Credit Risk Measurement And Analysis Of Listed Companies Based On Improved KMV Model

With the deepening of supply side structural reform and the breaking of "Rigid payment",we are gradually entering an era of more strict control and unified supervision,the financial industry is developing rapidly under the background of economic globalization.At the same time,the contradictions accumulated in the process of market expansion are becoming more and more obvious.In terms of bond market,there has been a disorderly wave of defaults since 2014.This inevitability caused by reform prompts us to be alert to the hidden dangers of credit risk.Establishing a credit risk measurement system that suitable for our country becomes an important measure in the market.The KMV model’s visibility,high availability of data,and the excellent characteristics of the results make it been widely recognized and used in different eras.The traditional KMV model is based on a large number of default market data in the U.S.,the setting of default coefficients reflects the effect of the debt structure in American companies on the ability to repay under the event of crisis.Nowadays,our country lacks a lot of default market data,our financial market is also immature,these make the default distance and default probability of the Chinese market are quite different from that in America.Therefore,the default level is also a very important parameter that needs to be re-considered when establishing the KMV model in the Chinese market.In the fact that default level is a core to the KMV model,starting from the reality of our country to find suitable coefficients for a more accurate measurement is very important to guard against default and to build a more perfect credit system.The purpose of this article is to find a better default point.The first part of the article illustrates the adverse effects caused by default events in the market,summarizes the development history and research results of credit risk measurement model,and mainly introduces the improvement and development of KMV model based on the latest research.The second part of the article introduces the modeling steps,their respective characteristics and application situations of the credit risk models commonly used by individuals and institutions.The third part is the main content of this paper,which introduces the Black-Scholes-Merton model in detail.KMV model is based on the Black-Scholes-Merton model.This part also introduces the modifications of the parameters by different experts.We propose genetic algorithm to modify default point,and to enumerate the common selection methods,crossover and mutation operators in the genetic algorithm toolbox in chapter 4.The last part of the article combines KMV model and genetic algorithm,selects the ST and similar non ST company in 2015 and 2016 as the training sample set,the test sample set is selected in 2017.In the B-S formula,we use the non-deterministic default level,so that it can update with the default point in the reproduction process.Finally,we can see that the default coefficients of the A-share market in China are 3.259 and 1.104 in the last two years,and the accuracy of the classification of default distance is increased from 73.3%to 76.7%.The difference of default distance between normal enterprises and companies under special treatment is significant.The ROC curve and AUC value are very good;the calculation results of the asset value,volatility of asset value and probability of default are also more rigorous.On the basis of empirical analysis,I evaluate the model and results.This paper provides some reference for the development of credit risk system.