Measure Research On Financial Portfolio Risk

Since progress in the internationalization of the economy and integration of financial activities,The world economy becomes more closely linked,Which is not only in the mutual restrictions and influences of economic activities,but also in the interactions of financial risks.In order to improve the market competitiveness of products and obtain more benefits,Financial institutions actively promote products research and innovation,So that new types of financial industry risks occur frequently.Recently,the stock market" De-leveraging "suggests the need to manage the risks of the stock market.Interaction of financial risk makes it more difficult to be measured.Nonnormality of the distribution of financial time series makes that the traditional risk measure model applications have significant limitations.Copula method can flexibly construct portfolio multivariate distribution and measure financial market risks.Therefore,the paper research has theoretical significance and application value.Thesis focuses on the advantages of Copula-VaR risk measurement model.Firstly,Paper describes the purpose and significance of the investment by the research of capital supply and demand balance and maximize investor returns,then combines the relevant literatures.Secondly,Paper researches the theoretical basis of the portfolio model,Which includes the principle of GARCH model、Copula model and construction method of GARCH-Copula model.Then paper studies on the common risk measure methods.Copula-VaR model can be easier to capture the risk characteristics of the current financial.Finally,Paper measures the risk of the stock portfolio and it is proved that the Copula method is beneficial to the calculation of the VaR value.Paper study has got two conclusions.Sample data with economies in transition still has "Leptokurtic ","Fat Tail"," Volatility clustering" and " Leverage effects" features.On the basis of GARCH(1,1)model,Thesis chooses GARCH(1,1)process which has a fat tail distribution as the edge distribution model and portrays a deviation from the normal of financial time series.Paper calculates VaR value of each model by choosing different Copula connection functions to describe relevant structure.The empirical results show that in terms of the calculation of VaR,Copula technique is superior to conventional models,but the effect of different types of Copula model produces a big difference.Therefore,Selecting the appropriate model can study the risk of financial data more accurately.