Real Option Pricing Based On RBF Neural Network

With the continuous development and progress of the economy and society,the business operation and management means of the enterprise are also changing constantly,resulting in the constant updating of the company's value type.Enterprises are faced with more and more unknown factors,and the flexibility of management decisions is becoming stronger and stronger.This makes the analysis method used to evaluate the enterprise value gradually fail,mainly because it cannot estimate the hidden value brought by the variability of the company.Real options make up for this.It can estimate the hidden value of the company described above,thus obtaining more accurate value evaluation results.To provide investors with more accurate and effective information and enable them to make more intelligent investment decisions.This paper will study the real option from the qualitative and quantitative perspectives.Firstly,in qualitative research.As the development of real option is not very complete,there is no uniform definition.Based on the characteristics of real options,this paper summarizes a large number of literature and summarizes the definition and classification of real options.It fills the blank of the real option.Secondly,in the direction of quantitative research.Research on pricing of real options.According to its pricing similarity with financial options,the present value S of the subject matter,implementation price X,the rate of change of company value,execution time T and risk-free interest rate r are selected as the variables of the real option pricing model according to the financial option pricing model indicators.Collect the financial data of 55 listed enterprises in shenzhen stock exchange from the CSMAR database,and calculate the above index value.Then the good index is selected and the traditional real option pricing model--black-scholes pricing model is used to price the real option.Thirdly,considering the artificial neural network which simulates the thinking mode of human brain,its nonlinear characteristics are highly consistent with those of real option price changes.In this paper,RBF neural network algorithm is introduced into real option pricing.A real option pricing model based on RBF neural network algorithm is established.Carry out case analysis and numerical simulation.The model effect and convergence rate are analyzed and calculated.Finally,on the basis of RBF neural network real option pricing model,the implicit layer center iterative algorithm is improved.K-means clustering algorithm is introduced to improve the training efficiency of the model.The three models of real option are compared and analyzed.It is found that the real option pricing model based on RBF neural network algorithm has good performance both in effect andefficiency,and is superior to the traditional pricing method in nature.Therefore,it is of great significance to introduce RBF neural network into real option pricing.