Um estudo dos direcionadores da distribuição de dividendos das empresas não financeiras participantes da B3

Abstract

The purpose of this study was to identify the dividend policy drivers of Brazilian companies listed on B3 through the Dividend Yield and Dividend Payout indicators, also considering these relationships given the division of B3's sectors. Among the theoretical approaches of the study, the theory of relevance of dividends, the theory of the irrelevance of dividends, asymmetric information, agency theory, the hypothesis of the life cycles of organizations, the theory of fiscal and tributary effects, clientele and specific aspects of Brazilian legislation, in which there is more than one form of distribution of earnings. Through a theoretical review that included several authors and studies for different markets, including within Brazil, variables such as size, indebtedness, shareholding concentration, corporate governance, profitability, liquidity, among others, were mapped as possible drivers of dividend policy. The sample used included 537 companies with annual data between 1996 and 2020. The initial analysis of the study points out that the Public Utility and Communications sectors are the sectors that most distribute earnings, while the Others sector can be characterized as the worst distributor. To estimate the dividend policy drivers, a panel data model and the econometric method of Quantile Regression with quantiles .25, .5, .75 and .9 were used as estimation parameters. For the estimations with the entire sample, without sector division, the results indicate that: i) the asset size is significant, positive and increasing for all quantiles of both variables; ii) earnings per share are positive and significant for all DY quantiles; iii) GDP, exchange rate and CDI are significant to explain companies' dividend policy; iv) belonging to differentiated levels of corporate governance positively impact companies that distribute less dividends and negatively impact companies that distribute more dividends; and v) variables related to profitability and indebtedness did not perform well to explain the companies' dividend policy. When we look at each sector of activity, the results show that: i) in general, the size of the asset is positive, increasing and significant for most quantiles considering both dependent variables; ii) GDP, CDI and Exchange rate show good performance to explain companies' dividend policy; iii) the control characteristics related to indebtedness, liquidity, risk and profitability present different relationships for each sector; iv) characteristics related to shareholding control and belonging to differentiated segments of corporate governance indicate the presence of a clientele and signalling effect in the dividend policy of companies, depending on the sector in which they operate.