论文部分内容阅读
应用时变条件t-copula函数描述股票指数收益序列之间的时变相依结构。时变条件t-copula模型的难点在于如何设定时变相依参数的演化方程,本文建立了用于描述包含时变自由度在内的所有时变相依模型参数的演化方程。进而采用蒙特卡洛仿真方法计算了各种指数组合的VaR,分析了道琼斯指数与标准普尔指数组合风险的演化趋势,并对结果进行后验测试,结果表明,时变条件t-copula函数仿真估计VaR可以覆盖最大损失风险。
Applying the time-varying conditional t-copula function to describe the time-dependent dependency structure between stock index return sequences. The difficulty of the time-varying conditional t-copula model lies in how to set the evolution equation of time-dependent parameters. In this paper, we establish the evolution equations for describing all the time-dependent model parameters including time-varying degrees of freedom. Furthermore, Monte Carlo simulation method was used to calculate the VaR of various index combinations. The evolution trend of the combination risk of the Dow and S & P index was analyzed and the results were tested empirically. The results show that the t-copula function simulation of time-varying conditions VaR can cover the maximum risk of loss.