The Estimation Of Credit Risk Based On The Fuzzy Integrative Evaluation Model

As a bran-new supervisory capital management frame, one of primary innovation in Basel New Capital Agreement is to raise the counting inner level method as the credit risk supervisory capital required. The method allows banks to use measure its credit risk with own inner data which also develop more business for the banks with high level of risk management than the other banks with same capital. Based on the partition of risk characteristic, the main point for inner level method is to confirm some risk factors on every risk assets such as the probability of breaking promise, loss rate, risk exposure and valid time and so on. According to measuring these risk factors, we measure the requirements of credit risk capital for all risk assets in banks.Because the banking is a kind of field with high venture, so it is key for banks to have competition with the others to evaluate and manage risks. The credit risk is always one of main venture for commercial bank to shoulder. It exists in close relationship between management and its measurement on credit risk in commercial bank.The credit risk points to the indetermination of safety factor for finance fund and the different reasons as follows: bank loan cannot be got back which increases the possibility of bad debt because of unwillingness or incapability for them to refund the primary interest. Moreover, the credit risk also includes the possibility of changeable debt market value and loss owing to the borrower's different credit standing and ability of insecure stipulation. Many international famous banks all begin to develop their inner risk control and managed model. More and more the analytical technique of capacity variety has become the principal part of study. At the present time, the commercial bank of China faces both the huge potential problem of credit risk and short of strong ability to shoulder risk. Hence, we should enhance technically the risk control ability for the commercial bank as urgent affairs. At the same time nowadays we strengthen the measurement of credit risk and evaluate the actual state in a corporation and forecast future by reason of immature Chinese capital market and new beginning on accumulating prevenient data. Simultaneously, we must reinforce the measurement for non-finance factors working on credit risk according to the lag financial data and low reliability in internal corporation. We should adopt the traditional model such as way of"6C"and credit evaluation with own characteristic in commercial bank to measure credit risk. We also should strengthen 5-class loan management, boost the level of regulations and standardization on credit risk control and reduce ill loan rate. Therewith we should create actively the condition to use gradually measurement of credit risk with advanced inner model.It's correlative between credit risk and man's subjective action which is also a result of thinking function for man brain. During thinking, brain will make a decision inevitably based on unclear thoughts and value standpoint. Therefore we may solve the unclear problem about credit risk with a way of math inkling.This paper adopts multi-level ways with blurry integrated estimate to construct Chinese commercial bank's credit risk evaluating model which combines qualitative analysis and quantitative analysis, and carry through the demonstration analysis to expect more choices offered by credit risk measurement.The process and main point is as follows: according to the"6 C"way, we build up target system, which involve lots of factors from different aspects about credit risk evaluation in commercial bank. However it's feasible for estimate multi-target credit risk with multi-level blurry integrated estimate model. It reflects different level among factors of objective things which can avoid the disadvantage for the main factor subset to from distributing easily if more factors exist.This paper adopts the way"AHP"to make sure the coefficient with important degree of every target factor. Level analysis way (AHP) is a kind of simple and convenient way which always adopted while making quantitative analysis on non-quantitative affairs in systems engineering. The theory for this way to make sure the principle of important degree is to compare every two factors at different level by means of level structure model and to build up the judge matrix which will lead to the answer of important degree. Due to the strict logic of blurry math method, we can have filter and repair disposal on assured important degree. Thus we try to eliminate subjective component to make the assured important degree to meet external facts much more.There are 5-class sorts for the evaluation about credit risk as follows: normal sort, attention sort, of a sort, questionable sort and losing sort. According to the way of combining qualitative index expert score with subject function which is relevant to qualitative index structure, we build the general judge matrix from target factor volume to estimate volume. While choosing the blurry judge model, we operate the blurry matrix with the considered important average model M (·,+) to get the general evaluation result of credit risk and judge its type according to the rule of supreme subject degree. The type judges generally and have a well-considered all factors with the size of important degree to avoid neglect from some meaning factors.