The Computing Research Of VaR And CVaR Based On GARCH Model In Shanghai And Shenzhen Index

The purpose of securities investment is to obtain benefits. In fact, revenue is alwaysaccompanied with risks in investment activities. Usually, the higher the income, thegreater the risks and vice versa. Accompanied by economic globalization and financialmarket integration，financial market is becoming more and more complex. Therefore,theeffective guard and the management of the financial risk is the urgently needed solutionmajor issue which the financial organ and the investor face together.In the financial macket，there are lots of methods for measuring investment risks.VaRis one of the most popular tecluuques for financial risk management and prevailing newbenchmark in the world. VaR method is used to measure the maximum possible loss of afinancial asset of a portfolio in a coming holding period, with the given confidencelevel.Since VaR is not a coherent risk measure, scholars put forward CVaR which is basedon further research of VaR. The CVaR meet consistency risk measurement and considerthe risk values of losses part which exceed VaR.This method gained acceptance andrecommendation in the academic circle.Today the research of the application of VaR andCVaR in Shanghai and Shenzhen Index is of great significance which contributes much tothe healthy development of Securities Market and the strength of the internationalcompetence.Firstly,the body section of this thesis outline the current relevant literature of riskmanagement of securities market, followed the financial risk definition and classifiedaccording to the different criteria. Secondly, this thesis introduces the theory of VaR andCVaR systematically including the computing method and their respective advantages anddisadvantages.Then, this thesis introduces the relevant theory of GARCH model.Lastly,this thesis empirical study the VAR and CVaR with GARCH models in risk managementof Shanghai and Shenzhen index based on different samples, quantitative studying theapplication of the new model and compared with the traditional VaR and CVaR modelfrom the risk forecasting.