The Research And Forecast Of Stock Price Volatility Based On The Multi-Fractal Theory

The stock market is a vital and essential component of the securities industry andthe financial industry, and it is widely concerned by the investors. So it is veryimportant to forecast the stock price effectively and accurately in the financialinvestment. The study of the stock price fluctuation law and its prediction is a veryimportant issue for many scholars.Fractal theory is a very active branch of nonlinear scientific research. It studies onthe smooth and irregular geometric form of nonlinear system. Multi-fractal researchesthe ultimate overall characteristics from the local system. It mainly uses statisticalphysics methods to discuss the probability distribution.This paper mainly employs multi-fractal theory to study the time series of stockprice fluctuations law. First, the multi-fractal analysis on the stock market of Chinaprove that stock price fluctuations obey multi-fractal random walk, and then weresearch on the Multi-spectrum characteristics change before and after the continuedstrong fluctuations time sequence of stock prices through empirical analysis. Then, westudy the relationship between the Multi-spectrum parameters and price changes inthe time series, thus proving that Multi-spectrum parameters has some predictivecapability. Finally we construct the stock price volatility forecast model, the clusteringforecast on the Multi-spectrum parameters access to the higher accuracy forecastresults, It show out a new research method on the stock price forecasting.