文档介绍：湖南大学硕士学位论文基于MCMC模拟的贝叶斯金融随机波动模型分析姓名:李峰申请学位级别:硕士专业:数量经济学指导教师:朱慧明20070401 I摘要经济或金融时间序列存在着普遍的波动性现象,而波动性是描述金融市场研究的一个核心问题。1986年Taylor提出了描述波动性的随机波动模型。近几年随机波动模型在我国得到了不断的发展,研究者提出了众多的扩展模型,例如厚尾SV模型,均值SV模型等。但是此类标准SV模型的扩展模型模拟我国金融时间序列的优劣程度并无确切标准。本文引入DIC准则,并运用贝叶斯方法,对SV族模型进行了比较分析。根据贝叶斯定理对SV族模型中的SV-N模型、SV-T模型、SV-GED模型、SV-MN模型和SV-MT模型进行了贝叶斯分析。构造基于Gibbs抽样的MCMC数值计算过程,借鉴国外先验分布的经验,然后通过WinBUGS软件对模型参数进行估计。再通过参数的贝叶斯估计值对上海股市和深圳股市进行比较分析。经过比较分析发现上海股市和深圳股市都表现出强的波动持续性,而上海股市比深圳股市具有更强的波动持续性,上海股市的噪声要比深圳股市的多。而深圳股市的波动水平比上海股市的要大,风险也更高。最后利用DIC准则对深证指数和上证指数在SV族模型下的模拟情况进行了比较分析。分析发现SV-MT模型更加深刻的描绘了我,波动聚集性以及尖峰厚尾的性质。模拟深证成指最优的为SV-GED模型,而SV-MN模型是模拟最差的。在模拟上证指数的收益率的SV模型中,SV-T模型是最优的,而SV-MN模型是最差的。因此,我们可以得出结论,我国股市的收益率存在明显的尖峰厚尾现象,应采用厚尾SV模型对我国股市进行分析。关键词:厚尾SV模型;均值SV模型;MCMC算法;Gibbs抽样;DIC准则 IIAbstract The economical or the finance time series have the universal phenomenon of volatility, but the volatility is a core research question which to describe a money market. In 1986 Taylor proposed the stochastic volatility model described volatility. In recent years the stochastic volatility model develops very fast in our country, the researcher proposed the lots of expanding model, for example heavy-tail SV model, SV model in mean and so on. But the quality of expanding model for simulating the finance time series does not have a conclusion. This article introduces the DIC criterion, using the Bayesian's theorem, pare the SV model system. The SV-N model, the SV-T model, the SV-GED model, the SV-MN model and the SV-MT model in the SV model system were analyzed according to the Bayesian theorem. Drawing on the experience of the parameters ‘priors distribution abroad. A Markov chain Monte Carlo algorithm procedure with Gibbs sampler was designed to estimate the models’ parameter through the WinBUGS software. Shanghai Stock market was analyzed through Bayesian estimated value of the pared to Shenzhen Stock market. After parative analysis we discovery Shanghai Stock market and Shenzhen Stock market all display the property