Study Of Stock Price Estimating Method

Securities Market seems to be a mysterious and bewilderingly changeable place, especially in China, and there is full of magic power. We have learned some theoretical pricing models and methods, but they are not work well. We can't invest by means of them. Therefore, penman wants to find a new method to guide investment. In this thesis, penman tries to apply Tendency Extrapolation to estimate, but it doesn't work well. There are some complicated nonlinear-coupled effect factors, so some common methods can't solve this problems properly. While Artificial Neural Network (ANN) are adapted to these well. Therefore, we may sure that Neural Net Work is more superior theoretically. On basis of such theory, penman apply Back Propagation Algorithm to estimate a stock-SanXiashuili(600116) . Compare the estimated price with the real price and analyze the reasons, we point out the advantages and the disadvantages of the new method finally and get some satisfying results. Penman also develops a Stock Price Estimating System by Delphi5.