论文部分内容阅读
CAViaR模型是常用的VaR估计方法之一,但通常面临参数估计和模型检验的困难.本文发展了贝叶斯CAViaR模型用于分析油价风险,并考察该模型在参数估计、模型选择、VaR预测等方面的作用.采用布伦特原油价格日数据,研究显示贝叶斯CAViaR模型有效控制了估计风险和模型风险,且具有较好的VaR预测绩效,优于传统CAViaR模型.本文同时指出,油价VaR存在自回归特征并受前期正负收益率的不对称影响.不对称斜率CAViaR模型有效刻画了油价VaR的动态变化模式.
CAViaR model is one of the commonly used VaR estimation methods, but it usually faces the difficulties of parameter estimation and model verification.This paper develops the Bayesian CAViaR model to analyze the oil price risk, and investigates the model in parameter estimation, model selection, VaR prediction The Brent crude oil price date data show that the Bayesian CAViaR model effectively control the estimated risk and model risk, and has better VaR forecasting performance, superior to the traditional CAViaR model.This paper also points out that the VaR There is an autoregressive characteristic and is affected by asymmetric positive and negative returns in the previous period.The asymmetric slope CAViaR model effectively characterizes the dynamic change of VaR.