The Cvar Measurement Based On The Study Of High Frequency Data

Var is a risk measurement techniques developed in recent years, because of its simplicity, comprehensive, practical and so widely welcomed by many organizations, has become the financial market risk management methods commonly used. There are many Var calculation methods, the more commonly used is the use of low frequency data, and exactly what benefit data Bing based on GARCH model estimation, then based on parameters calculated get Var value. But such data is bound to lose many days of low-frequency information, so that value can not get good coverage Var actual risk level.This study is based on starting such a problem, first low-frequency data based on family type model using ARCH Var and Cvar measure. Then the focus of this study is also the biggest innovations of the forthcoming high frequency data model (ACD-GARCH) introduced the field of risk measurement and risk measurement techniques Cvar combined measure of confidence of financial markets to some degree and distribution (GED distribution, normal distribution, T distribution) loss under the theory of risk, and low-frequency measurement results and previous data model, and Var technology measure to compare the risk of loss to see whether the value is closer to the actual loss, whether it is more good measure of risk. This process is the use of software such EVIEWS6.0 and MATLAB7.0 complete ACD-GARCH model estimation and use of programming measurement Cvar value. Final conclusion:the next five days with the actual loss of value comparison shows that both low-frequency or high frequency data under the Data Cvar Var higher risk than the precision measurement and high frequency data than the low frequency data under Cvar under Cvar has a higher risk of measurement accuracy.