Predicting the Financial Distress of Companies Listed on the Tehran Stock Exchange Using DEA-DA Technique and Artificial Neural Network

This study aims to identify financial criteria to evaluate and analyze the financial distress of companies listed on the Tehran Stock Exchange and the dynamic forecast of corporate financial distress. Therefore, after a comprehensive review of the research literature and the main financial ratios used in previous studies, eight financial ratios widely used in previous research were selected. The research data was collected from the Exchange and Securities Organization data sources and existing data systems such as Tehran Stock Exchange, Codal.ir website, and Rahvard Novin software related to 106 companies. Then, the clustering process was performed for 105 companies using the SOM artificial neural network method. In this study, the number of existing clusters was considered equal to two clusters (financially distressed and non-distressed companies). After clustering the companies, the proposed DEA-DA model was implemented. Finally, the membership of the new company was predicted in the appropriate distressed or non-distressed cluster. The study's results indicated that the membership of the new company in the financial distress cluster was correctly predicted, and the proposed method made it possible to dynamize the financial distress forecast for decision-makers including corporate managers and investors by considering various economic and financial criteria.