The Validation KMV Model For Credit Risk Assessment Of Listed Companies In China

Credit risk is one of the most important risks for the bank industry. Exploring an effective credit risk management method is an essential issue for both the steady development of bank industry and the whole financial industry. In June,2004, Basel Committee issued the Basel Ⅱ, which allows the superior banks utilizing their own internal risk assessment method to manage credit risk. This policy aims to help banks reach the capital adequacy ratio required by the regulatory agency and improve their efficiency in operating capital. Basel Ⅱ shows an new direction for banks in credit risk management.In our country, the skills of credit risk management in bank industry are poor. The main methods for assessing default risk are still credit scoring system and expert system, which, on the one hand, has negative impact on the efficiency of operating capital, on the other hand, increase the operational risk of banks. With the bank industry becoming more and more international, our local banks have to deal with increasingly fierce competition. In order to improve their own competitive power and credit risk management skills, modern credit risk assessment models have to be introduced and applied in our country.During the past twenty years, credit risk measurement approach has developed greatly. Excluding the traditional methods, some modern models were explored to measure the credit risk quantitatively, including the CreditMetrics, Credit Portfolio View, Credit risk+and KMV model. Whilst, the lack of formal credit rating mechanism makes it difficult to apply Credit Metrics and Credit Portfolio View in China, moreover, the default rate in Credit risk+model is hard to forecast in the credit market of China, which influences the applicability of this model.The KMV model is a credit risk measurement model explored by the KMV company. The inputs of this model are mainly based on the data of equity markets and financial reports. The stock market was created in China in the1990s and it has become increasingly regulatory through the continuous reform and development. The basic conditions for applying the KMV model are mature in China. However, there are some particular characteristics, such as the split shares structure, in China’s stock market which make it impossible to use the original KMV model directly without adjusting the parameter settings. In this paper, we will try to adjust the parameters of the original KMV model and apply the adjusted KMV model for listed company’s credit risk measurement. The validation of the model is an essential issue when testing the applicability of a model. It directly affects the judgment for a model, so the research will pay more attention on testing the validity of KMV model.This dissertation consists six chapters:The first chapter will mainly introduce the research background, research meaning, research contents and research innovation.The second chapter is the literature review. In this chapter, the development of credit risk measurement will be illustrated. Moreover, we will review the literatures about applying the KMV model in China, including both the theoretical and empirical studies.Chapter3will describe the theoretical basement and framework of the KMV model. It will also give a detailed interpretation of the parameter settings for the KMV model as applied in China.In Chapter4, we will introduce the techniques used to validate the credit risk measurement model in the previous study, containing Power Curve, Accurate Ratio, Cumulative Accuracy Profiles (CAP) and Conditional Information Entropy Ratio (CIER).Then the validation framework used in this paper will be introduced.The fifth chapter is the empirical study; it contains the description about sample selection and empirical processes as well as the empirical result analysis. The empirical process includes two sections:in the first part, we will apply KMV model to measure the credit risk of all the30sample listed companies which are divided into two groups—ST group and Non-ST group; in the second part, we will measure the dynamic credit risk of the sample company we choose.The final chapter will describe the weakness of dissertation and give advice on relative problems.The innovation of this dissertation is introducing the validation measurement of credit risk model and validation framework systematically. In Chapter four, we introduced several credit risk model validation methods, including power curve, accurate ratio, Cumulative Accuracy Profiles (CAP), Conditional Information Entropy Ratio (CIER) and other methods. What is more, we built a systematic framework of validation methods to test the applicability of KMV model using in China. We got the conclusion from this study that, for the samples wholes credit risk level is high, Z-Score model has better applicability than KMV model; for the samples whose credit risk level is relative low, KMV model is superior than Z-Score model.The main weaknesses of this dissertation are as follows:First, we selected the sample of Non-ST group based on the criterion of "special treat" in the "Rules for Stock listing" issued by Shenzhen Stock Exchange in September2008. This criterion is not absolutely same as the definition of default of company. Second, we use the original function of default point in our empirical study. While, there is not any authority having proved that the original coefficient of default point is also the most suitable one in China. Third, the relationship between asset volatility and equity volatility in our study is from original KMV model, but in fact, the relationship between asset volatility and equity volatility is still uncertain in China’s stock market. Forth, we cannot map the distance-to-default with expected default probability because of lack of default database in China, so we use distance-to-default as our final output.