CORPORATE CAPITAL STRUCTURE AND TAXES: SOME EVIDENCE (BANKRUPTCY COSTS, EFFECTIVE RATES)

The scientific method was followed to investigate and seek evidence to evaluate the DeAngelo-Masulis theory that there is an optimal capital structure for the firm, i.e., that firm-specific nondebt tax shields cause the firm to face a unique expected corporate tax rate and a unique expected tax benefit of debt. The testable prediction developed from the theory is that nondebt tax shields are negatively correlated with debt. If the prediction is supported, the implication for management is that there is a unique optimal capital structure for each investment opportunity of the firm, and this unique capital structure is a function of the amount of nondebt tax shields accompanying the investment. The optimal capital structure of the firm as a whole would depend on the investments the firm is making.The prediction was supported by the preponderance of the evidence. The study illustrates how two approaches to observation--behavioral and statistical--are complementary steps in the scientific method.The study involved a review of theoretical and empirical work in the area of capital structure theory an analysis of responses to an open-ended questionnaire asking managers what factors they consider in the capital structure decision and how they consider them difference-in-means tests to determine if a bivariate analysis adequately explains debt levels and to analyze the movement of debt over a one-year period formulation of predictive hypotheses from theory, prior evidence, and evidence gathered in the questionnaire and the difference tests and tests of the predictive hypotheses by means of multiple regressions performed over two time periods.