Study On Financial Market Risk Measurement Based On Extreme Value Theory And CoVaR Model

With the increasingly closely relation of the global financial market,the connections among financial asserts have become closely.Some signal market has been combined into the information network with the help of advanced technology,making the information transfer more and more quickly.It is much easier for individual financial market risk to form systematic risks by the high degree of market linkage.Such extreme events taking place from Latin American crisis,Black Monday in the United States,Japan's stock market crisis,Europe's currency crisis to Mexican peso crisis,Asian financial storm and American subprime crisis and so on,have led to sharp fluctuation of global capital markets and great destructiveness for the universal economy and finance.As a result,numerous financial companies go bankrupt and the real economy was in recession.Thus,risk management becomes much more important in the developed market and diversified investment channels.Financial Risk Management is one of the core contents of modern financial theory.Traditional theory of Financial Risk Management,relying on VaR model,which is connected with relevant financial volatility model and Copula theory,has achieved good results.But in terms of extreme risk events occurring frequently,the accuracy is not high to apply this method.Based on it,late researches bring in extreme value theory and overall effects still need to improve further.So,how to measure financial market risk effectively,especially for managing extreme value theory effectively is increasingly becoming an issue for financial institutions and investors.Currently,national risk management is at the stage of primary development,so financial risk management is based on borrowing and improving overseas relevant risk management theories and models to upgrade anti-risk capability of financial entities.Owing to the extreme risk features of financial market and nonlinear correlation of financial asserts,this paper will build the model of risk measurement to reflect the fluctuation of financial market based on assimilating existing relevant theories and research achievements,along the streamline of financial risk measures,from the risk measurement of signal assert to multiple asserts,fromVaR to CoVaR,by bringing in the organic bond of the model of quantile regression,typical factual financial volatility model,extreme value theory and Copula.Thus,based on improvement of financial market risk measurement model is the research speciality of this paper.Its mainresearch content and innovation as follows:Firstly,based on the improvement of financial market risk measurement between quantile regression model and extreme value theory.Firstly apply QR-GARCH to fitting the return rate of financial asserts on account of getting volatility and residuals,then introduce EVT,finally establish the extreme risk measurement model based on QR-GARCH-EVT.Meanwhile,introduce CAViaR model of quantile regression and establish the extreme value measurement model based on CAViaR-EVT model.The advantages of this model combination mainly shows that quantile regression need not to assume the distribution of the return rate of financial asserts in advance and it has good statistical property.What's more,EVT model is much more suitable for high quantile predication of heavy-tail distribution and the results of predication are rather stable.So,this paper uses the respective advantages both quantile regression model and EVT model to recombine and establish a new risk measurement model,which measures extreme risks.Secondly,use CoVaR to measure the risk overflow intensity and conduction effects of Chinese petroleum futures market and bulk commodity futures market at home and abroad.From the point of research content,the existing risk overflow effects researches focus on securities market much more and only a few researches refers to the risk overflow effects of crude oil futures market and relevant market.Aiming at the risk overflow effect of Chinese crude oil market and other commodity market,especially at the extreme situations,the researches are quite less.This paper considers financial features of crude oil futures market,by improving GARCH and applies Beta-Skew-t-EGARCH and EVT fitting its volatility.Based on it,we introduce Copula to describe nonlinear dependent structure of Chinese crude oil and bulk commodity futures market at home and abroad,and measure risk overflow effects at the extreme conditions in order to provide more clear perception of risk contagion for investors and more rational decision basis for risk supervision department.Thirdly,based on investment portfolio risk measurement improvement of CoVaR.By improving efficient frontier of Markowitz,we can consider the factor of changing the return rate of individual object assert into systematic risk measurement,then apply CoVaR to measure systematic risk diffusion and establish a new assert allocation model based on Mean-CoVaR.The main advantage of this model is that we consider the results of risk diffusion into the unified analytical framework optimized by investment assert portfolio,and that it contributes to reducing the risk of assert allocation portfolio.