Estimation Capability of Financial Failures and Successes of Enterprises Using Data Mining and Logistic Regression Analysis

Abstract

Given today's conditions, enterprises should be financially powerful in order to achieve their goals. Therefore, estimating financial failures is quite important in terms of determining possible future financial risks of enterprises and taking the required steps. In this study, annual balance sheets and income statements of 143 manufacturing firms, which were publicly traded in Borsa Istanbul between 2010 and 2013 and which displayed continuity, were used to estimate their financial failures and successes. Data Mining and Logistic Regression Analysis methods were used in this research. Models were developed to make an estimation prior to the first, second and third years taking year 2013 as basis, and classification accuracy, in other words estimation power of these models were compared. Among all models that were suggested to estimate failures and successes of enterprises, 2012 was determined as the year with the most successful estimation power as a result of our analysis.