Classification of Legal Customers and Prediction of their Profitability using Customer Lifetime Value and Markov Chain (Case: A private bank in Iran)

Nowadays, customers are the most important sources of income for financial institutions and banks. According to the privatization process in the country and financial restrictions of banks, it is necessary to maintain and attract more profitable customers. Though, one of the most important methods to identify profitable customers is the concept of customer lifetime value (CLV) but it is more important to estimate customers’ future conditions because a bank`s future profitability highly depends on the customers situation.
In this research, the issues about CLV, the necessity and different classification methods are presented. Then, considering weighting variables using Recency, Frequency, and Monetary (RFM) model, AHP technique, and experts opinion, customers are classified. Using Markov chain and probability matrix, the displacement of customers in different groups and their future status are predicted.
One of the major outcomes of this research is the calculation of profitability matrix to predict customers’ displacement in different groups. The probability matrix can also show the reluctance of large number of customers to move to the specified groups (the highest percentage of customers in the main diameter of the probability matrix). In this research, we observed that account balance (M) has the greatest impact on customers grouping and that the number of transactions (F) and recency variable (R) are ranked as the second and third impact factors. Also, the determination coefficient (C) is another result of the research. Finally, the presented research used financial information and proposed a mathematical model (Markov chain) to calculate the probability of customers’ displacement (switching from one group to another). The proposed model can be helpful to facilitate customer relationship management process in the banking system.