Stock Market Risk Measurement Based On Vine Copula-GARCH-VaR Model

With the continuous development and openness of financial market,people's investment in financial assets are more and more frequent,investors often choose the portfolio rather than a single asset to invest.Therefore,it is important to describe the financial assets between the dependent structure effectively to measure the risk,which is particularly important in preventing,resisting and defusing financial risks.This paper studies the portfolio of Shanghai Composite Index,Shenzhen Component Index,Hang Seng Index and S&P 500 Index.As the sequence of financial assets yield out of the peak and thick tail,fluctuations and other characteristics.In this paper,the GARCH model is used to characterize the yield,and the information process is analyzed.The descriptive statistical analysis of the time series data is carried out.Then,we introduce the theory of Vine Copula and traditional multi-Copula.We use the Kendall rank correlation coefficient to select the central variable of C-Vine Copula in each layer structure.Measure the inter-asset dependency structure and the correlation of the C-Vine Copula and traditional multi-Copula.The experimental results show that the fitting degree of Vine Copula is higher than that of traditional multivariate Copula function by AIC information criterion.In addition,the Monte Carlo simulation method is used to predict the risk.For the analysis of the 1000-day data for the multi-normal Copula function,the multi-Student t Copula function and the C-Vine Copula function,the simulation is based on the whole simulation.The results show that when the confidence level is 95%,the failure rate of the multi-Student t Copula function is close to the theoretical value.With the increase of the confidence level,for 97.5%and 99%,the confidence level of the unconditional VaR is compared with that of the simulated sample,the failure rate of the C-Vine Copula function is closer to the theoretical value.In general,the C-Vine Copula function can describe the complex form of correlation between markets more flexibly,and can measure the portfolio VaR of investors in the investment process more effectively.At the same time,this paper also calculates the VaR level of the four indices when the four indices and the equal weights are calculated.The VaR of the weighted portfolio is significantly smaller than the average of the four VaR.It can be seen that the diversified investment can indeed reduce the risk for investors'investment.As the correlation between financial markets may change with time and economic structural changes,this paper introduces the time-varying Copula model to describe the dynamic structural changes between the markets,the two indicators of the Shanghai Composite Index and the Hang Seng Index.And the time-varying characteristics of the two-asset yield series of the Shenzhen Index and the Hang Seng Index are used to analyze the advantages and disadvantages of the time-varying Copula function and its corresponding static Copula function.The empirical results show that the overall correlation between the Shanghai Composite Index and the Hang Seng Index has increased after the opening of"Shanghai-Hong Kong Link",but its relevance has become more unstable.We found that the time-varying function of the Copula function is better than that of the financial asset yield.It is shown that in the case of the optimal Copula function,the time-varying get better results."Shenzhen-Hong Kong Link" also has the same conclusion.At last,this paper summarizes the research,and expounds the shortcomings of this research and the future research direction.