基于ARCH-Expectile方法的VaR和ES尾部风险测量

发布时间：2018-07-21 19:34

【摘要】：甄别和确定风险因素的贡献是资产或资产组合风险管理的重要研究内容。近十年,下端风险越来越受到关注,在险价值(Value at Risk,VaR)和预期不足(Expected Shortfall,ES)是资产组合风险管理中两个常用的风险度量工具。Kuan等[1]在一类条件自回归模型(CARE)下提出了基于expectile的VaR度量-EVaR。本文扩展了Kuan等[2]的CARE模型到带有异方差的数据,引入ARCH效应提出了一个线性ARCH-Expectile模型,旨在确定资产或资产组合的风险来源以及评估各风险因素的贡献大小,并应用expectile间接评估VaR和ES风险大小。同时给出了参数的两步估计算法,并建立了参数估计的大样本理论。最后,将本文所提出的方法应用于民生银行股票损益的风险分析,从公司基本面、市场流动性和宏观层面三个方面选取影响股票损益的风险因素,分析结果表明,各风险因素随股票极端损失大小的水平不同,其风险因素的来源及其大小和方向也是随之变化的。
[Abstract]:Identifying and determining the contribution of risk factors is an important research content of asset or portfolio risk management. In the last decade, the lower end risk has been paid more and more attention. In this paper, value at risk and expected risk at risk are two commonly used risk measurement tools in portfolio risk management. Kuan et al. [1] under a kind of conditional autoregressive model (care), a VaR metric based on expectile is proposed. In this paper, the care model of Kuan et al. [2] is extended to the data with heteroscedasticity, and a linear ARCH-Expectile model is proposed by introducing the arch effect. The purpose of this model is to determine the risk source of the asset or portfolio and to evaluate the contribution of various risk factors. The risk of VaR and es is evaluated indirectly by expectile. At the same time, a two-step parameter estimation algorithm is presented, and a large sample theory of parameter estimation is established. Finally, the method proposed in this paper is applied to the risk analysis of the stock profit and loss of Minsheng Bank. The risk factors affecting the stock profit and loss are selected from three aspects: company fundamentals, market liquidity and macro level. The results show that, The source, magnitude and direction of the risk factors vary with the level of the stock extreme loss.
【作者单位】：对外经济贸易大学金融学院应用金融研究中心;中国科技大学数学科学学院;中国科学院数学与系统科学研究院;上海财经大学统计与管理学院;
【基金】：国家自然科学基金重点资助项目(71331006);国家自然科学基金资助项目(71203025,71271128) 教育部创新团队计划上海财经大学创新团队支持计划资助
【分类号】：F832.33;F832.51;F224

【参考文献】