Font Size: a A A

Research Models And Methods Of Chinese Commercial Banks' Risk Monitoring And Early Warning

Posted on:2006-01-12Degree:MasterType:Thesis
Country:ChinaCandidate:Z Z ZhangFull Text:PDF
GTID:2206360152475734Subject:Operational Research and Cybernetics
Abstract/Summary:PDF Full Text Request
Based on the theory of artificial neural network, data envelope analysis, fuzzy mathematics and analytical hierarchy process, this dissertation is devoted to some models and methods about risk management of commercial bank with a view to the current research background of risk management in China. The main work of this paper can be summarized as follows:1. Finance is the core of modern economy, and the efficiency and safety of commercial banks determine in a certain extent the efficiency and safety of the financial system of the whole country. The monitoring and early warning system of commercial banks can be helpful to preventing bank risks in advanced instead of settling risks after the event. In this paper, the models and methods of monitoring and early warning of bank risks are summarized briefly, at the same time the latest progress and development trend in this area are discussed.2. According to current conditions of Chinese enterprises, a credit risk model suitable for the account data of Chinese enterprises is established. The model is based on artificial neural network (ANN) technique, which has many strong points such as good learning ability, error tolerance and robustness. The ANN model can extract the common characteristic of financial ratios and overcome the mutual disturbance of interrelationships of financial ratios. An empirical study is done with the data offered by a state-owned bank via the above model and the results show that the ANN model has a high predictive accuracy.3. Based on rejected cases a data envelope analysis (DEA) model is presented and a classification method is proposed with piecewise linear separating hyperplane as its boundary. The question of how to appraise the credit of new units with only the credit information of bankrupt units is discussed. A specific method evaluating credit situations of decision making units (DMU) is introduced, which classify the DMU into two categories i.e. the acceptance or rejection of credit risk. The proposed DEA programming model can determine not only whether a new case is accepted, but also the location of the new case withrespect to the boundary determined by the samples previously classified. Empirical results show that the model and method are feasible. 4. According to current conditions of Chinese commercial banks, an index system of financial risk of Chinese commercial banks is designed. Then the member functions of the all indices are constructed. Fuzzy analytical hierarchy process (AHP) is used for the appraising of bank risks. That is to say, the importance weight of the indices is objectively determined by fuzzy AHP, which take the uncertainty, ambiguity of human judgment and assessment attitude of policymaker into considerations. The risk types of commercial banks are determined using the method of fuzzy synthetic evaluation, which make up the limitations of judging bank risk from single index and is an exploration of risk management from the degree of Total risk management (TRM). A case analysis is given and the empirical results show the model is feasible.
Keywords/Search Tags:Commercial bank, Risk management, Credit risk, Artificial neural network, Data envelope analysis, Fuzzy analytical hierarchy process.
PDF Full Text Request
Related items