and discussion is oriented to the interests of the stakeholders, as they appear in the reports. The hypothesis is oriented so that the companies that have the stakeholder management will have superior performance in net revenue, ROA, EBITDA and lower indebtedness, independent of the sector in which it operates. In order to begin the quantitative evidence, the Mann-Whitney test was performed to compare the groups with the net revenue in the three periods (Figure  1).

Figure 1 -
Mann-Whitney U test between groups, sectors and net revenue in the three periods

Source: Research results

From Figure  1, it is possible to affirm that there are significant differences at the level of at least 0.05 in the final sample for the group that carries out the materiality matrix as to the comparison of the performance in the net revenue, being superior for the group that carries out in the three periods: t_before (U = 1481.0; p = 0.000); t0 (U = 1136.5; p = 0.000); and t_after (U = 682.0; p = 0.000). It can be noted that the groups show significant differences, and the group that has the matrix tends to perform better than the group that does not have it.

By segmenting the comparison of the groups within the sector itself (cyclic consumption, non-cyclic consumption, and industrial goods) you can see that only the industrial goods sector at t_before showed no significant difference at the level of 0.05. The rest of the periods and other sectors showed significant differences at the level of 0.05. The Cyclic Consumption sector showed the values: t_before (U = 417.0; p = 0.042); t0 (U = 290.0; p = 0.000); and t_after (U = 143.0; p = 0.000). The Non-cyclic consumption sector showed the values: t_before (U = 11.0; p = 0.019); t0 (U = 18.0; p = 0.043); and t_after (U = 8.0; p = 0.002). The industrial goods sector showed the values: t_before (U = 199.0; p = 0.056); t0 (U = 133.0; p = 0.001); and t_after (U = 100.0; p = 0.000).

It was possible to identify that the sector has interference in the performance behavior, because while in the final sample the performance of those who had the matrix grows and those who do not have it reduces. In this context, having management for stakeholders presents favorable results in terms of direct improvement in net revenue. On the verification by sector, the behavior wasn’t uniform and neither there is a defined trend, considering that on the cyclic consumption was mirrored on the final sample, the non-cyclic consumption had growth of the group which doesn’t have the matrix at t0 and industrial goods had growth at t_after too. It was possible to identify better performance for net income in all sectors of those who have the matrix, in other words, of those who have the stakeholder management. To have a stakeholder management reflects on the performance of net revenue in the short and long term. The second indicator to be tested was the ROA in the periods and with the groups (Figure  2).

Figure 2 -
Mann-Whitney U test between groups, sectors and ROA in the three periods

From Figure  2, it is possible to affirm that there are significant differences at the level of 0.05 in the final sample for the group that performs the materiality matrix as to the comparison of the ROA at t_before (U = 1591.0; p = 0.002) and t0 (U = 1554.0; p = 0.001). It is noted, in the final sample, that the groups show differences, while at t_after the group with the matrix shows superior performance, but there is no significant difference at the level of 0.05.

It is worth pointing out that, in the course of time, from the second year on, the group that performs the materiality matrix has a superior performance, but has no significant difference. This finding has already been highlighted by Garcia-Castro, Arino, and Canela ( 2011), in an empirical discussion of the impact of the stakeholder management in the short and long term for shareholders, report negative effects in the short term, while finding positive effects in the long term.

It is possible to infer that having a stakeholder management causes a fairer ROA in the long term, that is, after the management is effective, the company’s shares are channeled in returns beyond the shareholders, considering that the average rank of the test for the ROA was reduced in relation to previous periods for companies that have the materiality matrix, but even so it remains superior in relation to companies that do not have the matrix formalized.

By segmenting the comparison of groups within sectors you can see that the cyclic consumption did not show significant difference at the level of 0.05 at t_before and t_after, industrial goods at t_after and non-cyclic consumption in the three periods (t_before, t0 and t_after). The cyclic consumption showed significant difference at the level of 0.05 for t0 and industrial goods at t_before and t0. The cyclic consumption sector showed significant difference at the level of 0.05 for t_before and t_after. The cyclic consumption sector showed the values: t_before (U = 456; p = 0.074); t0 (U = 414.0; p = 0.025); and t_after (U = 350.0; p = 0.115). The non-cyclic consumption sector showed the values: t_before (U = 33.0; p = 0.779); t0 (U = 26.0; p = 0.349); and t_after (U = 44.0; p = 0.758). The industrial goods sector showed the values: t_before (U = 167.0; p = 0.011); t0 (U = 162.0; p = 0.008); and t_after (U = 229.0; p = 0.300).

