基于遗传算法KMV模型的公司债券违约风险度量研究

Study on Corporate Bond Default Risk Measurement Based on Genetic Algorithm KMV Model

基于遗传算法对KMV模型进行了修正,并运用修正的KMV模型对样本债券在2017-2018期间的违约风险进行度量。结果表明:基于遗传算法改进的KMV模型在预测公司债券违约风险方面有着不错的表现,拟合正确率远高于改进前的原模型;并且公司债券所属行业的不同会影响模型违约点的选择,从而影响KMV模型度量违约风险的效果。

In this paper,the KMV model was improved based on the genetic algorithm,and the modified KMV model was used to measure the default risk of the sample bonds during 2017-2018.The results showed that the improved KMV model based on genetic algorithm had a good performance in predicting the default risk of corporate bonds,with the fitting accuracy being much higher than the original model.Moreover,the selection of default point of the model would be affected by different industries of corporate bonds,which would affect the effect of KMV model on the measurement of default risk.