ETF Option Value Evaluation Under Non-Normal Distribution Price Change Conditions

Financial asset assessment is an important part of asset valuation.With the advancement of science and technology,the evolution of assessment methods has become more and more frequent:from the early hypothetical conditional school to the mid-term data distribution school,the development has become a school of supervised learning;In the classic option evaluation,both the BS model and the Monte Carlo model assume that the data obeys a unary normal distribution(geometric Brownian motion).In order to study the value evaluation of ETF,it is found that the sample data(50 trading type open index fund)50ETF sometimes does not conform to the one-dimensional normal distribution.To thoroughly study the value evaluation of ETF,it is necessary to overcome this classic.The data distribution phenomenon of the model,starting from the weighted stocks of the 50 ETF,did a pilot study-the price fluctuations of the heavyweights were found as follows:1)The trading price of the stock group with a small market value does not conform to the geometric Brownian motion for a long time.(2)The stock group with large market value will conform to the law of piecewise geometric Brownian motion over time:the parameters ?.and ?^2 of geometric Brownian motion are different in different time periods;(3)The market value is extremely large.During certain periods of the stock group,some stock trading price fluctuations are in line with the geometric Brownian motion,and some stocks are not in conformity with the geometric Brownian motion law;and after a period of time,the original trading price fluctuations are extremely inconsistent with the geometric Brownian motion law.The transaction price will be very consistent with the geometric Brownian motion law;the original ones are very consistent with the geometric Brown Moving the stock trading prices become totally inconsistent with the geometric Brownian motion.Based on the discovery of this data distribution,this paper makes a theoretical analysis based on the characteristics of ETF market.In order to model the special evaluation method of option evaluation in this price change situation,this paper firstly makes a prediction model based on Jason's option value.It is found that this method can effectively find the change of the value of the option.Secondly,before and after the time of the price change,this paper finds the "asymmetric probability density"through probability analysis,and selects the noiseless features multiple times.The price change condition is based on the ETF option value evaluation of the LSTM convolution based on the fast cycle neural network(hereinafter referred to as Faster-RCNN).This new method has the characteristics that the late option value evaluation does not have.Finally,the model of this paper is a supplement to classic Monte Carlo model in the price change,and based on this supplement,this paper makethe s a joint value assessment reasoning(hereinafter referred to as joint reasoning)evaluation on the 50ETF value assessment.The method found that the accuracy of the joint evaluation method was 20%more accurate than the single Monte Carlo model.The supplementary model of this paper is a supplementary model based on price movement evaluation of Monte Carlo option evaluation.This kind of supplement and perfection establishes a good joint methodology for market fairness and fairness.