Research On The Impact Of Corporate Fundamental Data On Stock Price Fluctuations Based On Neural Networks

With the implementation of the "Made in China 2025" strategy,there are currently 156 listed companies in this plate of the stock market,and this plate has become an increasingly popular investment destination.Based on BP neural network,this paper studies the impact of stock market fundamentals data on stock price fluctuations,including static research models based on elastic analysis neural networks and dynamic research models based on autoregressive neural networks.The core of these two models is the input and output design of the BP neural network.The public factors obtained from the fundamentals data are used as the network input,and the stock price fluctuation is used as the network output to build a static analysis and dynamic analysis model.In the analysis model,through the different changes of the fundamental influence factors,explore the different degree of influence on stock price fluctuations.At the same time,the past value of the volatility element is added to the network input to build an autoregressive neural network to further analyze the stock price fluctuation.The article uses 156 listed companies in the "Made in China 2025"plate from April 2015 to September 2019 as the research sample.Based on research,it is found that "Made in China 2025" fundamental data of listed companies in the plate have a significant impact on the daily high and low spreads of listed companies and the return difference of individual stocks.The article analyzed the fundamental data through factor analysis to obtain 9 public factors,and classified the 9 public factors to obtain 6 categories of factors,including the scale factor,cash flow factor,profit factor,debt service factor,growth factor.Based on the elastic neural network analysis,when the nine factors change individually,the increase in asset growth factors and profit factors can increase the difference between the highest and lowest prices,increase the volatility of the stock price,and the increase in asset growth factors will also increase the return difference of individual stocks of the stock to increase the volatility of the stock price;when the six categories of factors are individually changed,the increase of the growth factor will increase the difference of the highest the lowest price,increasing the volatility of the stock price,and the increase of the scale factor will reduce the difference in returns of individual stocks,making the stock price volatility smaller,when the six categories of public factors increase the change in combination,the combination of the growth factor and the debt servicing factor,cash flow factor,and return factor combined with other factors(excluding the scale factor)can increase the difference of the highest and the lowest price,increasing the volatility of the stock price,and the combination of scale factor and growth factor,and the combination of return factors,operating factors,and debt servicing factors will increase significant reduction in return on stocks and will reduce the volatility of stock prices.Using the constructed autoregressive neural network,the current period of stock price fluctuations and fundamental data are used as network inputs,and the next period of stock price fluctuations are used as network output.Using Weichai Power Co.Ltd.and TCL Group Co.Ltd.'s 2015-2019 fundamental data and stock price fluctuation data as research samples,using autoregressive neural networks to predict stock price fluctuations,it was found that monthly fundamental data were used to perform stock price,the error value of the volatility forecast is smaller than it predicted by the quarterly data,and the prediction result is also more accurate.Through the above research conclusions,different investment opinions and suggestions have been provided for different types of investors,and relevant regulatory agencies have also provided opinions and suggestions.