基于GARCH族模型的风险度量研究-应用统计专业论文.docxVIP

基于 GARCH 族模型的风险度 量研究 II 万方数据 大连理工大学专业学位硕士学位论文Risk 大连理工大学专业学位硕士学位论文 Risk Measurement Research based on The Family of GARCH Model Abstract In recent years，China’S stock market development iS rapid．the volatility of the stock market and the financial sector have attention problems．Portfolio investors focused on earnings at the same time is also very pay attention to investment risks．Many of the portfolio average have higher yield，but there is big risk of loss．For effective risk management of the portfolio iS very necessary．In practice，the commonly used method iS to the measurement of risk in the value of risk(VaR)method．the VaR for effective has important significance to identify risk，estimate risk，SO how to quickly，accurately and easily estimate the VaR value become more attention of many scholars and this become now popular field． Using time series analysis method iS the mainstream of the financiaI sector．the family of GARCH model can effectively describe the different variance of time series。For the stock market often show”fat—tail”distribution characteristics．and through different scholars’data from a large number of empirical studies have shown that financial markets have asymmetry， namely the”positive disturbarlce”and”negative disturbance”showed different sensitivity． With further research on this field，many scholars have improved the existing GARCH model iS put forward to deal with many”fat—tails”and asymmetric GARCH model．these models suitable for analysis of the data of the stock market． In this paper，based on the family of GARCH model to fit the Shanghai composite index and the SP index，we selected the GARCH．N，GARCH．t、GJR—GARCH and EGARCH different GARCH model respectively for data fitting．We select the best fitting effect model of forecast．Then according to the result of optinml fitting，we use the Cornish．Fisher expression estimates quantile of the real distribulion．and then calculate the two sets of data of VaR value。We used different methods to calculate