文章摘要
Geng Zhixiang,Wang Chuanyu,Lin Jianzhong.Heavy Tailed Distributions of Financial Assets and Popular Risk Measures[J].The Journal of quantitative and technical economics,2013,30(2):49-64
金融资产厚尾分布及常用的风险度量
Heavy Tailed Distributions of Financial Assets and Popular Risk Measures
  
DOI:
中文关键词: -stable分布  厚尾  MDD、DaR和CDaR  蒙特卡洛模拟
英文关键词: alpha-stable distribution  heavy tails  MDD, DaR and CDaR  Monte Carlo simulation
基金项目:
Author NameAffiliation
Geng Zhixiang 安徽工程大学 
Wang Chuanyu 安徽工程大学 
Lin Jianzhong 上海交通大学 
Hits:
Download times:
中文摘要:
      本文首先运用正态分布、带有位置-尺度参数的t分布、logistic分布、极值分布、-stable分布和核密度估计对上证综指收益率分布进行拟合,结果表明核密度估计优于其他分布。其次,在进行尾部风险拟合和度量风险方面,通过设定相关指标,在显著性水平为1%时,-stable分布更适合衡量风险程度,在此基础上提出了调和-stable分布,并得到一个同构表示解。最后,本文给出了蒙特卡洛-stable分布模拟和经验值下的MDD、DaR和CDaR,并得到了模型值和经验值之间的乘离率。
英文摘要:
      The paper introduces normal distribution, t location-scale distribution, logistic distribution, extreme value distribution, alpha-stable distribution and kernel density estimation to fit the distribution of Shanghai composite index returns. The results show that the kernel density estimation fits the returns distribution better than others. Secondly, we define a fit index with focus on tail risk and degree index for risk measurement. The study shows that alpha-stable distribution is more suitable for degree in risk measurement at the 1% level. Moreover, we introduce the definition of harmonic alpha-stable distribution, and we find out a solution for isomorphic representation. Finally, MDD, DaR and CDaR are calculated empirically, and also in the Monte Carlo alpha-stable simulation. We obtain the bias between model-value and empirical-value.
View Full Text   View/Add Comment  Download reader
Close