Theories suggest that firms select capital structures depending on attributes that determine the various costs and benefits associated with debt and equity financing. With the object of proving this statement, Titman and Wessels (1988) suggested a method that assumes that, although the relevant attributes are not directly observable, we can observe a number of indicator variables that are linear functions of one or more attributes and a random term. In this article, we use a method very similar to this one, using Factor Analysis to extract a measure of correlation between the variable chose as proxy and the attribute of interest. The empirical results showed that, in Mexico, large firms use more long-term debt; profitable firms and growth firms use less short-term and long term debt; firms which have more non-debt tax shields use less short-term debt; but, without explanation, firms which face high volatility of earnings use more short and long term debt. In Argentina, just one relation: profitable firms use less short-term debt. In Chile, small firms use more short-term debt and large firms use more long-term debt; firms that have high levels of tangible assets use more short-term and long-term debt; profitable firms use less long-term (and also short-term) debt.
equity; debt; capital structure; factor analysis
