Research Of Credit Risk Contagion In P2P Lending Platform Based On The Complex Network Topology

The online P2 P lending came into China for ten years or so,experienced the lending platform increased like mushrooms after rain,the question platform failures and the network finance industry regulatory policy was approved,until now the P2 P developed steadily.The study of credit risk of default has always been the focus of the industry and the difficulty,it is also a process of sustained development.The recent research on complex networks has extended gradually from the physics at the earliest to human sociology,biology,finance,and many other industries and fields.In recent years,it becomes a hot research direction in the cross fields.The studied of credit risk contagion in interbank lending by using complex network topology has a large number of relevant literature,In this paper,We trying to establish the P2 P lending network evolution model,on the basic of combing the characteristics of the P2 P lending,consulting the relevant academic research,using the generating function,dynamic evolution theory and method of random network.After then,we do theoretical and empirical studies thoroughly about it.The main research content includes the following aspects:Firstly,We established the network evolution model based on node deletion and increase at the same time,according to characteristics of the formation and evolution in P2 P lending network.In the process of borrowing network growth by time,we found that the degree distribution of nodes in a network is power-law tail characteristics,that is most of the participants have less numbers of borrowing,a few part of participants have more numbers of borrowing.Secondly,we analyzed the P2 P platform's internal lending network topology characteristics.In this paper,we analyzed the related statistical characteristics of the lending network from the node degree distribution,clustering coefficient and network density,according to the relevant measures of complex networks and did the related empirical studies using the data of the Renren Dai platform.The results showed that with the increased of lending network scale,there is no obvious changes of nodes degree distribution characteristics of power-law tail and the network density decreased gradually and eventually evolved into sparse network,the network clustering coefficient of the vast majority nodes is zero.Finally,we checked the lending network structure characteristics to the effect of credit risk contagion.In this paper,we adopted the method of probability generating function analyzing the formation of susceptible nodes cluster and its scale.The theoretical results showed that the size of the susceptible nodes cluster exists a critical value,when the clusters' size are larger than the critical value,there is the possibility of systemic risk outbreak of default,and when the size of the cluster is less than the value,there is only partial default risk contagion.And then we the Renren Dai platform bids data for empirical research,it was found that the inter-network default contagion mainly came from the nodes which have both investors and borrowers roles,theses nodes' proportion decreased with network scale growth and couldn't make a system default risk.At the same time,the scale of default nodes grown along with the network increased,its proportion also increased gradually;The reason was that the average credit quality of the network nodes gradually declined,that is the borrower's own credit score was more and more lower,reduced its ability to repay its debts,the frequency of over time repayment increased and the repayment performance became terribly.