Commercial Banks To The Sme Credit Risk Management

Risk management is the core content and everlasting task of the administration of modern commercial banks. This dissertation aims at analyzing and studying the risk of bank credit granted to Small and Medium Enterprises (SMEs) that are closely related to the credit transactions of China's commercial banks. Some methods currently used in credit risk control by Chinese commercial banks are reviewed. The economic variables affecting the credit risk of commercial banks are demonstrated and analyzed by using quantitative economic approaches combined with collecting historical data of credit granted to SMEs. Taking into account the historical characteristics of credit risk of SMEs, practical credit-awarding control theories and methods adaptable to Chinese national conditions are explored.This dissertation is structured as follows:The first chapter is an introduction which begins with the historical lessons on the risk loss of SMEs granted by our commercial bank and provides the background and basis for further study. The research objectives and methods are then educed. Chapter 2, starting with the historical lessons on the financial risk loss at home and abroad, elucidates that modern risk management is a global task. Meanwhile the similarities and differences of financial risk losses between developing countries like China and western developed countries are discussed. The development process of the theories and methods of risk management and the practices of some major economic entities are then discussed. Finally the Basel Capital Protocol and its evolving procedures that are closely related to the risk management of commercial banks are briefly introduced. Based on a summarization of the idea of credit risk control and its evolvement of modern western commercial banks, Chapter 3 mainly discusses the theories and methods of credit risk control used by modern western commercial banks, including the theory and practice of credit risk control techniques in some major western capitalist countries such as in USA, for examples, the credit matrix that is familiar to us, KMV, Value-at-risk model, the American prevailing economic capitals allocating system, credit risk measuring models including structural models and set models, etc..Chapter 4 analyzes in detail the status quo of credit and financing of our SMEs, points out that the direct cause for the difficulty of our SMEs in financing is the low-rank of credit of our SMEs. Section 1 construes the credit characteristics of our SMEs according to the definition and classification of our SMEs. Section 2 expatiates on the financing circumstances and approaches of our SMEs, probing into the concrete reasons of the difficulties in financing. Section 3, by taking into account the status quo of the credit of our SMEs and comparing the commercial credit systems at home and abroad, puts forth some methods of keeping away and dissolving the credit risks of our SMEs. Chapter 5 starts with the practices of the credit-awarding operation of our commercial banks, and then expatiates on the practical methods of credit-awarding management of SMEs granted by commercial banks. Section 1 briefly introduces the concept and classification of credit-awarding operation, and the interrelations among different credit-awarding steps. Section 2 discusses the risk characteristics of SMEs and the risk-gaming relationship between commercial banks and SMEs. Section 3 presents the risk appraisal indices of SMEs granted by commercial banks. Both interior indices and exterior indices are discussed. Finally the traditional methods of credit-awarding management of SMEs granted by our commercial banks are expatiated.Chapter 6 mainly takes the interior credit grading system of commercial banks as research objects, applies econometric model to quantitative analysis and test of accuracy and rationality of the credit grade of commercial banks. Section 1 chooses the credit grading system of the Bank of China Limited as the testing object to carry through an essential introduction about the customer credit grading systems of the Bank of China Limited. For the sake of better understanding by readers about selected testing objects, the introduction covers a broad range of topics from the standard, object, realization method, examination and approval procedures etc.. Section 2 deals with the econometric testing method used in this chapter– ordered multivariable discrete choice model. Section 3 begins with the choice of testing variables, selects those variables that have larger effects on the credit grading system as independent variables of the model from many variables affecting the administration activity of the corporations. The selection is done by ordered models. Finally the ultimate form of ordered multivariable discrete choice models is determined and the affecting force coefficients of each explanatory variable to the customer credit grading system are obtained. In Section 4 the expectation prediction is conducted by using the established ordered model. The results of prediction show that it is feasible to use the ordered model to appraise the credit grade of various enterprises.Chapter 7 takes the risk quotas of SMEs'credit-awarding granted by commercial banks as samples, utilizes the weighted least squares to perform econometric tests on the financial and administrative indicators affecting the risk quota of SMEs' credit-awarding. Section 1 briefly introduces the concept and objective of the risk quota of corporations'credit-awarding granted by commercial banks, and the significance of a unified quota management. Section 2 begins with a discussion on the significance in testing the corporate risk quota, followed by an introduction of the design of the credit-awarding quota testing models and selection criteria of model variables. This section then expatiates on the establishment, estimation