RECOGNIZING OVERDIFFERENCED TIME SERIES
作者:
Ming Chun Chang,
David A. Dickey,
期刊:
Journal of Time Series Analysis
(WILEY Available online 1994)
卷期:
Volume 15,
issue 1
页码: 1-18
ISSN:0143-9782
年代: 1994
DOI:10.1111/j.1467-9892.1994.tb00173.x
出版商: Blackwell Publishing Ltd
关键词: Cointegration;inverse autocorrelation;unit roots
数据来源: WILEY
摘要:
Abstract.Differencing is often used to render a time series stationary. The decision of how much differencing to do is usually based on plots of data, the autocorrelation function or a statistical test. Hence, it may happen that an analyst mistakenly differences a stationary series. When that happens, the inverse autocorrelation function takes on a specific pattern. We characterize this pattern and discuss the behavior of sample estimates of the inverse autocorrelation function for such overdifferenced series.
点击下载:
PDF
(666KB)
返 回