用广义帕累托分布估计极端尾部风险——以上海、深圳股市为例  

Estimating the Extreme Tail Risk by Generalized Pareto Distribution——Taking Shanghai and Shenzhen Stock Market as an Example

作　　者：郭森 孙志宾[1] GUO Sen;SUN Zhi-bin(North ChinaUniversity of Technology,Beijing 100144,China)

机构地区：[1]北方工业大学,北京100144

出　　处：《中小企业管理与科技》2018年第32期65-66,共2页Management & Technology of SME

摘　　要：在股票市场,那些频繁发生的风险往往不是最重要的,反而那些极小可能发生的事件会造成重大损失。因此,极端风险度量是至关重要的。论文以我国沪深股市的收益率为研究对象,基于POT的方法计算沪深股市的在险价值。论文在阈值的选取采用自动顺序测试程序,运用Jin ZHANG(2009)提出的一个新的估计方法来估计参数,这种方法改进了最大似然估计并且避免了计算问题。实证分析表明我国沪深股市的收益率序列确实是明显的尖峰厚尾分布,实证分析算出的在险价值表明深市的风险比沪市要高。In the stock market, those frequent risks axe often not the most important, on the contrary, those events that are extremely unlikely to occur will result in significant losses. Therefore,the extreme risk measurement is essential. Inthis paper, the return rate of Shanghai and Shenzhen stock market is taken as the research obj ect, and the risk value of Shanghai and Shenzhen stock market is calculated based on POT method. In this paper, an automatic sequentialtest program is usedto select the threshold, and anew estimation methodproposed by Jin ZHANG （2009）isusedto estimate the parameters. This method improves the maximum likelihood estimation and avoids the computational problem. The empirical analysis shows that the return rate sequence of Shanghai and Shenzhen stock market is indeed an obvious ＂peak fat tail＂ distribution. And the value at risk calculated by empirical analysis shows thatthe riskin Shenzhen is higher than that in Shanghai.

关 键 词：POT极值理论 在险价值 沪深股市 

分 类 号：F832.5[经济管理—金融学]