Theoretical And Empirical Analysis Of Credit And Risk In Internet Finance

Rise and development of the Internet Finance in China is a fruit of innovative strategies and technological progress. China has been basically synchronized with foreign countries in Internet finance development. Moreover, China has made new innovation and development in specific forms of Internet Finance. Internet Finance showed its first appearance in China in 1997 and experienced a “blowout” and explosive growth in 2013. With its advantages in the organizational models and information processing capacities, Internet finance has gotten rid of existing channels, reduced transaction costs, partly alleviated the problem of asymmetric information, improved the efficiency of capital allocation, further integrated the capital supply and demand, satisfied demands of small retail investors and markets especially medium, small and micro-sized enterprises for financing services, and produced disruptive influence on social and economic life. However, it cannot be ignored that explosive growth and extensive development have been a main feature of Internet finance development in China in recent years. At present, China is still at the initial stage of “capital accumulation, risk aggregation, regulation absence” in its Internet finance development. Credit and risk-related problems existing in traditional financial transactions are still what Internet finance should invest great efforts to solve first.In order to solve problems concerning credit generation and evolution and risk control in Internet Finance, the author makes an empirical study of factors affecting credit in Internet Finance and investigates the evolution of the Internet finance credit embedded with social networks. On this basis, the alien invasion into Internet Finance is analyzed, and Complex Networks and other related theories and real evidence are used to analyze above-mentioned problems. Moreover, effective evaluations are made of credit risks to circumvent or weaken credit risks and impose targeted supervisions on credit risks. The Credit Risk + Model is applied to study credit risks of Internet Finance, which has great theoretical and practical significance to promote the healthy and orderly development of Internet Finance. This thesis is composed of 9 chapters.The first chapter is the introduction part. In this chapter, the author elaborates practical significance and research values of this thesis topic, summarizes research objectives, contents, framework and methods, puts forward several innovation points, and define main concepts of this research.In Chapter 2, literature review and an overview of related theories are made. The research status of Internet Finance at home and abroad is introduced and evaluated. Moreover, on the basis of the theoretical foundation of Internet Finance, an overview is made of the development features, model and trend of Internet Finance.In Chapter 3, the Internet financial information service industry is introduced. On this basis, several Internet financial service modes with different types of platforms are compared. Moreover, work is done to introduce and identify the development status and problems of China’s Internet finance, including the lack of a social credit system, the shortage of talent, technical defects of Internet finance, weak risk control, imperfect laws, incompatibility of regulation with development in the Internet finance, and other problems. The main problem lies in the credit and risk. In this thesis, focus is place on discussing and solving the credit and risk problem.In Chapter 4, a comparative analysis is made of the foreign Prosper platform and domestic P platform which use the borrower’s credit rating and other indicators. Analytical results show which factors have a significantly positive impact on the credit success rate and what kinds of loans enjoy greatest popularity. This can help to further explore the credit and risk problem existing in Internet Finance.In Chapter 5, the social network theory is applied to analyze credit theories and behaviors in Internet Finance. It is noted that in the Internet finance credit, whether actors show preference to credit or fraud is driven by integrated factors based on cultural choices. Therefore, an evolution analysis is made of the credit or fraud preference, and adjustments are made to balance the credit or fraud in the Internet finance credit to get the general and extreme equilibrium solution, abandon undesired ones and analyze convergence and material benefits of equilibrium. The theory of social network embedded is derived to offer paths to the credit problem generated by information asymmetry in the Internet finance development in China.Chapter 6 covers the defects in the analysis made in Chapter 5. Although the evolution analysis in Chapter 5 can well describe interactions between individuals, it cannot well describe group behaviors. In order to better describe the nature of group default risk, the author takes the lead in applying the alien invasion model to control Internet risks. Therefore, the credit default risk is compared to a “virus”, and the single population infectious disease model is transformed into a “multi-population