Özet:

This article aims to develop a financial failure risk model for businesses within the scope of the data set by using Discriminant Analysis and C5.0 Decision Tree Algorithm techniques. Industrial enterprises operating at the SME scale in Turkey were included in the scope of the study. The financial failure criterion is defined as “the company has closed the financial period with a loss for two and/or more years in a row”. According to this criterion, 20 financially successful and 20 financially unsuccessful enterprises were selected from those registered in BIST and included in the SME industry index. The financial data/rates of the selected businesses between the years 2014-2020 were used. Discriminant analysis and the C5.0 decision tree algorithm were applied to the data. As a result of the study, the test set accuracy rate of the model developed for discriminant analysis was 85.7%; for the C5.0 algorithm, this rate was 100%. In discriminant analysis, it was determined that the variables that contribute positively to the determination of financial failure are the active turnover rate and gross profit margin; the variables that contribute negatively are the age of the firms and the financial leverage ratio. In the C5.0 decision tree algorithm, it is understood that the companies with low gross profit margin, the share of financial expenses in sales, and high leverage are financially unsuccessful. In the same way, firms with a high gross margin, a share of financial expenses in sales, and a low degree of leverage were found to be financially successful.

Anahtar Kelimeler: