论文部分内容阅读
本文考虑了资产收益率遵循广义双曲有偏学生t分布(GHST),及其与波动率误差项存在相关性等的随机波动率模型,以分析其对收益率条件偏度的作用。采用上证综指实证显示,遵循GHST分布且存在滞后相关的SV模型,能更好地捕捉和复制上证综指的关键特征。其中,超额峰度的生成归因于厚尾分布;GHST分布的偏斜参数对收益率负偏性提供了解释;理论上占优的滞后相关性设定并不存在偏度生成机制,但其与偏斜分布的结合,使模型绩效表现更好。
In this paper, we consider the stochastic volatility model that the return on assets follows the generalized hyperbolic partial-student t-distribution (GHST) and its correlation with the volatility error term to analyze its effect on the yield bias. Empirical evidence from the Shanghai Composite Index shows that following the GHST distribution and lagging-related SV model, the key features of the Shanghai Composite can be better captured and replicated. Among them, the generation of excess kurtosis is attributed to the thick-tailed distribution; the deflection parameter of GHST distribution provides an explanation for the negative bias of return; theoretically dominant lag correlation does not exist the skewness generation mechanism, Combined with skewed distribution, the model performs better.