Bitte verwenden Sie diesen Link, um diese Publikation zu zitieren, oder auf sie als Internetquelle zu verweisen: https://hdl.handle.net/10419/190278 
Erscheinungsjahr: 
2018
Schriftenreihe/Nr.: 
ADBI Working Paper No. 857
Verlag: 
Asian Development Bank Institute (ADBI), Tokyo
Zusammenfassung: 
This paper verifies the impact of bank account information, such as information on deposits and withdrawals, that is not necessarily fully accounted for in conventional internal ratings and that can affect the accuracy of the default predictions of small and medium-sized enterprises (SMEs). The analysis demonstrates that the accuracy of default predictions improves when a model based on bank account information is used in addition to the default prediction model based on traditional financial information. The analysis also shows that the degree of improvement increases when the size of the company is small. For small companies, the quality of financial data is generally assumed to be low, but the bank account information model can complement the incomplete data. In addition, for small firms, the bank account information model shows better default prediction capability compared to the financial model, which implies the possibility that banks could extend loans even if only the bank account information is available. The correlation coefficients of the financial model and the bank account model are higher than 50% but not very high, suggesting that these models evaluate borrowers from different perspectives. This study suggests the possibility of analyzing credit risk more easily without past financial information, especially for small enterprises. If the bank account information model is utilized, banks can reduce credit costs and loan review times and costs, which will make SME financing more efficient and smooth. The empirical analysis in this paper focuses on SMEs in Japan, but the results can also be applied to other countries, particularly emerging countries in Asia.
Schlagwörter: 
small and medium-sized enterprise finance
credit risk analysis
big data
bank account information
JEL: 
G2
G21
Creative-Commons-Lizenz: 
cc-by-nc-nd Logo
Dokumentart: 
Working Paper

Datei(en):
Datei
Größe
614.95 kB





Publikationen in EconStor sind urheberrechtlich geschützt.