Considering The Non-Financial Factor Of Multi Classification Of Financial Crisis Early Warning

Nowadays, the social and economic environment is becoming increasingly complex, enterprise competition is increasingly fierce, and how to maintain sustained and healthy operation of listing Corporation has become the primary problem faced in the fierce competitive environment. The most intuitive crisis is the financial crisis as the enterprise faces, the research and prevention of it has become the focus of the company. This paper takes A stock manufacturing industry—the machinery, equipment, instrument listed companies as the subject investigated, and takes all samples in the same industry to research on financial crisis early warning abandoning the traditional research ideas of Crisis-Normal paired samples. First of all, in view of the one-sidedness of the early-warming indexes in the field of early warning crisis, consider our corporate governance characteristics, will take the non-financial indicators of owner-ship structure, board features, executive incentive, etc. Reaction the internal information characteristics of the company in the early warning model. Then, building the corporate governance index (G index) based on main-composition analysis, and unifying the related financial indicators to building an early warning indexes system which contains financial and non-financial information. Secondly, the K-Means clustering is introduced in the research field of crisis early warning to break the two classification of financial situation, using the ratio of intra-cluster distance and class distance to decide the number of clusters, and then, combining K-Means clustering, the listed companies’ financial situation is divided into four grades eventually. Finally, predicating the financial crisis of enterprise by the early warning model based on logistic regression. The empirical research shows that:most of the China’s A stock manufacturing industry—machinery, equipment, instrument listed companies have a bad financial situation. And, the non-financial indicators include the company interior governance factors have a significant influence on the warning model, the G index can significantly improve the accuracy rate of the model.