It was possible to identify that there is no uniform behavior and no defined trend. It identified better performance for ROA in all sectors of those who have the materiality matrix, i.e., those who have the management of stakeholders. A reduction in the average ROA rank at t_after, which may indicate a fairer value creation and distribution relationship for other stakeholders, being beyond the exclusive interests of shareholders. It is possible to infer that having a stakeholder management causes a higher positive impact on the ROA in the short and fairer in the long term. The third indicator to be tested was EBITDA in the periods and groups (Figure  3).

Figure 3 -
Mann-Whitney U test groups, sectors and EBITDA in the three periods

From Figure  3 it is possible to affirm that there are significant differences at the level of 0.05 in the final sample for the group that carries out the materiality matrix as to the comparison of the EBITDA, being superior for the group that carries out in the three periods: t_before (U = 1206.0; p = 0.000); t0 (U = 976.0; p = 0.000); and t_after (U = 819.0; p = 0.000). It can be seen in the final sample that the groups show important differences, while the group that has the matrix tends to improve the EBITDA performance, the group that does not have the matrix tends to get worse. It is possible to infer that having a stakeholder management causes a positive impact on EBITDA. In other words, the stakeholder management has an influence on the generation of resources, as well as on the improvement of investment capacity, payments to creditors and distribution of dividends to shareholders.

By segmenting the comparison of the groups inside the own sector, you can verify that only the non-cyclic consumption sector at t_before did not show significant difference at 0.05 level. The rest of the periods and other sectors showed significant differences at the level of 0.05. The cyclic consumption sector showed the values: t_ before (U = 322.0; p = 0.005); t0 (U = 329.0; p = 0.001); and t_after (U = 154.0; p = 0.000). The non-cyclical consumption sector showed the values: t_before (U = 21.0; p = 0.160); t0 (U = 9.0; p = 0.011); and t_after (U = 16.0; p = 0.014). The industrial goods sector showed the values: t_before (U = 135.0; p = 0.001); t0 (U = 95.0; p = 0.000); and t_after (U = 123.0; p = 0.001).

It was possible to identify that the sector has interference on the behavior of the EBITDA performance, because while in the final sample the performance of those who had the matrix grows and those who do not have the matrix reduces, when looking at the sector, the behavior was not uniform, and there is not a defined trend, since in the cyclic consumption was mirrored in the final sample, the non-cyclic consumption of industrial goods there was growth of the group that does not have the matrix at t0, but did not have identical behavior at t_after. It was possible to identify better performance for the EBITDA in all sectors of those who have the management of stakeholders (with the materiality matrix). It is possible to infer that maintaining a relationship with stakeholders has a positive and superior impact on EBITDA at t0 and t_after, which reinforces the superior performance of companies that manage stakeholders. The last indicator to be tested was net debt in the periods and groups (Figure  4). It is worth noting that the companies with the best performance are those with the lowest debt levels ( Mendes & Santos, 2018).

Figure 4 -
Mann-Whitney U test groups, sectors and net debt in the three periods

From Figure  4, it is possible to affirm that there are significant differences at the level of 0.05 in the final sample for the group that carries out the materiality matrix as to the comparison of net indebtedness, being superior for the group that carries out in the three periods: t_before (U = 1830.0; p = 0.042); t0 (U = 1735.0; p = 0.009); and t_after (U = 1338.0; p = 0.000). It can be seen in the final sample that the groups show important differences, while the group that has the matrix tends to improve the performance of the net indebtedness, the group that does not perform tends to get worse. It is possible to infer that the stakeholder management causes a positive impact on indebtedness. In other words, the results confirmed that the indebtedness restricts the behavior of managers towards stakeholders and, therefore, it is feasible that companies that have the matrix tend to have a favorable performance in relation to indebtedness.

By segmenting the comparison of the groups within the sector itself, you can see that the cyclic consumption sector did not show a significant difference at the level of 0.05 for any of the three periods, while in the industrial goods sector it was only at t_before. The rest of the periods and sectors showed significant differences at the level of 0.05. The cyclic consumption showed the values: t_before (U = 531.0; p = 0.443); t0 (U = 600.0; p = 0.746); and t_after (U = 435.0; p = 0.483). The non-cyclic consumption sector showed the values: t_before (U = 13.0; p = 0.031); t0 (U = 8.0; p = 0.009); and t_after (U = 20.0; p = 0.031). The industrial goods sector showed the values: t_before (U = 244.0; p = 0.313); t0 (U = 196.0; p = 0.049); and t_after (U = 130.0; p = 0.002).

