| The measurement of industry credit risk is an important research topic.The fluctuation of credit risk in various industries will not only affect their internal enterprises,but also affect the whole financial market.This paper selects nine representative industries in China to measure the credit risk in each industry and explore the cyclical and contagious dependence of credit risk among industries.The main work of this paper focuses on the following three parts:The first part is the construction of the modified KMV model and its measurement and analysis of the credit risk of the industry.Considering that China is different from foreign countries in the value of non-tradable shares and long-term debt,and the data shows different fluctuation characteristics,it is necessary to modify the basic KMV model when measuring the credit risk.This paper selects the stock prices and liabilities of normal companies and ST companies in nine industries from 2016 to2021 as sample data,taking the software industry as an example.Firstly,the KMV model is modified by five methods,namely,the net asset pricing method,GARCH(1,1)model and the change of default point coefficient,and three groups of modified KMV models are obtained and their credit risks are measured.Further,their effectiveness is verified by mean difference analysis,and the optimal KMV model of software industry is determined by rank sum test and variance analysis,thus the optimal credit risk measurement results of software industry are obtained.Secondly,Using the method similar to the software industry,the KMV correction models of other eight industries are determined and their credit risks are measured.Finally,according to the credit risk measurement results of nine industries and their time series diagram of default distance,it is found that the credit risk of mining industry has increased significantly in 2016 and 2020.The recovery trend began in the second half of 2020.The credit risk of manufacturing industry decreased significantly from2020 to 2021.The credit risk of wholesale and retail industry increased obviously in2017 and 2020,but then decreased rapidly.The credit risk of software industry and transportation industry increased obviously in 2017,while it was relatively safe in other years.The financial industry and electric hot water and gas industry fluctuated sharply in each year.The reasons for their risk changes were analyzed.The second part is the analysis of influencing factors and periodic dependence of industry credit risk based on BALQR model and Copula function.Firstly,the BALQR theoretical model is deduced and constructed.When solving the model,the full conditional posterior distribution of each parameter is derived,and the corresponding Gibbs sampling steps are given.Secondly,variable selection and data explanation are carried out.The risk measurement results of nine industries from 2016 to 2021 obtained in the first part are selected as dependent variables,and nine variables such as gdp and cpi are selected as independent variables.Then,according to the theoretical model constructed above and the specific variable data of each industry,BALQR models of nine industries are established respectively.Based on the above model,this paper analyzes the influence of each industry’s own variables on its credit risk at different points.The study finds that the influence of different independent variables on the credit risk of each industry is significantly different at different points.Taking the software industry as an example,increasing investment in fixed assets can reduce the risk of the software industry during the credit risk crisis.The influence of inflation rate is generally large,and fluctuates greatly with the increase of quantile value;The influence of interest rate is generally small;Money supply plays a significant positive role in the credit crisis;In the crisis period and safety period of credit risk,GDP will have a negative impact on the credit risk of software industry;The overall impact of the Shanghai and Shenzhen 300 index yield is weak;In both crisis and safety periods of credit risk,consumer confidence index can reduce the credit risk of software industry,and the effect is remarkable in crisis period.Finally,the existence of periodic dependence is verified by comparing the upper-tail and lower-tail result graph obtained by the above BALQR model with the industry tail coefficient calculated by using binary copula,and a series of periodic dependence results are obtained.The study finds that there is tail dependence of credit risk among industries.The tail dependence of some industries is caused by macro variables,such as manufacturing and construction,financial and transportation,which verifies the existence of periodic dependence.However,the tail dependence of some industries is inconsistent with the periodic response of macro variables,such as construction and transportation,wholesale and retail and real estate,which indicates the existence of infectious dependence between industries.The third part is based on the dynamic vine Copula model to study the infectious dependence between industries.First,the unconditional normalization measure between industries is calculated,and the industry pairs with large correlation are selected as the research object.Second,the tree structure of the model is determined by Spanning tree algorithm,and then the conditional normalization correlation measure between industries is calculated.Finally,By comparing the results of unconditional Pair Copula and conditional Pair Copula,the existence of contagious dependence is confirmed,and it is found that some industries can catalyze the credit risk contagion.For example,when the credit risk of mining industry is known,the credit risk correlation between manufacturing industry and electric hot water gas industry will be enhanced,while some conditional industries play an isolation role in the credit risk contagion between target industries,for example,when the credit risk of wholesale and retail industry and real estate industry is known,The correlation between electric hot water gas industry and financial industry will be weakened.These findings have certain reference value for studying the credit risk contagion mechanism and understanding the correlation and risk contagion between different industries. |
PDF Full Text Request