| 基于Gibbs抽样的贝叶斯超高频金融数据协整关系研究 |
| |
| 引用本文： | 李素芳,朱慧明,虞克明,郝立亚.基于Gibbs抽样的贝叶斯超高频金融数据协整关系研究[J].统计与信息论坛,2010,25(10):40-44. |
| |
| 作者姓名： | 李素芳 朱慧明 虞克明 郝立亚 |
| |
| 作者单位： | [1]湖南大学工商管理学院,湖南长沙410082 [2]布鲁内尔大学数学系,英国伦敦UB83PH |
| |
| 基金项目： | 国家自然科学基金，国家自然科学基金重点项目，教育部留学回国人员科研启动基金 |
| |
| 摘 要： | 针对传统协整检验不能适用于具有随机性特征的超高频金融数据的问题,构建贝叶斯超高频金融数据协整模型,结合参数的后验条件分布设计Gibbs抽样方案,提出基于超高频金融数据的贝叶斯协整检验方法,并利用中国股市超高频金融数据进行实证分析。研究结果表明：贝叶斯方法把参数看作随机变量的思想适合超高频数据随机性的特点,贝叶斯超高频数据协整方法能够不断更新参数信息,避免了OLS估计的有偏性问题,可以得到更符合实际的结论。
|
| 关 键 词： | 协整 超高频数据 贝叶斯分析 Gibbs抽样 |
Bayesian Co-integration Analysis of Ultra-High Frequency Data Based on Gibbs Sampling |
| |
| LI Su-fang,ZHU Hui-ming,YU Ke-ming,HAO Li-ya.Bayesian Co-integration Analysis of Ultra-High Frequency Data Based on Gibbs Sampling[J].Statistics & Information Tribune,2010,25(10):40-44. |
| |
| Authors: | LI Su-fang ZHU Hui-ming YU Ke-ming HAO Li-ya |
| |
| Institution: | 1.School of Business Administration,Hunan University,Changsha 410082,China;2.Department of Mathematical Science,Brunel University,London UB8 3PH,UK) |
| |
| Abstract: | The classic co-integration test is not suitable for the analysis of ultra-high frequency data because ultra-high frequency data are random.This paper employs Bayesian co-integration method to test ultra-high frequency data because more exact conclusion could be obtained through updating the information of parameters.Combined with Gibbs sampling,the empirical study on ultra-high frequency data of Chinese stock indices is carried out.As a result,it shows Bayesian co-integration test is feasible to conduct co-integration analysis for ultra-high frequency data,which could update the information of parameters to get more accurate conclusion,and avoid biased estimation of parameters to obtain more exact conclusion. |
| |
| Keywords: | co-integration ultra-high frequency data Bayesian analysis Gibbs sampling |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
