| Abstract [eng] |
The main purpose of this master’s thesis is to assess the models for predicting bankruptcy and perform the modification of the most accurate model in order to adapt it to Lithuanian retail companies. The first part of the study, where scientific literature analysis is conducted, examines models for predicting bankruptcy and presents their comparative analysis. In order to assess the tendencies in bankruptcies in Lithuanian companies and stress the problem of increasing number of bankruptcies in Lithuanian retail sector, an overview of bankruptcies in Lithuanian retail sector has been conducted. In addition, the analysis of scientific studies on the subject of assessment of bankruptcy predicting models has been performed that helped to identify which research methods have been applied in the studies of similar character and what outcomes of the research have been achieved. Although a common opinion is observed among the researchers that models of discriminant and logistic analyses due to their simple application and accurate results of prediction are the most popular, the analysis of the results of globally performed research revealed that there is no any single model that would predict a bankruptcy most accurately, as research results obtained when examining adaptability of the models in different countries and companies engaged in different economic activity, are different, therefore, it can be stated that certain models may be unsuitable to companies operating in a single sector, while the same may be suitable to companies engaged in other activity. The second part of the study presents research methodology and progress, based on scientific literature analysis that covers two key stages of the research – assessment of bankruptcy predicting models and modification of the most accurate model selected in order to adapt it to analysis of bankruptcy in Lithuanian retail companies and enhance accuracy of bankruptcy predicting. In the third part of the study, the assessment of the classical models of discriminant analysis and logistic analysis of bankruptcy predicting is presented. The assessment of models has been conducted using financial data of 56 successfully operating Lithuanian retail companies and 26 companies that have gone bankrupt in the same sector. On a basis of the data of the solvency and profitability relationship matrix and the calculated probability of bankruptcy, the accuracy of the bankruptcy predicting models being analysed has been assessed. The outcomes of the first part of the study have shown that discriminant analysis models for predicting bankruptcy are more accurate in Lithuania’s retail sector bankruptcy analysis, while Altman’s model for predicting bankruptcy identify bankruptcy threats most accurately and early. In the second part of the research, having applied the statistical discriminant analysis method and financial data of Lithuanian retail companies, a modification of the Altman’s model for predicting bankruptcy took place. The outcomes of the second part of the study have demonstrated that the potential of the modified Altman’s model for predicting bankruptcy to predict the probability of bankruptcy in Lithuania’s retail sector companies within two and three years before the bankruptcy, in comparison to the original Altman’s bankruptcy model, has increased. Thus, during the present research, a modified Altman’s model for predicting bankruptcy was created, with the help of which more precise analysis of bankruptcy risk in Lithuania’s retail sector could be conducted. Furthermore, the increased bankruptcy risk could be identified earlier that applying classical discriminant and logistic analysis bankruptcy models. The modified Altman’s model for predicting bankruptcy, created during the research, could be used successfully in the activity of retail companies in Lithuania, this way solving the problem of growing number of retail bankruptcies. |
