Study Of Credit Rating Method Based On Factor Analysis

The market economy is an economy based on credit; the credit is the basis of market economy and the soul. And the credit risk has been become the most important risks which financial institutions and the World Bank faced. According to a survey conducted by McKinsey & Company, the world's banking industry as an example, the credit risk of banks accounted for 60% of overall risk. Market risk and operational risk will only accounts for 20%, from the 1990s, China have came a lot of credit, didn't become genuine commercial credit environment. On the contrary, the reality is that people are to face credit crisis. This paper researched from China's actual situation of the listed companies in China's credit risk.The broad structure of the article is as follows: first expounded China's economic development facing a crisis of credibility at present. Then introduced the history of the development of enterprises credit rating and credit rating implications, the rating system constitutes relevant concepts; then introduced the model used by the major ratings; Finally researched the application of statistical methods for the empirical, and the empirical results have been analyzed and summarized.Chapter 1 introduced that credit crisis currently faced by domestic enterprises, and the analysis of the credit crisis. In view of the serious lack of credibility, the article established a credit simple and easy rating system to solve this problem.In Chapter 2, it discussed the credit rating; first it reviews credit risk management and the history of the development of credit risk rating. Then the meaning of the credit rating, the rating system elements and the role of credit rating are introduced. And this paper reviewed the existing rating models and indicator system.First it expounded on the credit rating model: 1.the expert analysis system, was often regard as the loan "SC" principle. 2. Multiple discriminate analysis (MDA) selected the variables which provide more information from some characteristics of the variable value of the objects (financial ratios), and established a more discriminate, and makes the discriminant function deduced the miscarriage rate for the smallest of sample classification, then discriminated type of the research objects. 3. Logit model applied the variable range of financial ratios to predict, the probability of default or bankruptcy; according the risk preferences of investors and bank risk police to set lines, it is in order to carry out the analysis of credit risk measurement and management. 4. Neural network model is a process of natural nonlinear; there is no need to distinguish any non-linear relationship, modeling brought to a great convenience. Then introduced the three categories of credit risk measurement which have modern representative: CSFP (Credit Suisse Financial Products) developed the CreditRisk+; J.P. Morgan developed CreditMetrics; KMV model was developed by KMV.It then describes the current corporate credit ratings of the financial index system, first introduced financial indicators selected by Olsen, They are : size (total assets 1980-2002 prices, the number of admission); capital structure (total liabilities / total assets); The rate of return on assets or funds from the business / total liabilities); short-term flows (working capital / total assets, current liabilities / liquid assets), the distinction between the right edge as high as 92% or more. Then the selected financial indicators by Moody's are introduced, these mainly include: analysis of mobile assets, liabilities ratio analysis, financial risk analysis and capital-benefit analysis. Finally, the selected financial indicators of China's commercial banks are introduced, according to the basic principles of indicators: overall choice principle, comparability principle, the principle of a scientific principle and practicality. The index system which commercial bank credit evaluation used include four categories that are profitability, viability and solvency capacity.Chapter 3 is used as evidence for this approach. It discussed the theory-factor analysis, cluster analysis which used in the exposition. This paper used principal component analysis in factor analysis, its main aim is to identify several elements to the public factor which concentrate most of the information, and these factors are the principal components of factor analysis. Then the cluster analysis method is used to determine the appropriate credit rating. In the process of cluster analysis, it can be used to define typed categories of an auxiliary tool which is the distance merged between each of the categories.Chapter 4 is the most important part of the evidence. According to the contents of the previous chapter, it studied the credit rating using financial indicators of China's listed enterprises.First, the sample space and determine contestant sampling method, and on the basis of the rating system and the actual situation of China, the paper elected eight representatives of the financial indicators.Next, using principal component analysis extraction common factor, then using cluster analysis method to determine the appropriate credit rating. In the process of cluster analysis, it can be used to define typed categories of an auxiliary tool which is the distance merged between each of the categories. On based of the above analysis, the paper divided enterprises into five categories in the final, and gave a comprehensive credit score rankings.Finally, the factors preferences of the five categories are analyzed through the plot chart, the results of the analysis reached the same score. This paper shows the rating model is feasible.