University of Macedonia Economic and Social Sciences
Η επιλογή λογιστικών μεθόδων και τεχνικών στην αποτελεσματικότητα των υποδειγμάτων πρόβλεψης της εταιρικής πτώχευσης
Abstract
dc:descriptionThe main objective of this thesis is to explore the relation between corporate bankruptcy the choice of accounting methods and techniques as well as the impact of such managerial decision on the efficiency of the corporate bankruptcy prediction models. In this study, multivariate discriminant analysis and logit analysis are employed for the construction of corporate bankruptcy prediction models with the incorporation of financial and accounting variables.The findings of the study provide evidence of significant relation between corporate bankruptcy and the choice of accounting methods and techniques. Moreover, the incorporation of an accounting choices portfolio in the corporate bankruptcy prediction models enhances their efficiency in terms of correct classification and both types of errors.The implications of the current study consist in the enhancement of the efficiency of the existing corporate bankruptcy prediction models which will lead to more accurate credit risk assessment of the banks and finally it will mitigate the bad debts of the banking system. Regulatory authorities should reconsider the current framework in the light of new evidence (Basel III, choice of accounting methods). The results of this thesis also provide evidence in support of Contracting Cost Theory and opportunistic behavior of managers, at least for certain accounting variables which we define as “prodrome” indices since they are signaling the future course of the corporation more timely and more accurately.
Degree
thesis:*- Grantor dc:publisher
- University of Macedonia Economic and Social Sciences
- Year dc:date
- 2015
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Kosmidis, Kosmas
- Κοσμίδης, Κοσμάς
Subjects
dc:subject × 10Rights
dc:rights- Statement dc:rights
-
- BY_NC
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/39423