Research On Credit Risk Early-warning Of Technology Bank Based On Bp Neural Network

With the evolution of economic globalization,technological innovation has become a "new engine" to promote the development of a country's economic,and China is in the critical stage of promoting innovative country construction and realizing economic restructuring and upgrading.Under the financial environment dominated by Banks,how can Banks not only meet the financing needs of enterprises in different life cycles and control credit risks rationally,which is the most important part of improving the innovation system of Technology and Finance,promoting industrial upgrading and economic growth in China.The technology bank has been developed to solve the financing difficulties of science and technology enterprises,improve capital allocation efficiency,and build a reasonable Technology and Finance System.However,China's theoretical research and practical exploration of technology bank are still in its infancy stage,the development of technology bank is seriously restricted by multiple factors,as for the high costs and risks nature of the technology loans,the information asymmetry between bank and enterprises,imperfect scientific and technological credit evaluation system,etc.Therefore,it is of great practical significance to study the credit risk early-warning problem of technology bank in this context.On the basis of combing the domestic and foreign related research,this paper starts from the concepts and characteristics of technology bank and sci-tech innovation enterprise,draw lessons from the credit risk early-warning theory and method of neural network model,building a technology bank credit risk early-warning model and carry on the empirical test.The paper puts forward related countermeasures and suggestions of controlling credit risk of technology bank.The specific research results and conclusions are as follows:(1)Analysis of the operation mode and credit risk factors of the technology bank.Based on the analysis of different operation modes of silicon valley,suzhou and Shanghai technology bank,this paper points out that the development goal of the future technology bank is to build a more advantageous technology corporate bank.Meanwhile,through the analysis of the credit participation subject of technology bank,the paper points out the category and cause of credit risk of technology bank.(2)Screening and construction of credit risk early-warning indicator system.This paper constructs the credit risk early-warning indicator system of technology bank from three perspectives for enterprise innovation ability,bank internal controlling and external environment.The paper uses combination weighting method to determine the final weight of indicators.(3)Establishing of credit risk early-warning model of Sci-tech bank.Applying BP neural network,this paper set up a credit risk early-warning model of technology bank.The parameters of BP neural network such as node number in each layer,activation function,learning rate are determined by the characteristics of the indicator system.Taking the SPD Silicon Valley Bank as empirical sample,the neural network model was trained and tested.Finally,we optimize the neural network through the genetic algorithm,the model error is small and gain a satisfactory pre-warning results.