Research On Credit Risk Assessment Of Listed Companies In Shandong Province Based On Combination Forecasting Method

Credit plays a pivotal role in the modern economy and society.Credit risk is an important part of financial risk.Accurate assessment and control of credit risk is of great significance to promoting the transformation of new and old kinetic energy in our province and maintaining stable economic development.At present,domestic and foreign researches on credit risk have been relatively sufficient,and a large number of credit risk assessment methods have been proposed.Regarding the combined forecasting method,the existing literature has also conducted some discussions and applications.However,there is still a lack of a unified evaluation result evaluation system to systematically compare and study different methods,and the application of the combination prediction method in the field of credit risk evaluation needs to be further deepened.Many scholars have also conducted detailed and in-depth research on the credit risk of enterprises in Shandong Province,especially state-owned enterprises.However,there are still some shortcomings in the discussion of the temporal and spatial distribution,industry distribution,related factors,and evolution trends of the credit risk of listed companies in Shandong Province..In this thesis,the blank control group and the standard control group are firstly introduced,and an evaluation system for the quantitative evaluation results of credit risk is constructed.The evaluation indicators are reasonably selected,and seven methods such as AHP,KMV,three principal component regression methods,decision trees,and random forests are used.The single method and the combined forecasting method of the four weight determination methods,including the least square method,the inverse error method,the equal weight method,and the correlation coefficient matrix method,are used to evaluate the credit risk of listed companies in Shandong Province for 44 quarters from 2010 to 2020.Quantitative evaluation and comparative analysis of the performance of different methods in reducing the two types of error rates in-sample and out-of-sample.The results show that different single methods have their own advantages and disadvantages.The results of the three principal component regression methods are relatively close,which can achieve a good balance between accuracy and robustness.The gap between the KMV method and other methods is large,and the accuracy has obvious disadvantages.Random forests have high in-sample accuracy but poor out-of-sample performance,while decision trees perform worse than random forests.The robustness of the AHP method is high but the accuracy is insufficient,and the effect of reducing the overall error rate is better than reducing the first type of error rate.Among the four combined forecasting methods,the least squares method has high accuracy and poor robustness.Both the equal weight method and the correlation coefficient matrix method can achieve a balance between accuracy and robustness,while the inverse error method is between the equal weight method and the smallest method.between the two methods.The weighted combination using the correlation coefficient matrix method is mostly superior to each single method in and out of the sample.Based on this,this thesis believes that the principal component regression performs better in a single method,and the combined prediction method with more average weights is more robust and accurate than the single method.According to the evaluation results,the credit risk of listed companies in Shandong Province is highly correlated with the debt repayment and profitability of enterprises,and is affected by the province’s resource endowment,comparative advantage and macroeconomic cycle.Due to industry distribution,the credit risk of state-owned listed companies in Shandong Province is higher than average.And compared with non-state-owned enterprises,there are certain differences in the factors related to the credit risk of state-owned enterprises,and the change of credit risk of state-owned enterprises is generally earlier than that of non-state-owned enterprises.In this regard,this thesis believes that it is necessary to explore the establishment of a more effective credit risk assessment system,so as to better play the role of the government,provide a more friendly business environment for the development of enterprises,and effectively control the credit risks of enterprises in Shandong Province,especially state-owned enterprises.