The Credit Risk Measurement Of Chinese Commercial Banks Based On LSD

Credit risk has become the main risks that the commercial banks have ever faced. The study of credit risk measurement is helpful to make effective risk prevention mechanism and reduce credit losses. According to the current reaseaches of credit risk for the domestic commercial banks, this paper aims to develop a scientific, effective and more practical credit risk measurement model to improve the accuracy of risk measurement, as well as provide some reasonable policy suggestions for the existing problems of the risk measurement for commercial banks.First of all, we mainly summarize the current reaseaches of credit risk for the domestic commercial banks, and especially present the commonly used credit risk measurement models meanwhile compare the advantages and disadvantages of the introduced models. In order to improve the accuracy of the credit risk measurement, combining the domestic actual situation and the advantages of the various credit measurement models, we propose a novel model based on data fusion named LSD to estimate the credit default risk for the domestic commercial banks.To better show the effectiveness and accuracy of our proposed model, we have a comparison of the estimation accuracy for the LSD model, Logit model and support vector machine based on the financial data. The results show that the LSD model is more effective and accurate than the Logit model and support vector machine, which means that our LSD model is feasible to measure the credit risk (refer to default or no default).Finally, based on the results of the LSD model as well as the real social and economic environment, some reasonable and specific suggestions are provided to overcome the challenges of the credit risks so as to improve the international competitiveness of Chinese commercial banks.