Determinantes da estrutura de capital e o nível de endividamento nas empresas de capital aberto: um estudo comparativo entre Argentina, Brasil e Estados Unidos

• Ano de defesa: 2014
• Autor(a) principal: Flávio Dias Rocha
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: Regressão quantílica; Estrutura de capital; Determinantes do endividamento; Modelos com termos de interação; Administração; Capital (Economia)
• Link de acesso: http://hdl.handle.net/1843/BUOS-9MBKJ5

One of the most important subjects in capital structure theory is which factors are their main determinants. However, usual approach considers average behavior of all firms, failing to take into account a less restrictive possibility, in which low-levered firms suffer different influences from high- or median-ones when choosing their capital structure. In order to check this hypothesis, the aim of this thesis was analyze, based on a sample of three countries (Argentina, Brazil and United States), the differences of the determinants for each kind of company (with low, mid and high amounts of debt). To achieve this goal, two methods were adopted: quantile regressions and regressions with interaction terms between lagged debt and other important variables. Yet, nine theoretical hypotheses were proposed, each one relating leverage to a common variable used in tests of capital structure determinants. Sample consisted of 1,192 firms: 61 from Argentina, 296 from Brazil and 835 from United States. Data were related to the period from 1997 until 2011. From nine hypotheses, one (referred to singularity attribute) has presented no relevant evidence; two were maintained as defined in their original versions those ones associated to profitability and property concentration ; in all other cases, empirical evidences suggested the need for adjustments in original proposals. Evidences signaled the existence of differences in the way firms with distinct leverage levels choose their capital structures. Low-levered companies adjusted their leverage when changes happened in profitability (with a negative influence), non-debt tax-shields (negative), size (positive) and payout ratio (positive); mid-levered firms decisions were influenced by changes in profitability (negative relation, but stronger than the influence in low-levered firms), non-debt tax-shields (negative relation, but weaker than in low-levered firms) and property concentration (following a convex relation); and finally, high-levered firms had have as main determinants of their capital structure: profitability (a negative and very strong relation), size (positive), bankruptcy risk (positive) and payout ratio (negative). A strong tendency to adjust debt level partially towards an optimal level in low- and high-leveraged companies arises, with a minor influence in all other cases. Thus, it is possible to check the importance of evaluate separately decisions made by firms in several different debt levels, as the determinants of capital structure choice vary in sign and intensity across the sample.