Please use this identifier to cite or link to this item:
Przanowski, Karol
Year of Publication: 
[Journal:] e-Finanse: Financial Internet Quarterly [ISSN:] 1734-039X [Volume:] 9 [Year:] 2013 [Issue:] 1 [Pages:] 44-59
This paper presents two cases of random banking data generators based on migration matrices and scoring rules. The banking data generator is a breakthrough in researches aimed at finding a method to compare various credit scoring techniques. These data are very useful for various analyses to understand the complexity of banking processes in a better way and are also of use for students and their researches. Another application can be in the case of small samples, e.g. when historical data are too fresh or are connected with the processing of a small number of exposures. In these cases a data generator can extend a sample to an adequate size for advanced analysis. The influence of one cyclic macro-economic variable on client characteristics and their stability over time is analyzed. Some stimulating conclusions for crisis behavior are presented, namely that if a crisis is impacted by both factors: application and behavioral, then it is very difficult to clearly indicate these factors in a typical scoring analysis and the crisis becomes widespread in every kind of risk report.
credit scoring
crisis analysis
banking data generator
retail portfolio
scorecard building
predictive modeling
C 53
Document Type: 

Files in This Item:

Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.