Forecasting the loan repayment behavior by banks loan receivers using discrete Markov chain

The banks charge due date and delayed payment is always one of the most important crisis in the loan lenders progression towards customers. The increase in the charges and delayed payments cause an inevitable losses to the banks. If banks estimate their future charges properly, they can prevent their capital losses by making some decisions such as surplus budget determination and confidence saving. In this research we investigated the forecasting modeling of the customers loan repayment behaviors. Theproposed model contains 4 states: 1. active; 2. payment with 1_2 month delay; 3. payment with 2_3 months delay; 4. delayed. The transmission matrix among various states has been obtained from historical data about the repayment manner of one of the domestic banks based on which the forecasting has been conducted. Results show that Markov chain model can forecast loan repayment behavior by customers well.