基于ARIMA-GARCH模型对WTI指数的实证研究.pdf
文本预览下载声明
• 38 • 价值工程
基于ARIMA-GARCH模型对WTI指数的实证研究
An Empirical Study on WTI Index Based on ARIMA-GARCH Model
李 丽 L I L i
( 东北石油大学数学与统计学院,大庆 163188)
( School o Mathematics and Statistics, Northeast Petroleum University, Daqing 163188 , China )
摘要:国际原油是资本市场兵家必争之地,原油价格受很多不确定因素影响,且各个因素之间的相关关系错综复杂,因此要从理论
上彻底弄清楚原油指数的变化机理十分困难。然而原油指数是一个运动的、特殊的系统,它必然存在着规律。本文基于ARIMA-GARCH
金融时间序列理论,对 WTI波动率进行实证分析,经过平稳性检验、ARIMA参数选择、ARCH效应检验和GARCH模型优化,建立了
ARIMA-GARCH预测模型,通过预测值与真实值的对比认为ARIMA-GARCH模型可以很好拟合WTI波动率并且进行短期预测。
Abstract : Crude oil is a vital factor in the capital market. Crude oil prices are a ected by many uncertain factors. The correlation
between the various factors is intricate. Therefore, it is vei了 di icult to find out the change mechanism o crude oil index completely in
theory. However, crude oil index is a dynamic special system, so there must be law in it. Based on the ARIMA-GARCH financial time
series theory, this paper analyzes the WTI volatility. The ARIMA-GARCH forecasting model is established through the test o the stability,
ARIMA parameter selection, ARCH e ect test and GARCH model optimization. By comparing the prediction and real value , it is argued
that ARIMA-GARCH model can well fit WTI volatility and make short-term prediction.
关键词:ARIMA ; GARCH; WTI 指数
Key words : ARIMA ; GARCH; WTI index
中图分类号:F713.35 文献标识码:A 文章编号:1006-4311( 2017 )02-0038-02
0 引言 间序列模型,同时还有学者将机器学习算法应用于原油
时间序列分析是从一段时间上的一组属性值数据中 价格预测。张珣、汪寿阳深入分析原油价格形成机制及影
发现模式并预测未来值的过程。ARIMA模型是目前最常 响因子提出D A C 方法论对国际原油价格波动进行分析
用的用于拟合非平稳序列的模型,对于满足有限参数线形 与预测。
模型的平稳时间序列的分析,ARIMA 在理论上已趋成熟, 1 ARIMA-GARCH 模型
显示全部