Determining Consumer Default Risk with Data Mining Techniques: An Empirical Analysis on Turkey

The aim of this study, which deals with consumer default risk, is to reveal the financial, socioeconomic, and demographic determinants of default risk at household level. Credit risk was investigated with various variables by applying data mining methods to the data set obtained from the Turkish Statistical Institute, Household Income and Living Conditions Survey covering the years 2016, 2017, 2018. Analyses were carried out using the WEKA data mining program. The findings of the study revealed that variables such as gender, age, marital status, education level, health status, employment status, region of residence and income status are important determinants of default. The findings of the study are thought to be an important reference for lenders in terms of risk assessment. In addition, the findings are expected to shed light on policy makers in terms of regulations to be applied to financial markets.