| 基于不同误差分布下ARMA—GARCH模型的国债指数实证研究 |
| |
| 引用本文： | 谭常春,张勇,张虎.基于不同误差分布下ARMA—GARCH模型的国债指数实证研究[J].合肥学院学报(自然科学版),2011,21(3):16-21. |
| |
| 作者姓名： | 谭常春 张勇 张虎 |
| |
| 作者单位： | 合肥工业大学数学学院,合肥,230009 |
| |
| 基金项目： | 教育部重大专项，中央高校基本科研业务费专项 |
| |
| 摘 要： | 根据上证国债指数收盘价数据的特征，对其收益率序列建立误差分布为正态、GED和t分布的ARMA—GARCH模型．比较不同分布条件下的拟合效果，得出误差分布为GED和t分布时，模型的拟合效果优于误差分布为正态分布的情况．应用交叉验证法对预测效果进行比较，得出误差分布为t分布的ARMA—GARCH模型预测效果最优．同时，在模...
|
| 关 键 词： | ARMA—GARCH模型 国债指数 收益率序列 预测 |
Empirical Research of Treasuries Indexes Using ARMA-GARCH Models with Different Error Distributions |
| |
| TAN Chang-chun,ZHANG Yong,ZHANG Hu.Empirical Research of Treasuries Indexes Using ARMA-GARCH Models with Different Error Distributions[J].Journal of Hefei University :Natural Sciences,2011,21(3):16-21. |
| |
| Authors: | TAN Chang-chun ZHANG Yong ZHANG Hu |
| |
| Institution: | TAN Chang-chun,ZHANG Yong,ZHANG Hu(School of Mathematics,Hefei University of Technology,Hefei 230009,China) |
| |
| Abstract: | This paper establishes ARMA- GARCH model with series of returns, via analyzing the characteristics of the data of daily closing prices of treasuries indexes, considering the different cases of its error distribution based on Normal, GED and Student＇s t distribution respectively. Comparing its fitting result among the different error distributions, the model based on GED and Student＇s t distribution are more influential than the model based on Normal distribution. With the Cross- validation, ARMA-GARCH model based on Student＇s t distribution is the most outstanding on the predicting result. In the model comparing, with the same error distribution, ARMA-GARCH model is better in the aspects of fitting and predicting than GARCH model, and more suitable for the research of series of returns. |
| |
| Keywords: | ARMA-GARCH model treasuries indexes series of returns predict |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
