The Study On Statistical Characteristics Of Financial Market's Volatility

Since 1970s, profound change occurred in financial market because of the change of macro-environment and the global fluctuation obviously intensified. The global economic integration is helpful to communication and development of economic in the world countries. Resources will re-distribute in the range of the world and financial market will open strongly. At the same time, the global economic integration enhances the dependence among the financial markets and leads to interactive effect of volatility. This situation makes local undulation of any financial market sweep sometimes to other markets quickly and this will increase the risk of global financial market and so as to cause financial crisis.This helical progress leads to a increase of probability and volatility of financial market. The breakthrough development of information technology,modern finance theory and financial engineering increases the circulation efficiency of capital and information as well as leads to a breakthrough increment of trade variety,trade volume and transaction velocity. But at the same time, it also heats up the volatility of financial market. So using scientific methods and means to measure financial volatility has a very important theory and practical significance value.The focus of this paper is the studying of volatility of financial market. This paper describes the variation theory of financial market and the properties of volatility, then we can understand the characteristics of volatility deeply and this makes a foundation for measuring of volatility later. Basing on the study of experts home and abroad, this paper quotes some methods and models to the china financial market and the research object is HS300 index. This paper analyzes HS300 index return series and the statistical characteristics of volatility series and then draws a conclusion that the two series are all non-normal. So the model selection is very important in this paper. Generally speaking, several kinds of models are usually used to estimate the volatility. They can be divided into two types: liner models and non-liner models. Before estimating the volatility of the day return of HS300 index using model, this paper also makes stationary test to the volatility and return series using the unit root method. By the empirical analysis, we can draw a conclusion that the two are all stationary time series.This paper uses several representative models to estimate the volatility of day return. Through analyzing the forecasting effect, we can find that the forecasting effect of GARCH-M (1, 1) is the best. Then modeling the effective drawing of actual volatility, the result testifies the correctness of empirical analysis. At the last, the paper puts forward some suggestion according to research range and depth.