The Study On Early-warning Of Financial Crisis For Listed Company Based On BP Neural Networks

With the deepening of economic reform in China,more and more companies will have to be faced with fierce competition .If a firm operate unsuccessfully,itwill be hard to avoid going bankruptcy . As to listed company, if falling into thefinancial crisis, it will bring enormous loss to investors, creditors. So to establish a effective practical pre-warning model on financial crisis, in order to meet interests' increasingly urgent need,has already not merely been an academic problem, but becomes the important factor influencingsound development of capital market of our country, and has very important realistic meanings.On the basis of forefathers' research results, this thesis has carried on more systematicresearch and discussion on the theory and model of pre-warning on financial crisis for listed company of our country. Firstly, the thesis expounds the pre-warning basic theory of the financialcrisis, defines the concept of financial crisis, and designs 25 pre-warning index. Secondly, based on the analysis and evaluation of thetraditional methods in pre-warning on financial crisis, the article puts forwards the combination forecast based on BP neural network, and structures the pre-warning model on financial crisis.Finally, regarded making company in Shenzhen and Shanghai stock market of our country as the research object, selecting 116 listed companies as the samples, the real example analysis have been carried on. The result of study indicates that the combination forecast model integrates the advantage of statistical model and artificial intelligence model, having 83.34 percent accuracy rate of discrimination in two years before listed company taking place the financial crisis, and takes on stronger superiority and application value.