Research On Financial Risk Early Warning Of Listed Companies In Coal Industry Based On PCA-SVM Model

For a long time,coal resources have been dominant in China's energy consumption structure.Although my country has abundant reserves of traditional conventional energy sources(such as coal,oil and natural gas),my country has a large population base,a small per capita occupation,and low utilization efficiency.,The demand for energy is far greater than the supply,and the status quo of perennial dependence on imports is particularly prominent.With the implementation of structural reforms on the energy supply side,the adjustment of my country's energy consumption structure is accelerating.During this process,the competition among listed companies in the coal industry will become increasingly fierce,which will give rise to a series of financial problems.How the listed companies in the coal industry seek to develop and find a way out in the course of reforms is very important.In this context,how to scientifically carry out financial risk early warning for listed companies in the coal industry is of great significance.Financial risk is an unexpected event that cannot be predicted in advance in the production and operation of a company.Once a company has a financial risk,it will affect daily operations at a small scale,and threaten the survival of the company or even lead to bankruptcy.The business activities of listed companies in the coal industry run through the existence of the company and will be affected by financing activities,investment activities,business activities,inventory management,and liquidity.Based on the impact of these activities,this paper selects a total of 35 indicators to construct a financial early warning indicator system for listed companies in the coal industry,and selects 36 listed companies in the coal industry as the research objects.The time range of the sample data is 2000-2018.Based on the research status at home and abroad,this paper constructs a financial risk early warning model based on principal component analysis and support vector machine combination model,and conducts comparative analysis through principal component analysis,breakpoint regression test,single SVM model test and PCA-SVM model.The following conclusions are drawn: First,by identifying the key factors of all the indicators selected in this article,it is found that the 12 key factors extracted in this article can explain the six dimensions of the indicator system,and illustrate the rationality and rationality of the indicators selected in this article.scientific.Second,the implementation of structural reforms on the energy supply side has a significant negative impact on the average rate of return of listed companies in the coal industry.Third,the impact of the implementation of structural reforms on the energy supply side on the average rate of return of listed companies in the coal industry is positively correlated with the shareholding ratio of executives.Fourth,the impact of the implementation of structural reforms on the energy supply side on the average rate of return of listed companies in the coal industry has a significant positive correlation with the size of the company.Fifth,through empirical research,it is found that the prediction detection rate based on the PCA-SVM model is higher and the error rate is lower,which verifies the applicability and scientificity of the PCA-SVM model for listed companies in the coal industry.Based on the above conclusions,it provides a theoretical basis and relevant countermeasures for listed companies in the coal industry in financial risk early warning.