基于多目标优化的信用风险管理

Credit Risk Management based on Multi-objective Optimization

信用风险管理是银行管理的核心内容。近年虽然不乏测度银行信用风险方面的研究,但都只限于考虑风险最低的单目标模型。以追求风险与收益系统均衡为出发点,提出了测度银行信用风险的二元目标优化模型,解决了信用风险管理过程中低风险和高收益的辨证统一问题。将遗传算法引进二元目标优化模型的求解,极大地提高了模型的求解效率。

Credit risk management is of high importance to banks, financial companies and investors. Recently,many models are adopted to measure the credit risk. However,these models usually ignore the optimization of portfolio. A multi-objective model is developed,which captures not only the minimization of risk but the maximization of returns of portfolio in this paper. There are many ways to deal with multi-objective decision. But, Genetic algorithm （GA） is used as the approach to deal with these problems. Compared with the other method of Pareto optimal solution, the experimental example shows the efficiency of GA approach on multi-objective decisions, especially on two-objective decisions.