Research On The Rating Of Micro-enterprise On Equity-Crowdfunding Platform In China

On the one hand,equity-crowdfunding is a product of the development of internet finance.The new online equity financing mode firstly appeared in the USA.Until 2014,the first batch of equity-crowdfunding platforms appeared in China,which has raised much attention of the government.On the other hand,since the reform and open,micro enterprises have occupied an important position in the national economy,which is the basis of the social development and national economy.However,micro enterprises always have difficulties in financing,which has been a stumbling block to the development of micro enterprises.The innovative mode,equity-crowdfunding,provide new financing channels for micro enterprises.A set of scientific small and micro enterprise rating system can help companies and stakeholders to reduce risk and make the right decision.The purpose of this thesis is to establish a model of micro enterprises on the equity-crowdfunding platform in China to help make decision.By reading the relevant literature at home and abroad,according to the five basic principles of the establishment of enterprise rating indicators,five first class indicators which can fully reflect the status of enterprises have been selected.Then,combined with the general characteristics of the usual enterprise and the unique characteristics of micro enterprises on the equity-crowdfunding platform,5 first class indicators and 18 second class indicators have been selected to build a set of rating index system of the micro enterprise.A sample of 200 start-ups has been selected and 18 second class indicators have been divided into three categories.Then,the SVM model of micro enterprise rating is constructed,the parameters of which is optimized by the genetic algorithm.And the Logistic model is used to carry out the comparative research.The research shows that the accuracy of the SVM model based on the genetic algorithm for micro-enterprises rating is 96.67%and the Logistic regression model is 85%,far lower than the performance of SVM.It means that the constructed SVM model can accurately predict the financial situation of enterprise and it has a certain reference value for investors.