VaR Analysis For Market Risk Of Crude Oil Price Based On Garch And EVT Modles

Due to the sharp fluctuation of oil price, the risk management of oil market has become more and more important.This dissertation studied the financial attributes of the oil market, quantitative analysis of oil price risk and the control of the oil price risk. In this thesis, we will do an empirical study mainly about Brent oil price and this will certainly make some reference to Chinese oil market risk management.Now Value at Risk (VaR) is the mainstream method of financial risk management and a tool which is widely used. This method is a more popular method used to measure risk, which has the characteristic of concision and perspicuity. But tradition VaR methods want to give a hypothesis of financial return data subjected to some distribution, which can reduce the reliability of the model. The normal distribution is often inadequate for the description of real financial data with heavy-tail distributions, however, especially very large quantile that interest to risk manager. The models of BMM and POT, which are based on the theory of extreme value, jus t analyze the distribution of the tail instead of the hypothesis. So it can avoid risk of model. At the aspect of researching the market volatility, GARCH model and its many relatives are well used for non-linear financial time series. So it provides favourable condition for the combination of GARCH models and VaR model.To begin with, this thesis introduced the knowledge about financial risk, reviewed the development of the technology of risk measurement, sumed up the present situation about domestic and foreign VaR research, introduced the concept and computed theorem in detail. Different market conditions, data level and precision have different computing methods of VaR, this thesis synthetically compared all these computing methods, analyzed the advantages and the disadvantages and pointed out the appropriate situation. Moreover, this thesis compared the accuracy of the GARCH family VaR, the extreme VaR and the GARCH-EVT VaR. Finally, we get a meaningful conclusion:GARCH-EVT model significantly improves the accuracy of VaR calculation.The main innovation of this paper is the introduction of Extreme Value Theory into the market risk measurement of crude oil price. And this thesis used a qualitative method to choose the threshold. Because Extreme Value Theory has a prerequisite for the application, we combined GARCH models with Extreme Value Theory to build up GARCH-EVT model. We try to get a more accurate risk measurement model.