A influência da governança corporativa no desempenho e na estrutura de capital das empresas listadas no N1, N2 e novo mercado da Bovespa

Abstract

Corporate governance is a broad subject, but, generally, it is related to an organization, its relation to shareholders and stakeholders such as customers, employees, suppliers, comunity, etc. The governance become the focus of the organization´s benefits when the enterpreneurs recognize that its practices could contribute to solve divergences between directors and shareholders, generated by the dilutions of the stock market´s property. In Brazil, the New Market is as example of a more rigorous politics of corporate governance, the level 1 and level 2, created by Bovespa. This study aims to evaluate if the adoption of corporate governance practices modify the capital´structure and the companies´performance. The sample is formed by 84 Companies, which possess stock markets negociated by Bovespa and listed on Level1, Level 2 and in New Market until December, 31st, 2006. This is a quantitative and descriptive research. It uses tests of differences to identify the period of the governance´s occurance. From the effect´s identification, it was applied a multiple linear regression, aiming to evaluate if the corporate governance influenced the variation in the structure in the Capital and the Companies´performace. The recurrent information from the administrative´s report bring out by Bovespa has been organized in a management´s index. Futhermore, the independent variables were formed by 5 of the major shareholders, the porcentage of stock market of 5 major shareholders and the board´s independence. The results showed little influence by the independent variables in the dependent variables. In this case, it can be inferred that the practice of corporate governance affects the performance and the structure of the companies´capital, at least, in this limited sample. In addition, to most of the models used, the insertion of controlling variables did not improve the quality of the models.