类似于二阶矩风险(方差风险)的时变性,高阶矩风险也具有时变性.同时,为讨论多个市场或多个金融资产对应高阶矩风险之间的关系,需要建立多元条件高阶矩波动模型.提出了多元GARCHSK模型并给出其向量表达,用独立成分分解技术来解决多元GARCHSK建模中的“维数灾难”问题,给出多元条件高阶矩波动率的估计方法.最后,利用该模型对我国股市4个主要股指的高阶矩风险进行了动态描述.
Higher moments risk has the time-varying character, which is similar to the character of the second moments risk, i.e. variance risk. To reveal the relationship between higher moments risk in different markets or different financial assets at the same time, it is necessary to establish multivariate conditional higher moments volatility model. Multivariate GARCHSK model and its vector expression are proposed in the paper. Based independent component analysis, estimation method for multivariate conditional higher moments volatility is discussed in detail to solve the problem of "dimension disaster" in multivariate GARCHSK modeling. Finally, the model is applied to describe the dynamic higher moments risk of the four main stock indexes in Chinese stock market.