Problems in Measuring Capital Structure: Empirical Evidence in Brazil

Abstract

This empiric-analytic paper, aims at verifying whether the firms’ capital structure is sensitive to leverage measures. It also verifies whether most of the leverage variability is due to financial debts. The sample consists of all non-financial companies with data available, positive equity and stocks listed in Bovespa between 1995 and 2007. Accordingly, we sampled on average 133 companies per year. The data required for the analysis were extracted from the Economatica database. In the methodology section we used cross-section linear regression in order to estimate the coefficients and other statistics of interest, following the methodology of Fama and MacBeth (1973). As main results, more than half of the cross-section heterogeneity of the leverage ratio comes from variations on the Non-financial Debts/Total Assets ratio. With respect to the sensibility of the measures used for the leverage, we conclude that profitability is the only variable that is reasonably consistent, in magnitude, sign and significance, with the different measures of leverage.