Application Of 4 Level Discriminant Analysis Model In Credit Rating Of Private Enterprises

With deepening development of socialist market economy in our country, Private economy and private enterprises have accessed to a great development and have made certain achievements .Private economy is an indispensable microeconomic foundation and the product of the economic system reform, which conversely promotes the establishment and perfection of our economy system. The rise of private enterprises much stimulates the development and expansion of native market, and fastens building the framework of our market economy system, as well as promotes the transformation of government function and the establishment of national macro control system, which make a large contribution to native economy ,social development and political and economic reform.However, financial difficulty still is the bottleneck of survival and development of private enterprises. It is significant for promoting our economy and society progress to solve those difficulties. The reason of financial difficulty of native private enterprises is complex, and the most notable obstacle is information asymmetric. To our native commercial bank, private enterprises are really new client who distribute dispersedly, and there are no authoritative agencies rating for them. So ,it seems much important to establish a complete creditmetrics index system of private enterprises by scientific method.The whole procedure of estimating the object's funds and reputation, testing and measuring their credit both from quality and quantity objectively is credit rating. Popular credit rating methods include classical credit analysis, discriminant analysis, Credit-Metrics model and KMV model. Many native scholars have also researched measure methods of estimating credit risk. Some scholars like Wang Chunfeng handled DA,Logit,Probit and NN to analyzed the credit risk of commercial banks, and compared advantages and disadvantages among these ways. Another group of scholars such as Shi Xiquan estimated the credit risk of listed companies and built a 2-level discriminant analysis model by DA.This paper simulated the way by which Altman built ZETA model, estimated many discriminant functions, established a 4-level DA model to measure the credit risk of private enterprises. We chose 16 indexes which represent profit ability, capital management, debt ability, cash inflow and growth ability, then removed indexes related and insignificanty by step-discriminant-analysis.5 indexes have been reserved at last.After that, linear discriminant functions have been set up with these 5 indexes. We tried establishing Fisher functions about 3 level,4 level,5 level and 6 level, and results showed the optimal MDA was 4 level. According that, we set up a 4 level MDA model and estimated the discriminant functions and critical values, whose conclusion was confirmed realistic and suitable by test. In addition, two derivable result were drawn: the first one is that native private enterprises could be measured 4 levels off; and the other is that abilities of profit and gaining cash inflow have more effect than debt ability to credit risk of our native private enterprises.