Nonparametric models of financial leverage decisions
This paper investigates the properties of nonparametric decision tree models in the analysis of financial leverage decisions. This approach presents two appealing features: the relationship between leverage ratios and the explanatory variables is not predetermined but is derived according to information provided by the data, and the models respect the bounded and fractional nature of leverage ratios. The analysis shows that tree models suggest relationships between explanatory variables and the relative amount of issued debt that parametric models fail to capture. Furthermore, the significant relationships found by tree models are in most cases in accordance with the effects predicted by the pecking-order theory. The results also show that two-part tree models can accommodate better the distinct effects of explanatory variables on the decision to issue debt and on the amount of debt issued by firms that do resort to debt.
|Date of creation:||Sep 2010|
|Date of revision:|
|Contact details of provider:|| Postal: na Rua do Quelha 6, 1200-781 Lisboa|
Web page: http://cemapre.iseg.utl.pt/
More information through EDIRC
When requesting a correction, please mention this item's handle: RePEc:cma:wpaper:1005. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Helena Lima)
If references are entirely missing, you can add them using this form.