of models, and the analysis of results. Having completed a collective evaluation of the model, an important conclusion is drawn that there exists an over-credit-awarding to major clients in our commercial banks at present. Most importantly, some administrative indicators that are mostly concerned by commercial banks in credit-awarding are got hold, which provides a foundation for further studies on the risk preference of commercial banks in credit-awarding. These indicators include corporation credit grade, owner's equity, working capitals, net value of fixed asset, and working liability etc.Combined with the econometric estimation results of Chapter 7, Chapter 8 establishes the Artificial Neural Network (ANN) Model, re-examines the credit- awarding risk quota from another aspect. Meanwhile a comparison is made between the methods and conclusions of two kinds of models. The advantages and disadvantages of the two models are deduced. Section 1 is an introduction of the ANN model. In Section 2, the econometric model of credit-awarding risk quota is reviewed first, then a five-variable ANN model is established followed by a comparison between econometric model and ANN model. The third model, a ten-variable ANN model is established subsequently and the three models are compared. It is found that the fitness of five-variable ANN model is basically the same as that of the econometric model, but the fitness of the ten-variable ANN model is obviously better than that of econometric model. The similarities and difference between the two types of modeling methods are stated as follows: the ANN models possess the characteristics that the input and output have nonlinear relationships, thereby can accommodate and utilize more information, which is superior to the linear econometric models. The linear econometric models can only accommodate information possessing linear relationships and exclude information possessing nonlinear relationships. However the linear econometric models can determine the coefficients of interactive force between economic variables, while the ANN models could not. All these show that each kind of model has its own strong point, and should not be substituted each other. It is suggested that selection between the two types of models should be done according to the need of the particular business.Chapter 9 is the conclusion of the dissertation. First, a recapitulative summary on this research work is provided and shortcomings are discussed. Then several aspects that are closely related to the credit-awarding transactions of the commercial banks and should be paid attention to, for example, obey the rules and regulations, to prevent from fraud and to secure assets etc, are put forth. These problems could not be solved just by mastering risk control theories and methods; they also play an important role in the credit-awarding risk control. In our present situations that financial markets and legal supervisions are not perfect and not transparent enough, it seems very important to focus on these problems. Finally, some forward-looking generalizations are provided as future research directions.The main innovations of this dissertation are: 1. The research aims at the customer credit grading system used by Bank of China Limited, selects actual data of a local branch. Tests are performed on the grade system by using ordered multivariate discrete choice econometric model to find: whether the index setting of the credit grading system is reasonable; whether there exists redundant index; whether the score setting of each index is reasonable; and whether the degree of effect on the grade results is consistent with original thoughts of design. All these will provide consulting ideas for the revision and perfection of the customer credit grading system. 2. It utilizes the credit-awarding risk quota granting to the manufacturing corporations by Bank of China Limited and financial data of the credit-awarding corporations, together with the credit grading situations of the corporations, to establish the weighted least square models. Results of the model appraisal are obtained by using the software E-Views. By analyzing the various factors affecting the credit-awarding risk quota and the degree of effect, consulting suggestions on constituting right credit-awarding risk quota and modifying present risk quota are put forward. 3. Through re-testing the risk quota by ANN methods, comparisons of similarities and differences are made between the econometric model and ANN models in testing the credit-awarding risk quota. Advantages and disadvantages of these two models are analyzed respectively so that a more satisfactory result can be obtained. 4. The research objective of this dissertation is to provide guidance on the theories and methods of credit risk control for the credit managers of commercial banks. All this is done based on case analysis with actual data, and it is an applied research. This dissertation adopts quantitative models that are relatively advanced in the world, combines qualitative with quantitative research approaches. It possesses practical guidance and is valuable in practical application. This is significant in theoretical exploration and in innovative practice operations for the present risk control and risk management of commercial banks.Being a study for a doctorate degree in Quantitative Economics specialty, this dissertation emphasizes particularly on testing the current risk control theory used by commercial banks in practice. At the same time, useful attempts are made to apply the quantitative economic analytic methods in credit risk measurement models for commercial banks. However, regarding the systematic and comprehensive aspects of the research, more in-depth expatiations on the risk control theories in commercial banks are still needed. Welcome comments from teachers and scholars who intend to undertake research in this area.