An empirical study on customer risk management in banking industry: applying k-means and RFM methods (evidence from two Iranian private banks)

This paper aims to study customer risk management in the banking industry. For this purpose, notions and backgrounds of customer relationship management (CRM), risk and risk management, classifying and clustering methods as well as multiple criteria decision‐making methods (MCDM) have been studied. Since, to manage customer credit risks, recognising and classifying them is critical, therefore 150 legal customers and 100 general customers from two private banks in Iran have been selected. K‐means algorithm has been proposed for clustering both general and legal customers, moreover a WRFM model has been applied to classify general customers based on customer loyalty properties. A technique for order preference by similarity to ideal solution (TOPSIS) has been used for prioritising general customers based on loyalty properties of the RFM model. On the other hand in order to calculate the relative importance coefficient or weight of loyalty properties in the WRFM method, the pairwise comparison matrix based on the analytical hierarchy process (AHP) has been applied.