infectious disease model” with various populations invading each other. On this basis, non-linear coupling differential equations about interactions between multiple populations are presented. In this chapter, some simplified hypotheses are used to solve these non-linear coupling differential equations. Moreover, this chapter also discusses how to control risks in a uniform network or scale-free network and measures the threshold value for controlling risks.In Chapter 7, attempts are made to use Credit Risk + model to make an empirical analysis of credit risks in Internet finance, because it would be quite difficult to collect relevant data to make an empirical study of the multi-population alien invasion model generated in this thesis( it would be another topic). According to empirical analytical results, when industry risk factors have equal covariance, compared with the CSFB model and Two-stage Model, the Mixture Gamma Model and the Poly-system Risk Model can better reflect the credit risk of loan portfolios in reality. When industry risk factors have unequal covariance, VaR values of the Mixture Gamma Model and the Poly-system Risk Model are still greater than the CSFB model and Two-stage Model. Compared with the Poly-system Risk Model, the Mixture Gamma Model underestimates the credit risk of loan portfolios. Therefore, the Poly-system Risk model can better evaluate the credit risk of loan portfolios, which verifies the applicability of the model.In Chapter 8, risks and problems in Internet Finance are summarized, and specific recommendations are given to improve credit, reduce credit risks and optimize Internet Finance in China.In Chapter 9, research conclusions and outlook are made. On the basis of concluding what are discussed in each chapter, the author gives some ideas for further research.Compared with previous studies, this study makes innovation in following aspects.Firstly, at present, Internet Finance is a new thing in China and still at its initial stage of development. In existing research on Internet finance, scholars mostly make comments, but rarely make empirical research and dynamic analyses. In this thesis, a comparative analysis is made of the foreign Prosper Platform and domestic P Platform as representative Internet finance credit platforms in China and foreign countries, to obtain valuable conclusions.Secondly, in this thesis, the social network theory is applied to analyze theories and behaviors related to the core issue(credit) in the Internet finance credit. Adjustments are made to balance the trust or fraud in the Internet finance credit to get the general and extreme equilibrium solution, abandon undesired ones and analyze convergence of equilibrium.Thirdly, on the basis of the equilibrium adjustment, the author takes the lead in applying the alien invasion model to control Internet risks. To better describe the nature of group default risks, the credit default risk is compared to a “virus”, and the single population infectious disease model is transformed into a “multi-population infectious disease model” with various populations invading each other. On this basis, non-linear coupling differential equations about interactions between multiple populations are presented. Some simplified hypotheses are used to solve these non-linear coupling differential equations.Fourthly,In the case of a uniform network, the possibility of a wide-range default risk proliferation is quite small. In the case of a scale-free network, the default risk threshold value would be significantly(negatively) affected by the value of network scale. The value is calculated and a simulation graphical analysis is made. This is conducive to control of financial risks. A larger value of the network scale indicates a closer link between different individual members in this scale-free network and a higher complexity degree of the network, showing typical features of a “small world”. Such a scale-free network is likely to experience a substantial increase in default risks. In fact, the outbreak of the subprime crisis in 2008 has a great impact on the complexity of the relationship network structure in the human world.attempts are made to use Credit Risk + model to make an empirical analysis of credit risks in Internet finance, because it would be quite difficult to collect relevant data to make an empirical study of the multi-population alien invasion model generated in this thesis( it would be another topic).These conclusions are of great significance for further analyzing the credit and risk problem in Internet finance. On this basis, the Credit Risk + Model is used to analyze credit risks in Internet finance. This model was used mainly for estimating credit risks of commercial banks. This model has been constantly evolved and improved through research. Several models derived from the Credit Risk+Model, including the CSFB Credit Risk Model, the Mixture Credit Risk+ Model, Two-stage Credit Risk + Model and Poly-system Risk Credit Risk + Model, are compared to figure out their advantages. Loan data of four industries on the Internet loan platforms is used to make a comparative analysis of Internet finance credit risks in different models, at different confidence levels.