Genetic Programming-based Model For Credit Risk Assessment In Commercial Banks

Credit risk is one of the most important financial risks in commercial banks. With the accelerated steps of financial globalization and continuous upgrade of financial markets’ complexity, the importance and metrical complexity of credit risk in commercial banks have also increased. Traditional credit risk assessment model can not meet the demands of financial risk management system. Further research on credit risk assessment is of great significance both in theory and practice.To begin with, the advantages and disadvantages of classical evaluation model are pointed out by reviewing the theory and methods of credit risk assessment in commercial bank at home and abroad. The features of the classical standards from the natural and wave properties are analyzed in this paper. Besides, the credit safety degree is took as credit risk metric to avoid the discreteness and hysteresis of the predict consequence.Nextly, due to its powerful heuristic search property, the genetic programming algorithm is especially suitable for resolving credit risk problem. By combination of the mathematical modeling technology and compute science, not only the model of credit risk is established, but also the model is realized and simulated on Matlab platform as well.Finally, an empirical analysis of the model is carried out. Given the financial data on credit objects offering by an anonymous commercial bank, the established model can make prediction fitted the change direction of credit risk excellently.The model constructed in this paper not only provides the basis for credit decisions but also improves the warning ability of credit risk in commercial bank, thus can be widely used in the fields of commercial bank’s loan examination and risk monitoring.