| Abstract [eng] |
96 pages, 32 tables, 26 pictures, 91 references, 22 appendices. The main purpose of this master thesis is to develop and evaluate a credit risk assessment model focused on decision-making for small and micro enterprises. The work consists of three main parts; the analysis of literature, the research and its results, conclusion and recommendations. Literature analysis reviews the credit risk theories, presents the main credit risk models, introduces the concept of small and micro enterprises, identifies their economic importance and the credit challenges due to their characteristics as well as indicates the potential and critical approach of automation in the credit decision-making process in the choice of assessment methods. After the literature analysis the author has carried out the study on the formation of a credit risk assessment model, taking into account the results of the analysis of the clients of an alternative financier operating in Lithuania during the reporting period. The model is built using a rule-based approach that incorporates client and offer information types through automation and allows it to act as an instantaneous decision-making system integrated with the financier's credit risk policy rules. The data file is based on information provided by the credit bureau ‘Creditinfo Lietuva’, the internal database and information provided by the client. The SPSS software package is used for data processing. The non-parametric criterion Chi-square (Pearson formula) is used to determine the relationship between categorical variables. The results of the study show that the credit risk assessment model developed for the credit decision works on a 100% compliance basis. The logical input-output set of the model, in terms of the rules defined, is ensured and allows to explain the decisions taken in the individual information segments. The key findings of the study, which reveal the benefits for the financial services provider, are that the automation process explaining the differences between human and model decisions guarantees that the credit risk model, as a primary assessment tool, can contribute to the impact in terms of improving the accuracy of the decision-making process regarding accuracy, compliance, objectivity and generalizability. The conclusions and recommendations summarize the main concepts of literature analysis as well as the results of the performed research. The author believes that the results of the study can provide useful guidance for companies considering the implementation of a credit risk assessment model and could help the company to implement the system in a real business environment. |
