Comparison Of The Risk Management Approach Based On The VaR And Empirical Test In The Stock Market

Securities market risk management is based on a measure of the stock marketrisk,We need to build an appropriate model and use appropriate methods, measuringSecurities Market Risk.This research is a hot topic at the head of the financial researchareas.Securities market risk measurement models and estimation methods are diverse,which is the most widely used is the Value-at-Risk method.VaR is a method to measurethe expected maximum loss given a certain time interval and confidence level, in anormal market environment.The VaR method provides a comprehensive measure ofmarket risk, and is built on a reliable scientific basis.The VaR method applies to thecomplex portfolio, the portfolio leverage effect and dispersion effect. First of all, it canbe provided to security holders risk quantitative indicators to guide the formulation of theinternal decision-making; followed when making investment decisions, VaR can also beweighed against the anticipated risks and benefits.On the theoretical side, the VaR method to do a more detailed description.First, theVaR calculation methods were described in detail at different levels,Including under thenormal distribution, t-distribution and the GED distribution.After this article details thevarious variance prediction model, on the basis of previous studies, the combination ofall models are introduced;After the third chapter, various calculation methods andvariance of the prediction model were more comprehensive theory comparison,whichalso briefly two other commonly used risk measurement methods, and made a simplecomparison of VaR method.Empirical test, the Shanghai Composite Index as the sampledata in different distribution under which the eight most effective model is applied to theempirical,by comparison of the results of the data output in the24models estimatedthrough the analysis parameters VAR calculation results of testing.The empirical resultsshow:Compared with normal distribution and t distribution, GED distribution able tobetter characterize the Shanghai Composite Index volatility,and under the GEDdistribution,the most appropriate model is PARCH model.