Study On Financial Risk Warning Of Listed Pharmaceutical Enterprises

With the implementation of China’s "14th Five-Year Plan",the transformation and upgrading of various fields have been accelerated,and more detailed reform programs have been introduced for the pharmaceutical and health industry,such as actively promoting the centralized procurement system for drugs and implementing Internet medical care,etc.These changes have brought opportunities for pharmaceutical enterprises and also revealed many problems in pharmaceutical enterprises,which are particularly prominent in financial data.These problems are especially prominent in the financial data,such as the unreasonable allocation of funds and the high distribution costs of the pharmaceutical supply chain.Therefore,in order to coordinate the security and development of the pharmaceutical industry,meet the people’s demand for high quality and high level of the pharmaceutical industry,and improve the financial risk response ability of enterprises,it is especially necessary to build the financial risk early warning model of pharmaceutical enterprises,strengthen risk prediction and take corresponding management measures.This paper first introduces the research background,significance and related literature.Next,the concept,characteristics,theoretical basis and model structure of neural network of financial risk are discussed.Then it introduces the current situation and characteristics of pharmaceutical industry,analyzes the causes of financial risks of pharmaceutical enterprises and the necessity of financial risk early warning.Based on the principles of systematic,comparable and moderate financial index construction,the BP and PNN neural network early warning model for pharmaceutical enterprises is constructed as follows: firstly,the eligible financial early warning sample enterprises are selected.Secondly,the selected financial indicators were classified into solvency,profitability,development capability,operating capability,cash flow,per-share category and capital structure with a total of 75 indicators in 7 categories.Then these raw data were pre-processed,and after passing the normality test,significance test and correlation test,the remaining 33 indicators were subjected to principal component analysis,and 10 common factors were extracted and substituted into BP neural network and PNN neural network for financial early warning analysis;finally,the specific indicators of J enterprises were substituted into the neural network model constructed above,and the accuracy of the model was checked and the results were analyzed,and the corresponding improvement suggestions were proposed.Suggest corresponding improvements.The main conclusions of this paper are as follows: first,the accuracy of PCA-BP model is 81.579%,which is lower than the accuracy of PCA-PNN model of 84.211%,so the PCAPNN model has better and more stable prediction.Then,in response to the four major problems identified in the research process,Company J makes management suggestions in four aspects: reducing debt risk,controlling sales expenses,increasing R&D efforts and improving operational capacity.Finally,the shortcomings of this paper’s research are proposed: there may be false disclosure of financial data;the study lacks consideration of the external situation of the company.