A proposal of sector-specific financial failure forecast model

Although there is no generally accepted definition of the concept of financial failure, it is a general concept that expresses the situations where individuals, businesses or organizations cannot reach the financially determined targets or the financial performance is not at the desired level. Financial failure affects not only the business where the problem arises, but also the social, economic and political environments. In order to prevent the problems arising from this situation, many financial failure prediction models have been created. The efficiency of these models, which were developed using a limited number of samples, regardless of sector, is questioned in today's conditions. Within the scope of this study, sector-specific financial failure prediction models were developed with the help of alternative methods using a large data set for each sector. Foresight models were obtained by using 4 different definitions of financial failure in 17 sectors whose data were shared by the CBRT. The financial ratios to be included in the sector-specific models were selected with the help of 3 feature selection methods. The performances of the prediction models developed specifically for the sector were performed with the help of four different classification methods. In the models developed using the concepts of profitability, net working capital and equity, the financial structure and dynamics of each sector were included in the forecasting models. The average accuracy rate of sector-specific forecasting models was 95.7%, and the precision rate was 98.1%. When a general evaluation is made in terms of the methods used in sector-specific financial failure prediction models, the Random forest method has made much more successful classifications than other methods.