Using Discrete Markov Chain Model for Predicting the Behavior of Banks Loan Portfolios

The main goal of total commercial banks is collect the saving of real and natural persons and allocate them in the form of facilities to industry, service and manufacturing companies. with the Non repayment of facilities from side of customers, the banks is faced with many problems such as disable in repayment of central banks, increasing the amount of facilities from customers repayment amount and disable in granted facilities. With increasing delayed demands and delaying in repayment of loans, the necessary of optimal allocation of facilities and investigation of bank credit portfolio is revealed. therefore, in current research, the modeling of predict the behavior of bank loan portfolios by using of Markov chain with limited state is investigated that includes (1) active loan or alive, (2) loan with one to three month delay in repayment, (3) delayed loan .transitional matrix between different states by historical information from the melli banks’ mortgage portfolio is achieved and then the timely prediction of payment number delays in repayment and Non repayment of facilities granted to certain portfolio is conducted. The results indicate that the proposed Markov model have the ability to accurately predict the behavior of credit bank portfolio. Key words: bank credit portfolio, Stochastic processes, discrete Markov chain.