On the appropriate thresholds in the logistic regression bankruptcy model

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F62156489%3A43110%2F24%3A43925249" target="_blank" >RIV/62156489:43110/24:43925249 - isvavai.cz</a>
Result on the web
<a href="https://doi.org/10.1063/5.0210401" target="_blank" >https://doi.org/10.1063/5.0210401</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1063/5.0210401" target="_blank" >10.1063/5.0210401</a>

Result language
angličtina
Original language name
On the appropriate thresholds in the logistic regression bankruptcy model
Original language description
This article focuses on the identification of a potential threshold for logistic regression models for bankruptcy prediction in the forestry sector in cases where the ratio of active and bankrupt companies in the dataset is not balanced. Model estimates are performed on a random selection of 300 active forest companies in EU countries along with eight (and 17) bankruptcy companies in 2019 (and in 2018). Analyzes are performed on the basis of accounting (micro) data of companies from the Orbis database. Empirical results show that in cases where bankruptcies are in the minority, it is necessary to replace the commonly used threshold with an optimized threshold, which will emphasize the balancing of the error rate in both the group of active and the group of bankrupt companies.
Czech name
—
Czech description
—

Publication year
2024
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Similar results(10)