Risk Measurement Based On Extreme Value Theory With Application In Interbank Offering Market

Interbank lending rates is the main benchmark interest rate in the currency mar-ket of China, looking for appropriate financial time series model to describe its random wave process, and choose appropriate risk measurement method to measure the inter-bank lending interest rate risk, for our country the interbank loan interest rate market risk management has theoretical and practical significance. Extreme value theory has been recognized as an important set of probable and statistical tools fot the modeling of rare events in that it plays a very important role in risk management as a major method of modeling and market risk. Because financial assets return data has clustering volatility and fat-tail characteristics, this paper will combine extreme value theory with GARCH model to build the GARCH-EVT modle. Firstly, an GARCH Model was built to fit the autocorrelation and heteroskedasticity of logreturn series, in order to obtain the residual sequence tha have appriximate i.i.d distribution. Then, POT method was utilised to analyze the innovations and estimate the interval of VaR, The parameters in POT model was solved by MLE. Finaly, Through bask testing, we find that our model is suitable to measure risk in the interbank offering Market in China.