The Application Of Financial Market Risk Measurement Based On EVT And Copula

With the globalization of economics, integration of finance, intensification of competition, relaxing of restriction and innovation of technology, the global financial market has made a great change on fundamental and structure, which appeared in size extending and efficiency promoting. At the same time, the volatility of the financial market becomes more fiercely, the stabilization of financial system declined obviously. So the management and measurement of the financial risks have become key abilities for financial institutions and industrial and commercial enterprises, and it also is the kernel content of the financial engineering and modern financial theories.The process of the financial risk management is very complex, which includes risk identification, risk measure, risk management decisions and practice and the control four stages in sum. Financial risk measure is the most important of all, which includes the measurement of the extent and the range of the loss that caused by all kinds of risk.At present the measurement of risk in financial market includes the sensitive method, volatility method, VaR method, ES method etc. In recent years, the extreme condition of the market made appearance of the market risk measure and the description of the variants' asymmetry, nonlinear and the financial practice field has paid more attentions on it. It has appeared many new risk measures such as the extreme value theory and Copula method. The extreme value theory only considers the tail of the distribution, but not the whole distribution, so it avoids the assumption of the distribution and the extreme value theory can also describe the tail quantile very accurately, which will help us to deal with the fat tail problem. Copula links the joint distribution and the marginal distribution together, which provide us a method to measure the dependence structure. Because of not confining the marginal distribution, we can construct the multivariate distribution neatly with the Copula. So when we make models with Copula, we can do research on the joint...