Research On Financial Early Warning Of P2P Lending Platform Based On CHAID Model

In the context of national policy encouragement and market demand,P2 P network lending platform began to flourish that its development plays a significant role to small and medium-sized enterprise financing,the expansion of investors' investment channels and the continuous growth of private lendingmarket.But if an industry only focuses on growth for a long time and neglects the relevant regulatory requirements,its path will deviate from expectations.A large number of informal platforms,such as campus lending and qian bao,have exploded,triggering investor panic and creating a bad reputation.Under the background of complex industry,we cannot only pay attention to the credit risk of single borrower,the quality of lending business depends on the network lending platform,the collapse of the platform is bound to cause a wide range of impact.So this paper takes the overall risk of P2 P network lending platform as the research direction.CHAID model,is one of the decision tree data mining technology,through the analysis of sample data differences,obtain some key indicators that affect financial status of the platform on the basis of index difference degree of sorting,generated a tree hierarchy,concluded that anticipation platform financial conditions of the time.Previous studies have shown that CHAID model is applied to the financial warning system of listed companies,and the early warning effect is improved.CHAID model to conduct financial warning research on P2 P lending platforms,aiming at promoting the sustainable and healthy development of P2 P lending industry.First,this article clarifies the present situation of the domestic P2 P network platform loan,regulatory developments,the policy environment of that platform has entered the stage of standard bat from the open inclusive,which led to the nature of the P2 P network credit platform,and the specific operating mode.Second,the role of macro factors in the industry is common.Secondly,considering that the macro factors are common to the industry,based on the CHAID model,the systemic factors can not be classified.Therefore,the financial characteristic factors of the platform are taken as the focus of the model.The risk of P2 P network loan is classified into financial risk and non-financial risk.The financial risk is further divided into three aspects: liquidity risk,credit risk and operational risk.In the following chapters,the risk index system is established.In summary,the definition of the problem platform is put forward,The problem platform is manifested in the following five aspects: the difficulty of makingmoney,the intervention of economic investigation,the running of roads,the closure of business,and the transformation.The financial status of the platform is determined as the target variable of the model output.The definition of financial early warning and the idea of CHAID model are expounded.In the empirical part,we begin to build model.The data of this paper mainly come from the Internet financial data collected by Cathay Pacific Database from 2007-2017.The manual supplementary data comes from the domestic authoritative third party net loan industry synthesis portal "net loan house","net loan heaven eye",after culling duplicate,missing data,finally selected 729 normal platforms and 157 problem platform data.Using SPSS20 software,the explanatory variables are analyzed descriptive,independent sample T test is used to show the significance level of index difference,after the training set is extracted,the test set is tested,then the CHAID decision tree operation is started.The tree-like result chart shows that liquidity risk is the most important risk in the operation of the platform,and the ratio of the number of investors to the number of borrowers is the key index of financial early warning.Secondly,we should also pay attention to the accumulative repayment amount and the index of investment number.The accuracy rate of financial early warning prediction of training sample and test sample is more than 90%,which has good early warning effect.The tree structure of CHAID model is clear and easy to understand,which has the value of popularization and application.For investors,the early warning model can be used before the investment decision,and for platform operators,it is possible to examine whether the key indicators have been met,for the government supervision department,it is possible to issue targeted policy recommendations according to the research results.The main deficiency of this paper is that the data resources are difficult to obtain,the final result is the traditional static financial early warning system,the effectiveness of early warning lags behind,With the continuous enrichment of information resources,the dynamic financial early warning system of net loan platform will become the trend of later development.