EDITORIAL POLICY
Abstract:
This paper applies a logit regression model on a vast database of Spanish firms to predict their insolvency using accounting information. The database is balanced, in the sense that more than 25% of the companies included are insolvent. In order to contrast the robustness of the methodology, as well as the impact of the selected database on the results obtained by the logit model, 1000 simulations have been undertaken. In each simulation two random groups have been created, one with 80% of the firms and another group with the remaining companies. The first group has been employed to generate the model, whereas the second group has been used to test the model obtained.
This paper applies a logit regression model on a vast database of Spanish firms to predict their insolvency using accounting information. The database is balanced, in the sense that more than 25% of the companies included are insolvent. In order to contrast the robustness of the methodology, as well as the impact of the selected database on the results obtained by the logit model, 1000 simulations have been undertaken. In each simulation two random groups have been created, one with 80% of the firms and another group with the remaining companies. The first group has been employed to generate the model, whereas the second group has been used to test the model obtained.
| Download PDF | -- | Tweet |
|---|
|
Finance, Markets and Valuation (e-ISSN 2530-3163) is an Open Access research e-journal. Contents published on the journal are licensed a Creative Commons Attribution 4.0 International License (CC BY 4.0) |