Research On Internet Enterprise Valuation Based On Exponential Smoothing Neural Network

With the promotion of the Internet and the development of technology,a large number of Internet companies have gradually entered the public’s field of vision.At present,the Internet and Internet products have entered all aspects of people’s daily life.Under this background,Internet companies have shown great potential for development.The development of the Internet industry is inseparable from the promotion of the economy.Therefore,the academic and physical circles have conducted intense discussions and analysis on the evaluation methods of Internet enterprise value.Accurate valuation results of Internet companies can help investors make decisions,and help companies go public,mergers and acquisitions,and reorganizations.Compared with traditional enterprises,Internet enterprises have the characteristics of high uncertainty of future cash flow,large proportion of intangible assets,and serious Matthew effect.According to the research and analysis in this thesis,there are many deficiencies in the traditional valuation method in evaluating the value of Internet companies,and it is difficult to fully reflect the overall value of the company.At present,the DCF model is the mainstream method in the enterprise value evaluation method.The DCF model selects the appropriate discounted value to calculate the present value of the forecast results of each period according to the forecast results of the company’s cash flow,and the enterprise value is the sum of the present values of each period.Compared with other traditional valuation models,the DCF model can better reflect the overall value of the enterprise when evaluating the value of the enterprise with relatively stable cash flow.However,the accuracy of the DCF model for the valuation results of assets with high instability is insufficient.Based on this,this thesis introduces the B-S model.The B-S model can reflect the uncertainty and risk of value,and make up for the lack of the ability of the DCF model to evaluate potential value.Through the analysis of the value composition of Internet companies,the value of Internet companies can be divided into existing asset value and potential value.Based on this,this thesis constructs a combination model of DCF and B-S.The DCF model is used to evaluate the existing assets and the business value that tends to be stable.The B-S model is used to evaluate the technologies under development and intangible assets that may generate profits in the future.value.When selecting the parameters of the valuation model,it is necessary to avoid the subjective selection of parameters as much as possible to ensure the objectivity of the valuation results.Therefore,this thesis introduces the exponential smoothing method and BP neural network.The exponential smoothing method has a good predictive ability for time series,and the BP neural network has a strong ability to predict nonlinear data.Therefore,this thesis constructs an exponential smoothing neural network to improve free cash flow and volatility.This thesis takes Kunlun Tech Co.,Ltd.as an example,combines the financial situation in recent years,uses the exponential smoothing neural network to predict the operating income and stock price,and calculates the overall value of the enterprise through the DCF model and the B-S model.Finally,the robustness test of the value evaluation results is carried out.The results show that the value evaluation model adopted in this thesis can more accurately grasp the value of Kunlun Tech Co.,Ltd..This thesis is mainly divided into the following six parts: The first chapter is the introduction,which expounds the research background and significance of this thesis.The research results are sorted and summarized,the research ideas and methods of this thesis are expounded,and the overall framework of this thesis is determined accordingly.The second chapter is an overview of the Internet business valuation theory,and expounds the related concepts of Internet companies and the Internet business valuation theory.The third chapter is the construction of the enterprise value evaluation method based on the combination model.It summarizes the principles and characteristics of the DCF model,B-S model,and exponential smoothing neural network model used in this thesis.A method to improve the combined model of DCF and B-S by using exponential smoothing neural network is proposed,and the advantages of the improved model are summarized.The fourth chapter introduces the case company-Kunlun Tech Co.,Ltd.,introduces the basic situation of Kunlun Tech Co.,Ltd.,and analyzes the applicability of the combination model to Kunlun Tech Co.,Ltd..The fifth chapter uses the DCF and B-S combination model based on exponential smoothing neural network to evaluate the value of Kunlun Tech Co.,Ltd.,and carries out the robustness test of the evaluation result.Chapter 6 draws conclusions and prospects based on case analysis and related theoretical foundations.The main conclusions of this thesis are as follows:(1)According to the comparison between the valuation results of this thesis and the actual stock price,it is found that the combination model of DCF and B-S can fully reflect the value of Internet companies.(2)According to the prediction error comparison of the model,it is found that the exponential smoothing neural network improves the prediction accuracy of free cash flow and volatility.(3)Machine learning methods have strong data analysis and learning capabilities,which can play a role in optimizing parameters and make the valuation results more accurate.(4)Each valuation model has its own advantages and disadvantages,and a combined model of multiple valuation methods is beneficial to make up for the limitations of a single model.