Empirical Analysis of Credit Risk Measurement in Commercial Banks:Credit Risk Measurement and Risk Management in Commercial Banks

I. Literature Review The credit risk of commercial banks mainly refers to the possibility of loss incurred by the creditor due to the default of the borrower who is unable to repay or unwilling to repay for various reasons. With the continuous innovation and development of the financial industry and the financial market, the competition in the financial market has become increasingly fierce, prompting the financial market at home and abroad the measurement of credit risk has become the focus of many scholars' research.

Credit risk measurement models include traditional measurement and modern measurement. The common measurement models are as follows: ZETA model, MDA model, Z scoring model, Logit model and neural network model belong to the traditional measurement models; while the most popular and researched models are Credit Metric model, Credit Risk+ model, KMV model, Credit Portfolio View model, KPM model are categorized as modern measurement models. In the West, the research and summary of credit risk measurement models using simulation began in 2000, Gordy and Crouhy (2000) and others conducted the corresponding evaluation of the measurement of the same point in time, i.e., static asset portfolios, and concluded that the results of the data analysis through different models are close. Meanwhile, Nickell (2004) according to the comparative analysis of risk measurement models to study, mainly for the actual assets of the portfolio data for empirical analysis, but the results verify the possibility of the model's assessment of risk is not accurate enough. Similarly, Zhang Zongyi (2009) and others conducted a comparative analysis of traditional and modern models for the credit risk of more than 1,000 lending enterprises in Chongqing Municipality, with data from four state-owned commercial banks in Chongqing Municipality. By analyzing the data, Zhang Zongyi came up with results close to those of Nickell (2004), and the conclusions of the study also show that for the prediction of risk, it is weak to rely on the model to derive the effect of significance. Another point of view is that the KMV model proposed by the American KMV company in 1993, i.e., using the idea of options and eliminating the Credit Monitor Model in the middle, the current practice in China shows that the KMV model has the widest scope of application, which is mainly attributed to the continuous standardization of the management of China's capital market, as well as the market value of the listed companies and their intrinsic value, and the application of the KMV model is constantly close to the market value. The main reason is that China's capital market management is constantly standardized and the market value of listed companies and their intrinsic value are constantly close to each other.

The KMV model is widely accepted and applied by academics and practitioners, and after the emergence of new features of financial development in China, the extensive study of the KMV model has strong theoretical significance and practical significance. As we all know, the main theoretical basis of KMV model is Merton option pricing theory, the input variables used are the market value of the enterprise's equity and its volatility, and at the same time, according to the liability status of the target enterprise to calculate the enterprise's default implementation point, and in this way, the default distance is calculated, and then according to the relationship between the default distance and the expected default rate of the two, to arrive at the enterprise's default rate, and the default rate is used as the enterprise's credit risk. The default rate is taken as the credit risk of the enterprise. Recently, many scholars in China have begun to conduct theoretical and practical research on the KMV model according to the Chinese market. Liu Yingchun (2004) and other scholars agree that foreign scholars on the use of the KMV model is based on the modern option theory, based on which the disclosure of relevant information in the capital market is predicted, the relevant information excludes the market's historical book reports, and the predicted results actually reflect the current creditworthiness of listed companies in the capital market more than the current creditworthiness of listed companies in the capital market. The prediction result is actually more reflective of the current credit risk situation of listed companies in the capital market. However, Liu Yingchun's study also has its shortcomings, that is, the application effect on non-listed companies is very poor, mainly focusing on listed companies. Wu Hengyu (2005) and others believe that the EDF indicator in the KMV model is mainly for the real-time data display of stock prices, and has continuity in the measurement of corporate credit risk. At the same time, the KMV model provides the relevant regulatory authorities and investors with a more reliable credit analysis and measurement model; the limitation of the model is that there is a lack of historical data on corporate defaults and related bankruptcies in China's capital market, and it is not possible to convert the distance between defaults and the expected default rate, which affects the wide application of the KMV model in China.

In summary, the theoretical study of the KMV model of the literature is relatively rich, but at present the empirical study of the KMV model of the domestic literature is relatively small. Therefore, this paper applies the KMV model in combination with the characteristics of the current development of the financial industry, and finally derives the probability of the loss of the EDF indicator by predicting and estimating the equity value and its volatility of all domestic listed banking companies.

II. Credit Risk Measurement Model Setting and Variable Selection

First, the introduction of credit risk measurement model. In the 1990s, some financial institutions at home and abroad began a related technical research on corporate credit risk, mainly researching some new credit risk measurement models, mainly including such as KMV model, Credit Metrics model, Credit Rist+ model and so on. The modern risk measurement model is based on the theoretical risk measurement in the financial market, and at the same time, mathematical and statistical methods and systematic engineering research methods are added into the model, which is mainly for analyzing, identifying, measuring and monitoring the risks faced by commercial banks.The KMV model was put forward by the KMV company in 1989, which is mainly for providing services such as credit risk management, etc. The model is based on the stock price in the stock market as the basis. The KMV model, also known as Credit Monitor, is a credit monitoring model that estimates equity based on the option pricing theory and predicts the likelihood of default of listed companies by analyzing the volatility of their equity value, i.e., their intrinsic corporate value, and its application is to listed companies. The data of listed companies are easier to obtain.

According to the KMV model, the relevant indexes are defined: the total value of the company is the equity value (owner's equity) of the enterprise plus the value of debt; according to the balance sheet, when the total value of the company is greater than the liabilities of the enterprise, the creditors of this portion of the debt will be repaid in full, and the shareholders will be given the remaining portion of the value; when the total value of the company is lower than the value of the debt, that is to say, when the assets and liabilities are not paid, the company will not be able to repay the debt. , the company will not be able to repay the debt, that is, there will be default, at this time, the value of shareholders is zero or even negative. Therefore, if the total value of the company is lower than a certain value, there will be a risk of default to shareholders and creditors. At this point, the corresponding total value of the company is set as the point of corporate default, that is, the point when the value of the company's assets is equal to the value of liabilities.

The model assumes that the total value of the firm's assets obeys a distribution characterized by the expected value (E) and standard deviation (volatility) of the firm's total assets in a given period. The difference between the mean of the future value of the firm's assets and the book value of the firm's liabilities is set as the Distance to Default (DD). Determine the corresponding linear relationship between the expected default rate and the Distance to Default, i.e., express the default rate in terms of the Distance to Default, establish a functional relationship, and estimate the corresponding default rate. The expected default rate indicates the probability that a firm may default over a certain period of time under normal market conditions.The KMV model assumes that default occurs when the mean value of a firm's future asset value is lower than the book value of the liabilities that the firm is required to pay off. Since it is not possible to determine precisely whether a default will occur for a lending firm, it is only possible to estimate the risk of default, i.e., the magnitude of the likelihood of default.

Second, the estimation of parameters in the KMV model. Specifically as follows:

One is the estimation of the company's stock value. This paper uses listed banks as the object of study, and assumes that the asset prices of these 16 listed banks are constant over time.Merton's model, assuming that holding equity in a company is equivalent to holding a call option on the company, then the value of the company's debt is also the final price of the option's exercise.