Kredito rizikos vertinimo modelis Lietuvos kredito unijoms: santykinių rodiklių ir jų analitinių kriterijų parinkimas

Direct Link:
Mokslo publikacijos / Scientific publications
Document Type:
Straipsnis / Article
Lietuvių kalba / Lithuanian
Kredito rizikos vertinimo modelis Lietuvos kredito unijoms: santykinių rodiklių ir jų analitinių kriterijų parinkimas
Alternative Title:
Credit risk assessment model for Lithuanian credit unions: selection of relative ratios and analytical criteria
In the Journal:
Vadyba [Journal of management]. 2013, Nr. 1 (22), p. 135-142
Kreditas. Paskolos / Credit.
Summary / Abstract:

LTReikšminiai žodžiai: Kredito rizika; Kredito rizikos vertinimo modeli; Kredito rizikos vertinimo modelis; Kredito unijos; Santykiniai finansiniai rodikliai; Bankruptcy prediction; Credit risk; Credit risk assessment model; Credit unions; Financial analysis; Relative financial ratios.

ENIn this article, relative ratios and the analytical criteria thereof of the credit risk assessment model for small and very small businesses are presented. In the article, small and very small businesses are understood as those defined in the Law on Small and Medium-Sized Business Development (Official Gazette 109-2993). Lithuanian and foreign scholarly literature which analyses credit risk assessment, quantitative solvency analysis, and bankruptcy and insolvency prediction models, is presented and analysed in the article. The relative financial ratios used in the credit risk assessment model are selected in several steps. First, in carrying out single-criterion analysis, a long list of ratios is compiled; second, in accordance with the principles of model compilation, the relative ratios which are used in the model are chosen from said list. Nine relative financial ratios are selected, which include solvency, capital structure and profitability ratio groups. In order to effectively assess credit risk, analytical criteria are prescribed for each relative ratio selected for use in the model, allotting the values of the ratio into five groups, from the best value to the worst. Based on the relative ratios presented in the article as well as the analytical criteria thereof, an aggregate solvency index is proposed which can be calculated at credit unions and other institutions which provide financial and consumer credit or deferred payment for small and very small businesses in Lithuania. In the article, the author presents an index ranking and solvency limits table, according to which the index values are divided into 10 ranks, and financial status data is assigned to the ranks.Assessment of the reliability of the proposed aggregated solvency index was carried out by comparing its effectiveness with those of logistic regression – the insolvency probability calculation models of Zavgren and Chesser, as well as multiple discriminant analysis models - Altman's Z-score formula and the models for predicting bankruptcy formulated by Lis, Springate, and Taffler and Tisshaw. During the study, 25 companies that correspond to the concept of small and very small businesses were analysed. The analysed companies either have or had a loan in a credit union, or have gone bankrupt. How many first and second types of errors are made by the models was investigated, i.e. how many reliable customers are ascribed to the unreliable customer group and vice versa. The results obtained were summarized, and with the help of the research it was proven that the solvency index presented in the article can be used as an alternative to the indices and/or bankruptcy prediction models used by solvency agencies operating in Lithuania. The results of the study on aggregated solvency index performance that was carried out are shown graphically, using ROC curves, from which it is evident that the proposed index works effectively - the curve diverts from the middle towards the upper left corner. The ROC curves also show that the proposed index for small and very small businesses, when analysing the data of one year's financial statements, is more effective than the multiple discriminant analysis and logistic regression models. A full credit risk assessment model for Lithuanian credit unions covering both quantitative and qualitative factors will be published in the author’s subsequent works. [From the publication]

Related Publications:
2021-01-25 11:45:44
Views: 34    Downloads: 20