Resumo:
The capital structure refers to the relationship between equity capital and third-party capital (debt capital) that maximizes the value of the company, assuming that there is an association between indebtedness and profitability. Several theories seek to explain this association, among them the tax advantage theory. The tax advantage theory suggests that there are benefits from company indebtedness due to the deductibility of interest payments on third-party capital. This theory was expanded in Brazil to also consider the benefits of the deductibility of interest on equity capital, peculiar to the Brazilian reality. By focusing on the deductibility of interest, whether on equity or on third-party capital, the tax advantage theory can be expanded. Nonempirical studies show that firms in Brazil choose to postpone the payment of taxes as long as possible, which implies the availability of cash reserves and a reduced need for (further) borrowing. These available cash reserves can then be used to finance the company's activities. This research aimed to verify the effects of long-term tax liabilities, considering Corporate Income Tax (IRPJ) and Social Contribution on Net Income (CSLL), on the profitability of companies in the industrial goods sector of B3 (Brazil, Bolsa, Balcão), measured by Return on Assets (ROA) and Return on Liabilities (ROE), thus formulating two research hypotheses. Data from the industrial goods sector of B3 for the years 2014 to 2019 were collected from Economatica, covering 70 companies, to test the research hypotheses, totalling 420 observations (company-year). The Ordinary Least Squares Method (OLS) with unbalanced panel was used. The results obtained did not confirm the research hypotheses, according to which companies with higher long-term tax liabilities have higher profitability measured by ROA and ROE. There is evidence that an increase in long-term tax liabilities, in relation to total assets and shareholders' equity, is associated with a reduction in profitability measured by ROA and ROE, respectively. The use of OLS with an unbalanced panel may have led to a loss of efficiency of the OLS estimator, suggesting the use of specific methods for unbalanced panels, such as the generalized moments estimator (GMM) or the random effects model in future research. The findings and considerations of this study serve as a guide for future research.
