Research On Warning Of Financial Crisis Of Listed Companies Base On Genetic Algorithm

When economy develops rapidly in an unpredictable market,each enterprise cannot escape from the influence of macro environment and they need to continuously embrace opportunities and deal with challenges to adjust to the progress of times.When an enterprise is trapped in financial risks,managers need to make adjustment.Each enterprise is not separated and its decisions tend to affect associated enterprises,so more and more people pay attention to the detection and timely prevention of financial risks.The research of early warning system of financial crisis begins to gain more attention.The more accurate the warning system is,the more helpful it will be for enterprise managers to make appropriate decisions.With the advent of Internet era,it is not difficult to retrieve financial information.It is also of great significance to extract accurate information out of loads of trivial messages when facing information explosion.When analyzing the research outcomes of financial crisis both at home and abroad,pros and cons of different early warning systems have been studied.The paper first makes a clear definition of financial crisis and then lists the reasons and results of these crisis.The genetic algorithm is applied to research in the field of early warning of financial crisis.Through combining the financial indicators with the nonfinancial indicators,a financial indicator of early warning is set up to test whether the selected warning indicators can meet the requirements.Eventually adoptable early warning indicators are retained to be data of empirical research.Samples selected in this paper are all listed companies from Shanghai and Shenzhen Stock Exchange,and 68 companies with normal or abnormal financial conditions are selected for empirical research.The original early warning model of financial crisis and the genetic algorithm-based early warning model of financial crisis are established respectively to run the stimulation test with the retained data.After comparing and analyzing the results of two models,the conclusion is reached.