Study On Business Distress Prediction Of Chinese P2P Lending Platforms Based On Cox Proportional Hazards Model

The P2P lending industry is in the ascendant in our country where lots of new online P2P platforms start business and even more old ones are involved in business distress.When P2P lending platforms get into such trouble,it will probably brings investors much loss and also makes some problems for the P2P industry regulatory authority.We know that a company or platform falling into business distress is a continuous worsening process.If we can find symptoms at the beginning of business distress and predict the dynamic process of that by a forewarning model,that will means a lot to operators,investors and regulators.On the base of domestic and oversea document literatures,this paper tried to construct a forewarning model for Chinese online P2P lending platforms and showed its developing process dynamically.First,this paper analyzed six kinds of risks and summarized four categories of business distress.Second,after looking up the document literatures,this paper found 10 forewarning factors,including district,registered capital,shareholders background,guarantee institution,trustee agency,average annual yield,automatic bidding mechanism,management experience,governance structure and alternative financing channels.During the empirical analysis,first,we analyzed the lifetime of platforms through the life tables method and summarized the characteristics.Then with the Kruskal-Wallis rank sum test and Kaplan-Meier test,we found 8 factors that can be used for Cox proportional risk model.Finally,we built a forewarning business distress model based on a cox proportional risk model that can provide outstanding forecast accuracy,in which we found trustee agency,average annual yield and management experience are protective factors.This article firstly applies the survival analysis model in the forewarning research of P2P business distress which has been widely used in the medical statistics and achieves nice outcomes.This forewarning model will help a lot in controlling the business risks.