基于Gibbs抽样的厚尾SV模型贝叶斯分析及其应用

我国的金融时间序列存在普遍的波动性现象,而波动性又存在尖峰厚尾现象。首先对反映波动性特征的厚尾金融随机波动模型(SV-T)进行贝叶斯分析,然后构造基于Gibbs抽样的MCMC数值计算过程进行仿真分析,最后利用DIC准则对SV-N模型和SV-T模型进行优劣比较。研究结果表明:在模拟我国股市的波动性的方面,SV-T模型比SV-N模型更优,更能反应我国股市的尖峰后尾的特性,并且证明了我国股市具有很强的波动持续性。

Bayesian Heavy-tailed Stochastic Volatility Model in Finance Analysis Based on MCMC Simulation

Our country's finance time series exist the universal phenomenon of volatility, and the volatility has the property of Peak and heavy-tail. The first is to analyze Bayesian heavy-tail finance stochastic volatility model reflecting the volatility characteristic. The second is to design a Markov chain Monte Carlo algorithm procedure with Gibbs sampler to carry on simulation analysis. At last the SV-N model and SV-T model in the quality were compared using the DIC criterion. The findings indicate that, in simulating the volatility of stock market of China, the SV-T model is superior to the SV-N model, which can characterize the leptokurtic of stock returns in stock market of China. It is proved that the stock market in china has a high persistence of volatility.