A Study On Crisis Early-warning Theories And Positive Research For Listed Companies Of Our Country

The prediction of crisis is a big problem that concerns the vital interests of companies' investors and managers. With a constant improvement of security market's environment in our country, investor s ' thoughts of investment becoming gradually ripe and the managers' psychology becoming rational, there will be more extensive demands for the early warning of business crisis. Just in this background, this paper selects this hard-to-get problem for the thesis. And it has launched analysis and expounds in detail about the crisis early warning for listed companies of our country.It provides a specifical review of the literatures and development framework for this problem firstly. On base of it , to avoid some limitations and insufficient that exist in the present studies, it will take analysis and make discussions in two parts, one is in theory field and the other is in empirical study. In theory study, it examines closely and makes again clearly on the concept and characteristic of the business crisis. Along with the empirical analyses to origins, it describes and probes into the forming mechanism, growth and spreading of business crisis deeply. Finally, it puts forward a design philosophy to the crisis early warning system of enterprises. In positive research, using selected early warning index system and datum from Machinery, equipment, instrument type listed companies, it chooses and uses Factor Analysis approach, Cluster Analysis approach and Logistic Regression approach, three methods together for the first time , to get a prediction model. At the end, it successfully presents a early warning model with higher percentage that it correctly predicts.