风险度量VaR贝叶斯估计的渐近行为

asymptotic behavior of Bayesian estimation of VaR in risk measureAbstract Nowadays,with the rapid development of the world's economy and thedevelopment of economic liberalization,financial risks have been concerned by more andmore scholars and researchers at home and abroad.As a key part of risk management,risk measurement is very important to study risk measurement in the current economicglobalization.VaR's research and proposals have made a huge contribution to risk management,but with the progress of research,researchers have found that VaR has certain limitationsin many cases.In order to solve the situation that VaR cannot be applied in certain fields,scholars from various parties began to try to find other risk measures,and then putforward some other risk measures,such as CVaR,TVaR,entropic risk measures and soon.Risk is always the uncertainty of the problem,and risk measure is the way to expressand analyze this uncertainty.In fact.the distribution of the risk random variable has acertain dependence relationship with certain risk parameters.Therefore,it is necessary toestimate and analyze these specific parameters,and at the same time to explain the limitpropertiesAt present,there are many estimation methods available,such as maximumlikelihood estimation,non-homogeneous reliability estimation and homogeneous reliabilityestimation.In addition,there is Bayesian estimation,which has many limiting propertiesfor estimation,such as the commonly used large the principle of deviation,stronginterchangeability and asymptotic normality,and the principle of medium deviation.By analyzing the asymptotic behavior of the Bayesian estimated VaR risk metric,theBayesian estimate of the risk metric,and further combining the Bayesian estimate of therisk metric given by the exponential-gamma model,it is verified that the estimate satisfieslarge deviations and compatibility in principle.Keywords risk measure large deviation principle central limit theorem