论文部分内容阅读
组合风险的估计和预测一直都是风险管理中非常重要的一个方面。本文使用了利用高频数据信息的实现协方差矩阵、DCC-MVGARCH多元波动率模型、Risk Metrics模型和多元正交GARCH模型对沪深两市的指数资产组合风险在险价值的预测失败率进行了对比,并利用动态分位数检验方法对各模型的组合风险测度稳健性进行了对比研究。研究结果证明,基于高频数据的实现协方差矩阵模型能够显著提高组合风险测度的预测精度,且严格符合Va R置信区间所要求的失败率,能够很好地在提高资金使用效率与管理资产组合风险敞口间取得平衡。
Portfolio risk estimation and prediction have always been a very important aspect of risk management. In this paper, the covariance matrix, DCC-MVGARCH multivariate volatility model, Risk Metrics model and multivariate orthogonal GARCH model are used to predict the failure rate of the index portfolio risk in Shanghai and Shenzhen stock markets In contrast, the dynamic quantile test was used to compare the robustness of the portfolio risk measures of each model. The results show that the implementation of covariance matrix model based on high-frequency data can significantly improve the prediction accuracy of portfolio risk measure, and strictly meet the failure rate required by VaR confidence interval, which can well improve the efficiency of capital utilization and management of asset portfolio Balanced exposure to risk.