It was possible to identify that the sector has interference in the behavior of the debt performance, because while in the final sample the performance of those who had the matrix grows and those who do not have the matrix decreases, when looking at the sector, the behavior was not uniform and there is not a defined trend. It identified better performance for indebtedness in all sectors of those who have the management of stakeholders (with the matrix). The materiality matrix relates to the performance and importance of managing primary stakeholder groups to obtain results ( Barbosa, 2019) and that the inclusion of stakeholders in the management process brings competitive advantages to organizations ( Stocker & Mascena, 2019). It inferred that maintaining a relationship with stakeholders has a positive impact on net indebtedness, but that for some sectors the difference is not significant. This finding may be associated with the idea that indebtedness is influenced by exogenous and endogenous variables ( Mendes & Santos, 2018).

Finally, it was possible to infer that the stakeholder management has a positive impact on net revenue, EBITDA, ROA and net indebtedness (here, inverted thinking in relation to other indicators, because the lower the indebtedness, the better the result), being superior in companies that have the materiality matrix. Even if in some cases there is no significant difference at the level of 0.05, there is statistical evidence that shows that the values of the group with the materiality matrix (stakeholder management) are higher than the values of the group without the matrix. These results suggest that the materiality matrix can be a useful and positive tool for companies that intend to strengthen their relationship with stakeholders. It also showed that the sector can influence differences between groups, but always with higher values for the group that has the materiality matrix, i.e., that has the stakeholder management orientation.

The idea that stakeholder management can be correlated with the organization’s performance was evidenced in the study and, therefore, the study brings new empirical inclinations to the field of stakeholder theory. In other words, as evidenced by Stocker, Sarturi and Barakat ( 2020, p. :12): “an important premise of the stakeholder literature refers to the expectation that companies that manage their stakeholders superiorly also present superior financial performance”, as evidenced by empirically in this research.

From the results of this study, it was possible to understand, quantitatively, that there is a relationship of improvement in the organizational performance of companies that have the stakeholder management. In other words, it brings empirical evidence that stakeholder management unlocks and enhances value creation ( Harrison, Bosse, & Phillips, 2010), as well as being positively related to the company’s performance ( Bridoux & Stoelhorst, 2014; Harrison, Bosse, & Phillips, 2010), in view of the best results and significant statistical relationship.

The study contributes with empirical evidence of the stakeholder management in the context of value creation, because although previous studies have investigated the relationship of indicators in the context of sustainable actions or social responsibility, the insertion of the stakeholder management variable to understand value creation is unprecedented. It empirically supports the idea that the creation of connections between companies and stakeholders open invisible opportunities for value creation ( Camilleri, 2012), as well as contributing with evidence of the existence of a direct relationship between the stakeholder management and the superior organizational performance, given the highlight of Bridoux and Stoelhorst ( 2014), Harrison, Bosse, and Phillips ( 2010) and Harrison and Bosse ( 2013).

The study improves the understanding that it is not the fact that it belongs to the index, as it is the case of ISE-GRI ( Maia, Carvalho, Klotzle, Pinto, & Motta, 2017), that the company’s results point to a superior performance of those that do not belong, but the effective management of stakeholders for a positive result in the short and long term, as it was evidenced in this study. The implications of this study lead to important insights into the use of the materiality matrix, contained in sustainability reports or integrated reports, as a step towards integral stakeholder management, with the emphasis on the fact that primary stakeholders have a greater influence on the organization’s processes and actions than secondary stakeholders.

As a limitation, the study has the composition of the sectors, which can be expanded in future research, for all sectors of the BM&F Bovespa, including the 362 companies. A sample with several sectors can improve the inferences. It is suggested to perform other statistical tests, such as regression, including other characteristics such as size, GRI entry time, materiality matrix execution time, and other characteristics, which could also include other control variables (company size, GRI time, materiality matrix execution time, and others). An alternative way of future research would be to try to relate the stakeholder management beyond the materiality matrix, in order to complement the results of this study and to highlight whether or not the behavior of the field resembles the results of this study. Therefore, strategies and stakeholder groups are not similar and should not be understood as a single block.