Predicting Financial Distress in the BIST Industrials Index: Evaluating Traditional Models and Clustering Techniques

Abstract

Financial distress, which can lead to bankruptcy or liquidation, is important for companies, creditors, investors, and the economy. Recent financial crises and global economic fluctuations have brought this issue to the forefront. In an effort to foresee financial distress, methods like Altman's Z-score have been proposed while, recent developments have allowed for the incorporation of recent techniques like machine learning. The purpose of this study is to forecast the emergence of financial distress in BIST Industrials Index (XUSIN) companies by using the k-means clustering algorithm, Altman Z-score and Springate S-score models with firm level financial indicators where we investigated successful and unsuccessful companies. Our findings show that two companies met all three Altman Z-score, Zꞌ-score, S-score and financial situation criteria in 2011, 2012, 2015, and 2017; 2 companies in 2016 and 2018; 5 companies in 2013 and 2014; 4 companies in 2019; 1 company in 2020 where no companies are grouped in the same groups in 2021, which means the methods reach different results. It has been determined that the k-means clustering algorithm, particularly due to its higher separability, provides more accurate clustering results for the concerned parties compared to other methods.