|
| 文章编号：1672-3104(2012)01-0131-05 |
| |
|
ARCH模型族在深圳成指中的应用 |
| |
丁扬恺 |
| |
|
（浙江师范大学数理信息学院，浙江金华，321004） |
| |
| 摘 要： 金融资产收益率一直是经济研究人员和投资者关注的焦点，ARCH及GARCH模型族可以较好地拟合金融资产收益率序列存在的尖峰厚尾、波动聚集性以及杠杆效应等特征。本文收集深圳成指(399001)二十年的日收盘价，通过GARCH-M模型论证了市场中预期风险增加一个单位，就会导致其收益率相应增加0.112个百分点，收益率的波动冲击影响会持续很长一段时间。利用EGARCH模型说明深圳成指收益率存在着信息不对称性，利空信息的冲击使得波动的变化更加大一些。同时根据相应的统计检验量，发现EGARCH模型比GARCH-M模型具有更好的拟合度，故实际中投资者应选用EGARCH模型预测深证成指的收益率。 |
| |
| 关键词： ARCH效应；随机游走模型；GED分布；尖峰厚尾；波动聚集性；杠杆效应 |
| |
|
| |
Application of ARCH Model in Shenzhen Stock Index |
| |
DING Yangkai |
| |
（Department of Mathematics, Zhejiang Normal University, Jinhua 321004, China） |
| |
| Abstract: Financial asset yield alwalys has been the focus on economic researchers and investors, which can be fitted well by the models of ARCH and GARCH, with its features of leptokurtosis, volatility clustering and leverage effects. This paper collects twenty years close in Shenzhen stock index (399001). GARCH-M model demonstrates that the increase of the expected risk I a unit will lead to asset yield corresponding increasing 0.112 percentage points, and volatillity impact will last for a long time. EGARCH model demonstrates that Shenzhen index has the asymmetric information. Bad information impact makes the larger changes in volatility. In short, we find EGARCH model has better fiting degree than GARCH-M model by corresponding statistical test. Therefore, investors should use EGARCH model to predict Shenzhen Stock Index in practice. |
| |
Key words: ARCH effects; Random-walking model; GED distribution; leptokurtosis; volatility clustering; leverage effects |
| |
|
