Research On Financial Early-warning Model Of China's Manufacturing Listed Companies Based On BP Neural Network

In the fierce market competition,only development can obtain the living space.Manufacturing industry is the pillar industry of our national economy,and the proportion of manufacturing listed companies in China is over 65%,which has a significant impact on the national economy.But in recent years,China's manufacturing listed companies because of financial difficulties by special processing,and manufacturing the existence of enterprise financial risk serious impact on the healthy,stable and sustainable development of the enterprise.The importance of the manufacturing industry,combining with the characteristics of manufacturing enterprises,and the financial risk factors,through the financial data and non-financial indicators,accurately is a financial risk early warning of manufacturing listed companies,this research to strengthen the financial risk early warning of industry,therefore,research on manufacturing listed companies how to prevent the financial crisis has a certain theoretical and practical significance.This paper will set the research object as deep Shanghai a-share listed companies of manufacturing industry in our country,and defines the enterprise financial crisis for the enterprise for abnormal financial position and the special treatment(ST and*ST),and select the deep Shanghai,two cities in China in 2014-2016 for the first time by ST 60 group financial crisis as the financial crisis,at the same time,adopt the method of matched one by one year after year to choose 60 normal financial health as A financial group.From the debt paying ability,growth ability,profit ability,operation ability,the ability of cash flow and earnings per share index six aspects of financial and other non-financial factors,according to the selecting principle of early warning indicators,36 were selected 11 financial indicators and non-financial indicators,to set is suitable for the financial early-warning index system of our country's manufacturing industry.And through the kolmogorov-m's normal distribution test(K-S inspection),the significance test(independent sample t-test and the mann-Whitney-wilkerson inspection),the significance of early warning indexes,and the significance of early warning indicators extracting principal components to reduce the number of variables and eliminate multicollinearity;Selection was first ST index of the first three years of data as the cross section data to establish the BP neural network model,model prediction accuracy can reach 90.83%.The early-warning model of early-warning model established in this paper is ideal,so it is possible to find out the possibility of financial crisis in the enterprise in advance.