基于贝叶斯MSSV-ST金融波动模型的股市特征及机制转移性研究

Study on Volatility Characteristics and Regime switching of Stock Markets Based on Bayesian MSSV-ST Financial Volatility Model

针对有偏厚尾金融随机波动模型难以刻画参数的动态时变性及结构突变的问题,设置偏态参数服从Markov转换过程,采用贝叶斯方法,构建带机制转移的有偏厚尾金融随机波动模型,考量股市不同波动状态间的机制转移性,捕捉股市间多重波动特性。通过设置先验分布,实现模型的贝叶斯推断,设计相应的马尔科夫链蒙特卡洛算法进行估计,并利用上证指数进行实证。结果表明:模型不仅刻画了股市的尖峰厚尾、杠杆效应等特性,发现收益率条件分布的偏度参数具有动态时变性,股市波动呈现出显著的机制转移特性,而且证实了若模型考虑波动的不同阶段性状态后,将降低持续性参数向上偏倚幅度的结论。

The skewed and heavy-tailed financial stochastic volatility model cannot characterize the dynamics time-varying skewness and structural breaks.For this problem,we employ the Bayesian method and construct the skewed and heavy-tailed stochastic volatility model with markov-switching to explore the stock regime switching feature between different volatility state to capture multiple volatility characteristics,where the skewness parameter is allowed to shift according to a markov switching process.With a prior distribution for the Bayesian inference,we design the corresponding Markov Chain Monte Carlo sampling algorithm to estimate model parameters and use data of SSECI to make an empirical study.Results show that the model not only can capture the volatility characteristics,such as fat tail and leverage effect of stock markets,but also find out that the skewness parameter of distribution of return exists dynamic time-varying characteristics.In the meantime,it indicates that the stock volatility presents an evident regime switching character.Moreover,we verify that considering different periodic stages will reduce the volatility persistence value.